The changes which man has made in the floral landscape are, as a rule, slight. The limits of the forest zone have scarcely been altered. The beech forest of the southern Andes seems to be less tenacious than the monte which surrounds the Pampa, and it has been ravaged by fire along the whole edge of the southern steppe at 37° S. lat. The work of man is generally confined to changing the primitive complexion of the natural formations, without altering their general appearance. Thus valuable essences are disappearing from the forest and the scrub, the larch and the cypress from the district of the Patagonian Lakes, and the red quebracho from Santiago del Estero.

A change that is scarcely visible, but is of considerable economic importance, thus takes place in the vegetation of the prairie owing to the presence of herds. The pasto fuerte, composed of rough grasses, which is the natural vegetation, is being succeeded by the pasto dulce, in which annual species, soft grasses, leguminous plants, etc., predominate. It is mainly composed of plants of European origin. The difference between the pasto dulce and the pasto fuerte or duro is so important for the farmer that there is hardly a single work on Argentina which does not dwell on it. The idea, however, that the pasto dulce has advanced steadily westward, starting from the vicinity of Buenos Aires and constantly enlarging its domain, is not strictly accurate. In 1895 Holmberg3 traced the western limit of the zone of the pasto dulce through Pergamino, Junin, Bragado, Azul, Ayacucho, and Mar Chiquita. When we compare this with earlier observations, we see that in the course of the nineteenth century the zone of the pasto dulce has extended by about a hundred miles on the southern Pampa. When Darwin travelled from Bahía Blanca to Buenos Aires in 1833, he found no pasto dulce except round Monte, on the right bank of the Salado. Further north, on the other hand, the extent of the pasto dulce does not seem to have altered appreciably. The expedition to the Salt Lakes in 1778 found that there were already thistles beyond the line of the ranches, and these are characteristic of the pasto dulce in the Chivilcoy region on the Salado, which was then abandoned to herds of wild cattle. "There was thistle enough to cook," says the journal of the expedition. The difference is connected with the history of colonization in the province of Buenos Aires, where ground was gained only toward the south between 1800 and 1875. Since 1895 the pasto duro has been eliminated by agriculture rather than by the feet of the herds. Hence the advance of the pasto dulce is no longer in a continuous line moving toward the west. It is sporadic, depending upon the construction of new railways which open up the plain to the plough.4

Colonization does more than emphasize the individuality of each of the natural regions. It connects together different features, and blends them in a complex vital organism which goes on evolving and renewing itself.

The occupation of the whole of the soil of Argentina by white colonists is quite a recent event. The second half of the nineteenth century was characterized by a rapid territorial expansion, and over more than half the country the expression "new land" must be taken literally. It is only one generation since it was taken from the Indians. There can be no question here of tracing the history of the relations between the white population and the free Indians of the Chaco and the Pampa. The most formidable of these were, in the north, the Abipones and the Tobas. On the Pampa, the foes of the colonists were Indians of Araucanian descent, Ranqueles, Pehuenches, etc., who came down from the mountains and took to horses. At the close of the eighteenth century the frontier of Buenos Aires was on the nearer side of the Salado, and was bordered on the south-east and north-west by the fortresses of Chascomus, Monte, Lobos, Navarro, Areco, Salto, Rojas, and Melincue. The proposal of D'Azara to extend it as far as the Salado was not carried out, and it was not until 1828 that there was a fresh advance westward.5

The new frontier, which would not be altered until 1875, passed by Veinte Cinco de Mayo and Blanca Grande, at the north-western extremity of the Sierra de Tandil. It included the entire region which lies between the Sierra de Tandil and the lower Salado, where the village of Tandil had been established in 1823. In addition, a line of forts stretched from Blanca Grande in the south-west to Bahía Blanca. The expedition sent in search of a port south of the mouth of the Plata had not found any nearer site that was suitable. But Bahía Blanca was to remain an isolated advance post until 1880, sharply separated from both the colonized zone of the Pampas and the establishments on the Patagonian coast.

While the cultivated area was thus growing toward the south, it was being reduced in the north of the province of Buenos Aires and the south of Córdoba. The lands of the lower Rio Cuarto were not occupied. About 1860 (Martin de Moussy) the farthest establishments in this sector were S. José de la Esquina and Saladillo on the Tercero. The road to Chile by the Rio Cuarto, Achiras, and San Luis was threatened. The advance of colonization in this zone was at first in the west to Villa Mercedes on the Rio Quinto. The line of the Rio Cuarto by Carlota was reoccupied, and before 1875 the frontier had been pushed back to the Rio Quinto, where it joined the forts of southern Buenos Aires by way of Sarmiento, Gainza, and Lavalle.

At last, in 1878, General Roca abandoned the classical methods of fighting the Indians, and took the offensive. He deprived the Indians of their refuges to the south of San Luis and the Central Pampa, and threw them back toward the desert. The Argentine troops followed in their steps as far as the Andes and the Rio Negro. There are to-day few traces in the immense territory that was won of the indigenous population. Its extreme mobility had masked its numerical inferiority.6

The history of the northern frontier is much the same. At the end of the eighteenth century the Spanish outposts ran along the course of the Salado. To the north of Santa Fé, at Sunchales, Soledad, and San Javier, they protected the direct route from Santa Fé to Santiago del Estero. These outposts were abandoned during the revolutionary period, and the Indians advanced as far as the suburbs of Santa Fé. The roads both to Santiago and, by the Quebracho Herrado, to Córdoba were cut.7 Urquiza reorganized the Santa Fé frontier, first as far as San Javier, then below 29° S. lat. between Arroyo del Rey on the Paraná and Tostado on the Salado. The expedition of 1884 brought the Argentine army as far as the Bermejo, and broke the resistance of the Tobas. The forts which, more to the north, guarded the province of Salta, on the further side of the Sierras de la Lumbrera and Santa Barbara, had been dismantled at the beginning of the nineteenth century, as the tribes in this part of the Chaco were not hostile.8

The memory of the fights with the Indians is so completely blotted out to-day, and the menace of invasion by the tribes has been so rapidly extinguished, that it is difficult to realize fully the profound influence they once had on colonization. The line of forts was a frail barrier that was constantly broken through. The Indians of the Pampa stole cattle from the ranches of Buenos Aires, and sold them in Chile. Colonel Garcia calculates in 1816 that about 40,000 animals were stolen every year.9 Colonel Roca gives the same figure in 1876. The Pampa put no natural difficulties in the way of the movements of the Indians, no points which might serve as bases for the frontier. D. Pedro Pablo Pabon points out that the proximity of the Sierra, instead of giving protection to outposts in the Tandil region, would be an additional source of insecurity, as it increased the difficulty of keeping watch. In the north the Indian incursions followed the clearings in the scrub, avoiding the dense and impenetrable parts. The lagoon of Mar Chiquita, to the west of Santa Fé, was a valuable rampart, in the shelter of which a fairly large population had established itself round Concepción del Tio.

The enlargements of the frontier were sometimes due to expansive movements of colonization, the breeders occupying new land beyond the line of forts and demanding protection, and sometimes to the arbitrary action of a Government which was eager to extend its territory, though it was still without the means of exploiting it. Roca has well shown the defects of this system of premature military occupation. "To go far away from the populated districts in acquiring new territory is, in my opinion, only an aggravation of the inconveniences of defensive war, and it places a desert between the new lines and the settled regions.... Invasions occur at once."10 We should therefore be likely to make serious mistakes if we were to identify the history of colonization with that of military occupation. Moreover, the garrisons of the forts did not take a very active part in the exploitation of the soil. The plan which D'Azara proposed, of making blandengues (lancers) colonists and rooting them to the soil by distributing it amongst them, seems to have been purely Utopian. His description of the frontier shows clearly how slight a hold the early colonization had on the Pampa, where the only relatively industrious element was represented by the groups of civilians (paisanos) who gathered about the works and moats of the forts. It was different on the Santiago del Estero frontier, where there was agriculture as well as breeding. Here the fort was identical with the village, and each soldier had his plot of wheat, maize, or water-melons.11

The provinces which were to combine in forming the Argentine Republic had no economic unity. They were really two countries, two separate worlds, the coast regions and the mountain regions (de arriba), joined together, but not blended, by the main road from Buenos Aires to Peru, by way of Córdoba, Tucumán, and Salta. They represented two different branches of Spanish colonization. "Two human streams," says Mitre, "contributed to the peopling of the vice-royalty.... The first came directly from Spain, the mother country. It occupied and peopled the banks in the basin of the Rio de la Plata, in the name of the right of discovery and conquest, and fertilized them by its labour. The other stream came from the ancient empire of the Incas, already subdued by the Spanish armies. This spread toward the interior of the country as it passed from the Pacific to the Atlantic, occupied the land in virtue of the same rights, and exploited it by means of a feudal system.... The same year, 1535, saw the foundation of the two towns, Buenos Aires and Lima, and was the centre of these two cycles of discoveries and conquests. Thirty-eight years later, in the same year, 1573, the Conquistadores who came from Peru founded the town of Córdoba, two hundred miles away from the Paraná, while those who came from the Rio de la Plata founded the town of Santa Fé on the banks of that river."12

Tucumán and Salta were established by conquerors from Peru, while San Juan and Mendoza were built by the Chilean Spaniards. The line of demarcation between the two zones of colonization crosses the immense desert plains of the interior, not the elevated tablelands of the Andes.

The two types of Argentinians differed in every respect, in blood as well as in environment. The indigenous race, which was eliminated on the coast, mingled intimately with the conquering race in the interior.

The establishments on the Rio de la Plata had originally been merely stages on the road to Peru, and had no value of themselves. The elevated tablelands of the Andes long remained the economic centre of Spanish America, and the provinces of the interior, which sold them cattle and mules, depended very closely upon them. The end of the eighteenth century was marked by more rapid progress in the region of the Pampas. The vice-royalty of La Plata was created.

Freedom of trade was secured between Buenos Aires and the Spanish ports. The export of hides increased. The influence of Buenos Aires spread over the interior and, in spite of the Córdoba tariff, reached the regions of the north-west. "The creation of the vice-royalty," says Dean Funes, "and the new direction taken by commerce had the effect that Buenos Aires became the centre of considerable and important business."13

This commercial development, which seemed destined to bring closer together the two halves of Argentinian territory, was interrupted in the first half of the nineteenth century. This did not, however, break the connections between the provinces to the north-west of the tableland and those on the Pacific slope, and indeed, they became more varied and more binding. Packs of mules, carrying the ore of San Juan and La Rioja to the foundries of the Chilean side, added life to the Cordillera. When Chile, transformed into an agricultural country, could not meet its own demand for cattle, the oases of the Argentine side were sown with lucerne for fattening the cattle which were to cross the mountains. The provinces of Mendoza, San Juan, La Rioja, Catamarca, Tucumán, and Salta were held within the orbit of the Andes districts.14 There are historical reasons for this set-back to the influence of Buenos Aires. The wars of the revolutionary period and the conflicts between the Buenos Aires Government and the maritime powers checked the commercial enterprise on the banks of the Plata. This political isolation of the province of Buenos Aires, under the Rosas Government, lasted until 1853. Poncel gives us statistics of the imports of Catamarca which show the great importance of this date in the history of Argentine commerce:

In 1854-5 the Cordillera route definitely ceased to be of commercial importance to Catamarca, and it was afterwards used merely for the export of cattle.

But the attraction of Buenos Aires after 1853 was not merely due to its commercial life and its intermediate position between the provinces of the interior and Europe. It was chiefly based upon the economic development of the region of the Pampas, which began about this date, and altered the balance between the two halves of Argentina. The exploitation of the Pampa, the improvement in breeding methods, and the introduction and expansion of agriculture on the plain of the Pampa, which fill all publications on modern Argentina, are in themselves one of the great events in the economic history of the nineteenth century. They had also an indirect but profound influence upon the life of other parts of Argentina. The consuming capacity of the Pampa increased simultaneously with its wealth and population. It absorbed the products of the neighbouring provinces and in turn made customers of them, distributing amongst them, according to the services they rendered, part of the gold it obtained from beyond the Atlantic. One after the other the provinces lost the relations which had hitherto connected them with foreign lands. There was the same development all over the zone of cereals and lucerne—the direction of the stream of commerce was reversed. In some places, as at Tucumán and Mendoza the change was accomplished a generation ago. In other places, as at Salta and San Juan, it is still going on. In yet other places, the more remote valleys, like Jachal and Santa Maria, it will occur in the near future. By a singular anomaly the Far West of North America, which sprang up half a century ago, tends to withdraw more and more from the influence of the eastern States, which provided it with capital and immigrants, while the Far West of Argentina, which is just as old as the east and by no means a creation of the east, since it developed in isolation and freedom, and was already adult and rich when they came into contact, has nevertheless fallen into complete dependence upon the east in the course of a few years.

The life of the whole country depended upon the great colonization movement which transformed the plain of the Pampas. This brought about an economic unity which was at once reflected in the political world. The railway from Buenos Aires reached Tucumán before 1880; Mendoza, San Juan, Salta, and Catamarca before 1890; and La Rioja before 1900. The establishment of closer economic relations between the coast and the provinces of the interior has nearly always inaugurated a period of great prosperity for the latter. In every case the influence of Buenos Aires vitalized them, put an end to their slumbers, and made them rich.

Not only did the coast take for itself the products of the western provinces, which had hitherto found their way to other markets, but new centres of production had to be created to meet its needs. The forests of the Chaco received a great influx of wood-cutters, to provide the sleepers for the railways. The valley of the Rio Negro was planted with vines, to provide the wines of the colonies in the district of Bahía Blanca. The attraction of the Pampa was felt as far as the frontiers. Paraguay competed with Corrientes in the supply of tobacco and oranges; with Misiones in the supply of yerba maté. Each district chose the particular crop which was best suited to its climate, in order to secure the highest possible advantage from its relations with Buenos Aires.

The two most brilliant satellites of the Pampa, the most important productive centres of the interior, are Tucumán and Mendoza. All the other important towns of Argentina belong themselves to the region of the Pampas. Tucumán and Mendoza, which live by supplying the Pampa with sugar and wine, have become in turn secondary centres of attraction. They are a sort of regional capitals, and they have their own spheres of economic influence. A network of commercial streams has developed about them, and this has led to the formation of new roads. These lines of local interest are easily recognized on a map of the railways, where one sees them superimposed upon the regular fan of lines which converges toward Buenos Aires. La Rioja provides the props for the vines of San Juan and Mendoza. From the north of Córdoba to Salta, a distance of about 250 miles, the wood is cut for the fuel of the sugar-works of Tucumán. Santiago dries the fodder for its troops of mules. The prairies of Catamarca, which once fattened the cattle that were intended for Chile, and often came even from Tucumán, now sell their beasts to the butchers of Tucumán. The wines of San Juan find their best customers at Tucumán. Even the nearest portions of the plain of the Pampas, to the north-east of Santa Fé and the south of San Luis, supply maize and wheat to Tucumán and Mendoza, instead of sending them to the ports for export.

While Argentina lives on the Pampa, the Pampa lives on export. It has been developed through the inflow of European immigrants, and Europe pays by sending its manufactured products and capital. Except as regards emigration, the United States had, before the war, much the same relation to Argentina as the countries of Western Europe. Thus the economic prosperity of the Republic binds it more and more closely to the life of the whole world. Its position in the temperate zone of South America had retarded its entrance into world-commerce, and this explains the slowness with which its colonization proceeded at first. Its climate and products were too similar to those of Spain. Not only the mining and metallurgical centres of the Andes and of Mantiqueria, but even the sugar and cotton regions of Brazil, the Antilles, and the Guianas, were developed before the plains of the Pampas.

The turn of the Argentine Republic did not come until the growth of population in the industrial countries of Europe made them dependent upon foreign lands for their food, and until the application of steam to ships made it possible to export wool, meat, and cereals on a large scale.

When we compare the economic organization of Argentina with that of the United States, we see that it is both less complex and less capable of being self-contained. The difference is due to the architecture of the country. I said at the beginning of this chapter that Argentina has no equivalent for the zone of the Atlantic tablelands, which is now the great industrial region of North America. The industrial prosperity of eastern North America provides a safe home market for the farmers of the west, and relieves them of the need of exporting their produce. Moreover, the Atlantic tablelands, the original centres of population, where the first generations of colonists lived on land that was often poor, have seen the gradual formation of reserves of labour and capital which were afterwards used in colonizing the west. The east sifted, in a sense controlled, the influence of modern Europe in the colonization of the United States. It classified and assimilated the new emigrants who set out for the west, mingled with the troops of native pioneers on their way to the prairies. In the same way, when European capital flowed into the United States, it found in the eastern cities a large treasury and a body of financiers in whose hands it had to remain.

In Argentina, on the contrary, everything speaks of the close and direct dependence of the country upon oversea markets. The soil itself bears the marks of this solidarity. It is seen in the network of the railways, the concentration of the urban population in the ports, and the distribution of the cultivated districts in concentric circles which are often limited, not by a physical obstacle, but by the cost of freightage between the productive centre and the port. Thus we get a geographical expression of facts which seem at first sight to belong to the purely economic or sociological order.

CHAPTER II
THE OASES OF THE NORTH-WEST AND PASTORAL LIFE IN THE SCRUB

The whole life and wealth of the arid provinces of north-western Argentina depend upon irrigation; the water-supply definitively settles the sites of human establishments. The water resources are irregularly distributed. They are especially abundant in the south (San Juan, Mendoza, and San Rafael), where the torrents of the Cordillera are fed by the glaciers, and on the outer fringe of the hills above the Chaco, at the foot of Aconcagua, which gathers masses of cloud and rain on its flanks (Tucumán). In the intermediate district, on the contrary, in the regions of La Rioja and Catamarca, and in the interior of the hilly zone to the north-west of Tucumán, the amount of available water is small; the oases shrink into small spots far removed from each other.

This natural inequality was not felt at first. For a long time the spread of cultivation and the progress of wealth were restricted only by the scarcity of population, the difficulties of transport, and the inadequacy of the markets. The best endowed oases paid no attention to the surplus supply of water, for which they had no use. We have to come down to the close of the nineteenth century to find men reaching the limits which nature has set to colonization, and mapping out their domain. It is not until then that La Rioja ceases to compete with Mendoza, or Catamarca with Tucumán. While large industrial enterprises develop at Mendoza and Tucumán, strong centres of urban life arise, the population increases, and immigrants stream in, the oases of the interior scarcely change. Their population does not keep its level. Life has an archaic character that one finds nowhere else in Argentina. The physical conditions have retarded, one would almost say crystallized, the economic development. The living generation exploits the soil in ways that to some extent go back as far as the indigenous tribes, the masters of their Spanish conquerors in the art of irrigation. The industry of fattening and convoying cattle, which was once the chief source of wealth of the whole country, is still alive in those districts.

The zone of the elevated tablelands of the Andes without drainage toward the sea—the Puna—has still, below 22° S. latitude on the northern frontier of Argentina, a width of about 250 miles. This breadth steadily contracts southward as far as 28° S. latitude, where the Puna ends about the level of the road from Tinogasta to Copiapo.

To the east and south of the Puna the Argentine Andes are cut from north to south by a series of long gullies and large basins, between which there are lofty and massive chains with steep flanks. Some of these lie in the heart of the mountains, while others often open like gulfs upon the edge of the plain. These depressions with rectilinear contours are a common feature of the topography of the Andes in this latitude. The central plain of Chile is closely related to them. In the Argentine speech they are called valles: Valle de Lerma, Valle Calchaqui, Valle de Iglesias, de Calingasta, d'Uspallata. They are, however, not "valleys" in the sense of hollows made by erosion by running water. They owe their formation to tectonic movements, subsidences of the surface. The scanty rivers of the arid Anacs are not capable of doing work of that kind. When they enter the already formed bed of a valle, they seem to be lost in the immense space. Often they dry up in it, leaving behind the sediment and salts with which the water was laden. In other places they cut at right angles across the valle, escaping by narrow breaches in it, while the depression continues its course on either side, taking in sections of a number of independent streams.

Opposed to the valle is the eroded ravine, carved out by water, the quebrada. It opens upon a valle with a V-shaped mouth, which widens out at the top, and one can recognise at sight the various slopes and the successive stages of erosion. Narrow and winding, a level bed of shingle filling the entire base of the valley, it rises rapidly toward the mountains and provides a route from the valle to the puna. These valles, quebradas and puna are the three inhabited zones of the Andes. The first is the richest. The inhabitant of the valle, proud of his comparative comfort, has for his neighbour in the quebrada or the puna—the coyada—a contempt such as one finds the inhabitants of the good land in Europe feeling for the people in poorer districts.

The narrower the valle, the less rain there is. The observations give 112 millimetres of rain per year at Tinogasta, 290 at Andalgala, and 200 at Santa Maria. Salta and Jujuy have a much moister climate, and have no less than 570 and 740 millimetres of rain annually. This is because the eastern chain of the Andes, which stretches from the Sierra de Santa Victoria on the Bolivian frontier to Aconcagua, sinks lower at the latitude of Salta, and lets in the moisture of the Chaco to the heart of the zone of the Andes. The rains of Salta and Jujuy are suspended during the winter, but they are so heavy during the summer months (November to March) that maize, which needs only the summer rain, can be cultivated without irrigation. But when we follow the Valle de Lerma southward from Salta the maize harvest becomes more and more uncertain, and it is no longer sown in dry soil when we get to about twenty miles from Salta, in the latitude of the confluence of the Arias and the Juramento. However, the summer rains, which are good for maize, are very injurious to the vine; they spoil the grapes. Thus the southern limit of the cultivation of maize in dry soil almost coincides with the northern limit of the vine. At that point we have the real beginning of the typical scenery of the valles.

The need of irrigation is due to the scarcity of rain, but it is accentuated by a number of causes which tend to increase the aridity. The valles are the scene of scorching day-winds, the zonda, like the Föhn of the Swiss Alps, which, there being no snow, dry up the water of the springs and of the irrigation trenches, or use the deposits left by the waters to form dunes, which they push southward, sometimes like veritable glaciers of sand. Moreover, the soil of the valles is generally composed of coarse and permeable alluvial deposits, which absorb the rain-storms immediately. There is at the foot of both sides of the hills which enclose each valle an immense and far-lying bed of imperfectly rounded shingle. This double zone of detritus is strangely desolate, for the vegetation on it is restricted to isolated bushes of jarilla and tola. From the sheepfolds on the mountains to the oases in the valleys one hardly meets a single house. The bed of the valley is not so desolate. A broad ribbon of sand marks the dry bed of a torrent, and on the clays of its banks, if the sheet of water underground is not too deep, one finds, in spite of the goats and asses and charcoal-burners, little forests of algarrobas, which the foundries use for fuel.

The modern alluvial beds, gravel and sand, represent the upper stratum of a considerable series of continental deposits which lie on the Paleozoic crystalline rock of the Andes.16 They chiefly consist of red sandstone and coloured marls, which crop up here and there through the alluvial covering and give the landscape a rugged character, worn by water and wind. There is no trace of humus: nothing to soften the vivid colours of the rock. Bodenbender, to whom we owe the first general attempt to classify the series, points out the importance of distinguishing the different strata in connection with the question of water supply and the conditions of human life.17 A complete geographical study would have to follow the geological description in detail. In places—on the eastern edge of the Sierra de los Llanos—the fine modern clays are in contact with the granites of the hills and form above them a thick bed that is rich in fresh water. In other places—south-westward of the Sierra de la Famatina, as far as the Bermejo—the outcrop is of red sandstone only. The tablelands of Talampaya and Ischigualasta, which are cut across by the gorges of the tributaries of the Bermejo, form one of the most conspicuously desert regions in the whole Republic. Wherever the gypsiferous marls of the Calchaqui are near the surface, the springs are saline. The undulations of the impermeable rocky substratum bring to light the water that gathers in the alluvial beds. Thus the streams which come down the Famatina range in the west disappear in the alluvial beds on the fringe of the Sierra, but re-appear presently in the oasis of Pagancillo.

Hence the valles are by no means wholly productive. The oases represent only a limited portion of them. It would be impossible to imagine a more striking contrast than that of the freshness and life of the oases compared with the surrounding desert. Screens of poplars shelter them from the zonda. The water runs along trenches paved with round pebbles under the spreading vines, at the foot of which, to economize water and space, lucerne is sown. Each garden feeds a family. Near the raw-brick houses there are large earthenware vessels, as tall as a man, in which the corn is kept. The hammering of the cooper fills the air.

In places the oasis is watered by a stream. In those cases there is on each side of the bed of the stream a narrow fringe, a continuous ribbon, of smiling gardens, which hide the path. Above and below Santa Maria a trench is opened every mile in the wet sands of the Rio. The water rises in it and fills it, and is directed by it toward one of the banks, where it is jealously collected and distributed. The water which flows from the irrigated fields and returns to the river, as well as that which the porous side of the trench has permitted to escape, goes to fill another trench and supply other fields farther on. The region of Los Sauces, in the northern part of the province of La Rioja, to the south of Tinogasta, shows a different type of irrigated cultivation, on account of the sandy course of the stream. The fields follow the feeding artery for about fifty miles. It is bled at the beginning of each bend, the waters remaining underground like hidden wealth.

In most cases however, the valle has no running water. What reaches it from the lateral quebradas is lost in the alluvial beds accumulated at the point where the quebrada enters the valle. In order to make use of it the cultivated areas are grouped on the cone of deposition; at least, that is the position in the great majority of the oases. A costa is a line of separate oases with their backs to the same slope. When the valle is narrow, the costas on either side of the sterile depression face each other, like two parallel roads. The water of the quebrada is never sufficiently abundant to irrigate the whole of the cone of the torrent. In order to create an oasis there, they have selected the most easily cultivable zone, which is usually the foot of the cone, where the deposits are finer and more fertile, retain the moisture better, and require less watering. The summit of the cone is composed of coarse stones, the first to be dropped by the torrent as it loses its strength. These are bad lands, where the water is wasted.

To meet the occasional drought and the danger of sudden floods in this fluvial zone, which is entirely the domain of the torrent, there is need of constant care and ingenuity. At Colalao del Valle the cultivated fields are five or six miles from the summit of the cone. After a number of successive years of drought the stream of water which reached them on the flanks of the cone lost half its volume and threatened to disappear altogether. They then built a stone dam at the outlet of the quebrada, and the water accumulates behind this during the night. At three o'clock in the morning the sluices are opened, and the stream, having thus nursed its strength, reaches the fields down below about seven o'clock. Then the sun and the wind rise, just at the time when the reservoir is empty, and by the middle of the day the stream ceases, and irrigation is suspended. At Andalgala, above which rises the glittering crest of Aconcagua, the waters of the melting snows which feed the torrent have not time to be "decanted" before they reach the valley. They come down laden with mud and sand. Above the points where the irrigation-channels begin the people make, in the bed of the torrent, a dam of branches of trees which filters the water. It is swept away by every flood that occurs, and is at once restored.

What is even more admirable than the ingenuity of the vallista in utilizing the natural resources is the minute detail of the water-rights. It seems as if the vallista is even more cunning in protecting himself from his neighbour than in dealing with nature. The water-customs of these Andean valleys are worth an extensive study. The water does not belong to the State, and is not used by concession from the State. It is private property. The owner uses or abuses it as he pleases on the lands which he has selected. A man may be poor in land and rich in water, which he accordingly sells. There are frequent business deals in regard to water-rights, just as in regard to the soil and its produce. Appropriation of water often precedes appropriation of the soil. Many oases are communities where the non-irrigated lands are common to the whole population, and the irrigated fields alone are divided.

A primary group of customs regulates the relations to each other of communities higher up and lower down the same stream. At Catamarca the water of a certain stream is shared by Piedra Blanca and Valle Viejo. Piedra Blanca, in the upper part, absorbs the whole of the water for a week, but it must then suspend its irrigation during the following week and permit the stream to flow down the valley. The same evening, or the next morning, according to the season, the water reaches Valle Viejo. It is a custom known as the quiebras in the southern valleys of the desert side of Peru, where it allows different stages of cultivation to proceed simultaneously. In the same way, above Santa Maria, where several communities (S. José, Loro Huasi, etc.) receive the water brought by a channel from the Rio Santa Maria, each of them has a right to the full output of the channel for three days. At the end of that time the sluices are closed, and the water passes to the next community. There is grave trouble for any oasis that has its rights infringed or does not compel the communities higher up to respect them.

Amongst individuals the water-right is generally defined by a measurement of time, a certain number of days or hours—during which the owner controls the entire flow of the spring or stream. It is only when the water is more abundant that we find another method of fixing the right of water, defining it by bulk. The water is then said to be demarcada, as the unit is customarily the marco, or the volume which passes through an opening about twenty-one centimetres in width and eight in height. The marco has infinite divisions, and each subdivision has its own name—the naranja, the bombilla, the paja, and so on.

As all the water is utilized, and the rights of all are equally entitled to respect, the division of the water into marcos (demarcacion) is in practice merely a proportional distribution of it amongst those who have rights to it. If the sum total of rights expressed in marcos represents something like the total flow of a stream during an average season, in the time of low water it is disproportionate, and the water no longer flows to the tops of the marcos. In other words, the quantity of water granted to each rises or falls with the rise or fall of the stream itself.

Theoretically, when the water-right is defined in marcos it is permanent. Often, however, it is impossible to grant each proprietor a permanent title to the water. Even in oases where the water is "demarked," the turno—that is to say, the turn of the proprietors to have water—which is the absolute rule in the poorest oases, reappears during the months of scarcity, in winter, when there is no rain, and at the beginning of summer. It reappears also when the right of ownership has been broken up into fractions that are too small, and it is better to grant a larger volume of water for several hours instead of a constant stream of water which would be too scanty for profitable use. At Andalgala the "turn" is sometimes obligatory, and regulated by custom, in channels where the irrigating proprietors are too numerous; at other times optional, and settled by convention amongst the owners themselves, when water is scanty. At Valle Viejo (Catamarca), when the water runs low, they set up the mita; that is to say, the sluices remain closed in each channel during four days out of eight, each proprietor in turn giving up his right to a permanent supply in order to have a double allowance when his turn comes. The turno is, therefore, a general practice. Everywhere we can see the farmers on the watch along the acequias, waiting for the moment to close their neighbour's trench with a pellet of clay and to let the stream into their own trenches with a blow of the spade.

The most minute precautions are taken in order that no one shall suffer injury. As the irrigation is always slower and less thorough during the night, they take it in turns to have the day and the night alternately. When the community receives the water from another community higher up the stream, the succession of "turns" amongst its members differs every time. The water comes down charged with sediment, pushing in front of it a mass of liquid mud, as the flush of a torrent does. It takes some time for the stream to become regular and clear. The first irrigator therefore exercises his right under unfavourable conditions. In the local phraseology the volcada de agua is not as good as the corte de agua, which means the irrigation that begins when the acequia is full.

Irrigation entails the services of quite a staff of arbitrators and administrators. The head men, who have jurisdiction of a higher order and secure the accurate distribution of the water amongst a number of channels or communities, are now, as a rule, officials of the administration, appointed by the provincial authorities (juez de Irrigacion at Catamarca, juez de rio at Rosario de Lerma). But the juez de agua of each community or each channel is a syndic elected by the interested parties. At Santa Maria the juez de agua is elected by the owners and confirmed by the Government. He controls irrigation throughout the department, settling all differences, submitting plans of work to a meeting of the owners, and assigning their respective charges in labour and contributions according to their rights.

This land of customs and traditions is also a land of lively movement. The briskness of the traffic is primarily due to continuous exchange between the various zones of the mountainous district. This large trade, so scattered that the railways could not dream of satisfying its needs, is carried, in the old fashion, on the backs of mules. The lively aspect of the roads between the tableland and the lower valleys of the region, the brisk interchange of goods between zones with different climates, is one of the common features of life on the Andes.

But the classic spectacle presents a different aspect in different latitudes. In Peru, and in southern Bolivia, the higher valleys—Jauja, Cuzco, the Pampas of Cochabamba and Sucre—have centres of dense population and agricultural wealth at a height of between 9,000 and 11,000 feet. They raise cereals, and receive from the tropical districts (montañas and yungas) sugar, cane-brandy, cocoa, and coca-leaf. The valleys of the Argentine Andes are usually at a less elevation than the yungas and montañas of Bolivia and Peru. But they are not hot districts, and have not tropical vegetation. Frost prevents the harvesting of sugar-cane at Salta, at a height of 4,000 feet. As to the coca-leaf, which is not as much used here as in the north, the Argentine valles do not send it to the tableland, but receive it indirectly from there, through the southern yungas. In default of tropical crops, the Argentine valles sow wheat and maize, which they sell to the Indians of the cold districts of the Puna for wool and salt.

These commercial currents are of very ancient, probably pre-Columbian origin. Boman has discovered ears of maize in the prehistoric tombs of the Puna de Atacama.18 The Puna, at a height of 11,000 to 12,000 feet, is permanently inhabited, unlike the high valleys of the Cordillera de San Juan, which are occupied only during the summer season by Chilean shepherds. It is primarily a pastoral and mining region, but it has some tilled land, at more than 6,700 feet above the level of the valleys. The higher limit of annual cultivation in the cold districts, which is fixed by the summer temperature, does not fall in the same way as that of arboriculture in warm districts, because trees suffer from the winter frosts. The Indians of Cochinoca and Susques sow lucerne and barley for fodder, and the quinoa and potato for food. Transport between the Puna and the valles is carried on by the inhabitants of the Puna, and is not shared by the vallistas. They are especially active in the north, in the province of Jujuy. Belmar shows how important the sales of the Puna woollen goods were by the middle of the nineteenth century.19 These fabrics were used by the mill-owners of the Rio Grande de Jujuy to pay for the work of the Indians of the Chaco, whom they employed in the sugar-cane harvest. The competition of the manufactured products of Europe now menaces the domestic weaving of the Puna, just as the competition of the flour of the Pampa menaces the cultivation of cereals in the valles.

Besides this traffic of local interest the valles serve for a traffic of a higher, almost a continental character. It seems certain that during the pre-Spanish period the road from the Peruvian tablelands to Chile avoided the inhospitable desert of the Puna de Atacama, entered the region of the valles to the east, and crossed the Cordillera in the latitude of Tinogasta, or even a little further south. That was the route of the armies of the Incas, which in the fourteenth century came as far as Maule. The pre-Columbian roads, of which Boman has found traces between the Valle de Lerma and the Valle Calchaqui, seem to correspond with this direction of traffic. By this route the long quechua passed amongst the Diaguites populations. The conquerors followed the Indian guides. Almagro, in going from Peru to Chile, passed through the valles at the eastern edge of the Andes.

Later the valles were incorporated in the many variations of the historic high road, one of the first and busiest of Spanish America, which goes from the Rio de la Plata to Lima: a route both for armies and merchants. The plan proposed by Matienzo (1566) to make a road from the silver mines to the estuary of the Paraná, through the Valle de Calchaqui, seems to have been intended merely to improve a line of communication that had already been in use. Buenos Aires for a long time received European goods by this road. About 1880 the Salta route recovered for a time its continental importance, during the Pacific War and the occupation by the Chileans of the maritime provinces of Bolivia.20 At that time it was the only outlet for Bolivia.