Spanish domination in the last quarter of the seventeenth century extended from the Tropic of Cancer in New Galicia to 40°S in Chile. As with other intruding empires, its rule was most easily and firmly based in areas such as Mexico and Peru where existing well-organised political structures could be taken over; and it frayed out into the merest ‘presence’ on the more arid or jungly margins. Both the northern and the southern frontiers were far from being well settled in any sense. On the arid marches of New Galicia, the Chichimecas around the silver town of Zacatecas were not fully pacified, after several decades of ferocious guerrilla war, until 1600, and then ‘not by the sword [our ‘military solution’] but by a combination of diplomacy, purchase, and religious conversion’, including the settlement on the frontier of Tlaxcalans, stout peasant types with a ‘special relation’ who could be exemplars to the Chichimecas of a more civil existence.^[4] It was just at this time that in the far south the Araucanians compelled a local but long-enduring retreat. Spanish control was exercised from no more than three- 178 -large cities—Mexico, Lima, Potosi—with a handful of second-rank towns, and thence through scores of small district centres, tiny ports, missions and mining camps, many of them wretchedly rough places. There were many gaps where climate and terrain were more effective defences than armed force: so near the heart of the Indies as the Isthmus of Panama, Spanish power was so feeble, or so ill co-ordinated, that in 1698 the Scots could insinuate their Darien colony, unmolested by the Spaniards for over a year—though by the same tokens of climate and terrain, it was quite debarred from any growth or useful activity of its own.

Within these limits and limitations a small (though constantly increasing) Spanish minority was able not only to control a large (but constantly decreasing) Indian majority, but also to maximise the latter's productive capacity by a most ruthless exploitation. It is true that exploitation had its indigenous precedents; as Lynch says, ‘Whereas before the Peruvian Indians had toiled to build temples of the Sun, now they laboured to satisfy the bullion demands of the world economy’,^[5] and both the old masters and the new were powerfully aided by the bonds of religion. But the old system had more reciprocity—at least in the understanding of the people—and the Spanish exploitation was often so intense and uninhibited (except by laws promulgated far away in Spain, and disregarded) as to become counter-productive. Its workings were at first sustained but in the long run undermined by Indian fatalism in face of the break-up of their world, a fatalism at once a cause and an effect of appalling population decline.

Despite the apologetics of writers like Salvador de Madariaga, using selective data and faulty method, this demographic disaster, probably the most catastrophic in all human history, cannot be gainsaid.^[6] The pre-conquest population cannot of course be known with certainty. The most serious recent proponent of a low figure is Angel Rosenblat, who in 1935 estimated that of Mexico and Central America as 5,300,000 in 1492; but the very careful work of Simpson, Borah, and Cook, using Aztec tribute lists as well as Spanish data, indicates a 1523 population of 16,800,000 in central Mexico, roughly the Aztec domain; since initial depopulation is known to have been very rapid, a 1519 population of 20–25,000,000 is possible. By mid-century it was about 6,000,000 and in 1605 only 1,075,000; two or three decades later a slow recovery was in train. The earlier figures have been challenged by Rosenblat on grounds of ‘manipulation’; but, as Borah says, he simply repeats himself in 1967 ‘without change in the estimated figures, although with enormous additions at the foot of the page’. Such exact adherence over thirty years—and years of remarkable progress in the techniques of historical demography—is the more strange in that Rosenblat's figures were originally put forward as tentative. There seems a certain rigidity of mind, and Rosenblat's arguments are as a whole unconvincing.^[7]

It seems likely that there was already pressure on resources in the fifteenth century, and the Spanish shock may have simply tipped over an already precarious Malthusian balance; but discount the earlier figures even to Rosenblat's, and the story is still terrible. First place among the responsible factors must undoubtedly- 179 -go to epidemic disease—‘bacteria and viruses recognised the unity of the planet long before man’, and areas that were so to speak kept out of this unity by oceanic barriers—America, Australia, Oceania—owed to this spatial quarantine a total lack of immunity to new infections. The rest of the world had knocked together for millennia; now for the Americas the barriers to epidemic invasion were destroyed in decades, with results of course the most devastating precisely where population was densest.^[8] But there was also a great toll of life in the wars, by actual slaughter, economic disruption, and famine; Spanish labour demands did the rest, together with that most universal solace of the dispossessed and oppressed, alcohol; and probably a weariness of living in the confused regimen of a disordered world, where ‘To castrate the Sun, for that the strangers came’.^[9]

For Andean America, Venezuela through Chile, Rosenblat gave a 1492 total of 5,100,000, of whom only 2,000,000 were in the present Peru, and this last figure, falling to 1,500,000 in 1570, is often quoted.^[10] A priori, it seems too small, especially in comparison with Mexico; Chaunu speaks of the richness of Peru lying in its mass of 4–5,000,000 Indians. A recent review by Nathan Wachtel, using Spanish local enquiries as well as overall estimates, suggests tentatively but quite reasonably a pre-Pizarro population for the Inca Empire of 7–8,000,000, possibly 10,000,000; whatever the starting point, there was a steep decline to 2,500,000 in 1560, then a slower fall to 1,500,000 in 1590. The causes were as in New Spain: ‘Abus, guerres, epidemies' and, except for a few favoured Inca grandees, a trauma perhaps more acute than in less centralised and absolutist Mexico, for with the violent death of the Inca the arch of the world's fabric had collapsed.^[11]

One may reckon, then, on an initial mass of ten million souls at the very least, more probably forty million, dwindling with fearful rapidity and hence supplemented, especially in hot coastal lowlands, by slaves imported from Africa, though for a century these were numerically but a small element barely on a par with the all-controlling whites. In 1570 Europeans may have numbered 63,000 in New Spain (18,000 in Mexico City), 25,000 in Lower Peru and 7000 in Upper (roughly Bolivia). By 1630 these figures had risen respectively to 125,000 (48,000 in the capital, say half the population of Madrid or Vienna), 20,000 and 50,000, this last great increase reflecting the rise of Potosi, whose wealth became proverbial in other languages than Spanish, and whose total numbers, including Indians, rivalled all but a handful of European cities.^[12]

The raison d’être of Spanish settlement (the glory of God apart) was essentially the extraction of treasure from the earth, and both then and now the world's gaze has been caught by the dazzling mining economy of Spanish America. But the miners had to be fed and clothed, and to this end, often in disregard of metropolitan interests, there was a remarkable development of agriculture and stockraising, with some consumption manufactures, and these were oriented to the market—even if in some places a very local market with the mine, corn-supplying arable farm, and ranch for meat, leather and tallow, in one- 180 -close-knit complex. Farms and plantations and ranches of course depended on the acquisition of land, and of the labour to work it.

Whether by right of conquest or of Papal donation, the Crown of Castile considered itself the owner of the land of the occupied Indies, and it was in its own interests quite as much as those of the Indians that it strove to prevent the encomienda becoming effectively an hereditary landed fief. Direct land grants were initially carefully restricted, but laxity soon set in, and there were many opportunities for expansion: taking over land left vacant by the dying Indians, forced purchase or legal chicanery at Indian (or ‘poor White’) expense, simple squatting in a country where some delineations of title long remained ‘by eye’—that is, as far as could be seen from a given point.^[13] In New Spain at least the difficulty was not so much in acquiring an estate as in manning it, and as the encomienda declined the repartimiento and its Peruvian equivalent the mita (for which there was Inca precedent) grew in importance. This was a forced levy of a proportion of the working population of each Indian community, drafted for several short periods in each year to work at low wages either on public corvée or for private applicants.^[14]

The system was obviously cumbrous, with great wastage—especially in travel time—and disruption of Indian subsistence farming, and liable to vicious abuse. Early in the seventeenth century the norms were changed in New Spain: the labourer could now choose his employer, and with falling population there may have been some reality in the choice; but offsetting this, the proportion in the village so compelled to choose was raised from one-seventh to one-quarter. Finally, in 1632, the Mexican repartimiento was abolished, except for mines and public works; not so much a concession to the Indians as an effort to retain labour for these essentials. By this time the encomiendas were being replaced by great landed estates or haciendas, and the repartimiento was simply unable to meet their labour needs, which were supplied by the wage-labour, at least nominally ‘free labour’, of the old occupants. Much of this new-old labour force was naturally soon gripped by debt peonage into an almost serf-like status. In the more northerly mining areas of New Galicia the locals were far too well armed and mobile, and far too little tamed, to be shared out by repartimiento, and here the mines attracted labour from older centres, in marked contrast to the ever-attempted flight from mining in Peru. On this isolated frontier it was easy to get the workers so attracted into debt-bondage and to hold them there; nevertheless the new conditions did mean some betterment for many Indians, and it may well be significant that it is around the 1630s that the Indian population begins, however slowly, to increase in numbers.^[15]

But the solution reached by so much trial and error was essentially the creation of latifundia: Iberian America was basically a ‘big man's frontier’, with little place for the simple squire or the intensive small farmer, except around a few city markets such as Lima or Panama or in such favoured areas as Antioquia. For well over three centuries, until the Mexican revolution of the- 181 -1920s, the latifundist solution gave social, though not political, stability of a sort; it articulated ‘property into a system which, though it was neither feudal nor seignorial, established a rigid social hierarchy’.^[16] But after the Mexican ejido and Cuba, the price of this stability is still being paid today, in blood; even in Chile, so long the model of orden, progreso, libertad, those three words which sum up so much of the slogans of Latin America, so little of the realities.