Colonial Silver Mining: Mexico and Peru

Colonial silver mining began when European techniques of production were introduced into the New World to satisfy the sustained European demand for precious metals. At first the industry formed little more than an overseas extension of the great central European mining boom of the years 1451-1540. Certainly it was then that an association of high capital investment with a remarkable range of technical innovation succeeded in pushing German silver production to unparalleled heights. Problems of drainage were solved by cutting adits several miles long beneath the lode, or by the installation of great water wheels and whims moved by teams of up to a hundred horses. As early as 1451 the invention of lead smelting facilitated the separation of silver from the copper compounds with which it was usually found. The construction of stamp mills, impelled by either water power or horses, completed the circle of improvements.¹ The publication of G. Agricola’s De re metallica (1556), a lavishly illustrated work, hastened the diffusion of practical knowledge.²

Despite these obvious technical debts, colonial silver mining soon acquired a structure of production radically distinct from its Old World predecessor. The difference in scale was striking. Although when at its peak in the decade 1526-35 the European industry cut 350,000 marks (of 8½ ounces) a year, by the close of the century its output had fallen to 100,000 marks, a mere tenth of the American bullion imports then registered at Seville. Equally important, as early as the 1550s, miners in Mexico developed the amalgamation process, a cheap, simple method of refining large quantities of low-grade silver ore. In the years immediately following the discovery of Potosí in Upper Peru, Indians smelted its rich ores in their own clay furnaces. Then again, the labor demands of the industry drove the viceregal authorities to levy work drafts from the Indian peasantry.

These differences suggest, therefore, that American silver mining offers fewer comparisons with its German antecedents than with that other great colonial industry, the manufacture of sugar. The similarities are striking. Unlike tobacco or gold, both sugar and silver required an elaborate refining process only rendered possible by considerable capital investment. In both industries labor costs dominated the primary sector of extraction or cultivation. Both devised labor systems peculiar to America, the one relying upon imported African slaves, the other— at least in the first instance—upon enforced recruitment of Indian peasants.³ Finally, both operated within strict geographical limits, sugar growing best in the tropical lowlands, and silver being most abundantly located in the high ranges of the inland cordilleras.

The general effects of American silver production have long been the object of academic debate. Its role in the price revolution of the sixteenth century is still in dispute.⁴ Similarly, its contribution to Europe’s trade with Asia, despite contemporary concern, has yet to be justly weighed. But for Spanish America no one can doubt its importance. The mining camps and towns generated sufficient purchasing power to stimulate not merely trans–Atlantic and Pacific commerce but also long-distance internal trade. Potosí received cloth from Quito, mules from Buenos Aires, sugar and coca from Cuzco, and brandy from Arequipa.⁵ Merchants as well as miners built then fortunes upon the industry. It was the existence of this export sector in the colonial economies of Peru and Mexico which prevented them from becoming simple agrarian or feudal societies.

The aim of this article is to review the state of existing knowledge about colonial silver mining. We offer, however, a guide rather than a catalogue: the focus is more upon problems than bibliography. In areas where information is available but uninterpreted, we advance hypotheses.

I. Geologic Aspects

If the agrarian historian must take into account the effects of climate and soil formation, so too the student of mining should examine the geologic record. The location and composition of ore deposits in good measure determined the organization of the industry.

In general, nearly all American silver has been found in the great chain of mountain ranges which stretch from Canada to Chile. Geologically, the cordilleras of the western Sierra Madre of Mexico and the Peruvian Andes derive from the volcanic eruptions of the Tertiary Age. It was then, apparently, that magma intruded into the multiple faults and cavities in the earth’s crust. Upon cooling, the argentiferous magma crystallized for the most part into fissure veins of varying silver concentration. These formations were as various and particular in their distribution, depth and dip as the mountains which encased them. The bulk of these primary ores took the form of low grade sulphide compounds, but near the surface the prolonged effects of weathering tended either to isolate the ore into particles of native silver or to oxidize it, associating in the latter case with reddish soft minerals known to colonial miners as colorados or pacos. Furthermore, the incessant percolation of rain water carried a considerable portion of these oxidized ores below the water table level, there to be re-precipitated in a process now described as secondary or supergene sulphide enrichment. An alternative mode of enrichment at this level, defined by geologists as hypogene concentration, derived from interaction of the primary ores with further interior flows. The hard, dark ores so formed, called negrillos, were frequently as rich if not richer than the colorados above them.⁶ Finally, beneath this middle enriched zone lay the great mass of low grade primary sulphides, which slowly diminished, merging into a variety of base metal compounds.

This general pattern, needless to say, was interrupted by local idiosyncrasy. At Guanajuato, little oxidation occurred; nevertheless, its mainly sulphide ores had experienced a hypogene concentration process almost to surface levels. At Potosí the conical shape of its isolated peak facilitated the formation of an oxidized encasement 250 yards deep, beneath which lay a zone of high grade sulphides. Moreover, the multiple veins of its several lodes all contained a considerable scattering of native silver or argentite.⁷

Used with caution and respect for local variation, the geologic record can provide the historian with invaluable evidence as to the physical framework governing the miners’ exertions. It helps to explain the familiar cycle of bonanza, abandonment, renewed bonanza and final depletion which characterized the history of so many American mines. For although quantities of native silver could be found near the surface, in all likelihood the most consistently rich zone of the lode lay at water table level–in Mexico at 400-600 feet. Few mines during the colonial period, the Valenciana and Potosí apart, pushed beneath this middle region; indeed most did not reach it until the eighteenth century.⁸

The economic deductions that can be obtained from a comparison of the geologic formation with recorded shaft depths are extensive.⁹ Once the days of primitive surface bonanza were past, then, within the observed scale, the farther down the miner pushed the higher his likely yield. With the water table attained, both his capital and maintenance costs inevitably rose, but so too did his gross production. Naturally this presumption could always be disappointed by local variations or difficulties. Moreover, such a policy required considerable capital investment in the form of adits or shafts, frequently on a scale beyond anything available to the colonial miner. The cycle of abandonment and bonanza referred to above sprang from two combined causes: the nature of the ore deposits, and inadequate capital resources to counter their effects.

II. Mining Techniques

The establishment of mining as a rational industry rather than as a smash and grab raid upon the earth’s crust primarily depended upon the construction of deep shafts and long adits. Unfortunately we know little about the development of underground excavation. True, at the end of the eighteenth century European visitors poured scorn upon colonial technology.¹⁰ They exclaimed over the absence of cross galleries, the tendency to cut work tunnels which corkscrewed through the lode, the lack of mules for carriage, and the reliance upon whims. Humboldt noted with surprise that Mexican miners used crowbars similar to those employed in Germany in the sixteenth century.¹¹ In consequence of these criticisms the impression lingers that American technology remained both backward and immobile, isolated from later European developments. In particular, the failure to adopt the use of the steam engine has confirmed the view. Yet, in point of fact, colonial methods of excavation underwent several marked phases in the course of two and a half centuries; important differences, moreover, existed between Peru and Mexico. And for the most part these developments corresponded to local attainments and cost possibilities. If steam engines were not introduced, it was because their fuel costs were exorbitant.¹²

To understand technical development in New Spain three main problems have to be considered: the dimensions of shafts and adits; the use of underground cartridge blasting; and the type, capacity and number of whims. Dimensions are significant because they provide not merely a clue as to the size of a mine but serve as an index of capital value. To construct a deep shaft cost as much as to build a factory or a church. Moreover, to know the average depth is to gain a guide as to geological possibilities and problems. The same considerations apply even more to the cutting of adits, since these horizontal tunnels rarely yielded any short-term returns to cover working costs.

In 1617 at San Luis Potosí the local magistrate helped to organize the construction of a drainage adit some 250 yards long. In Mexico this is the only reference to this drainage method available for the seventeenth century. At Zacatecas a shaft 112 feet deep was cut during the 1640s. At Parral the deepest level was 130 yards. Nevertheless, by 1697 Gemelli Carreri inspected shafts at Real del Monte and Pachuca that had reached 200, 184, and 160 yards.¹³

To penetrate to these levels, the Mexican miner needed two weapons: gunpowder to break the hard rock and whims to keep the shafts drained. As yet, we do not know when cartridge blasting was introduced into New Spain. Guanajuato tradition avers that it was José de Sardaneta, owner of the Rayas, who first used it during the 1720s. Certainly Zacatecas in 1732 consumed remarkably little of this commodity—a mere 1,300 pounds.¹⁴ The great excavations of the later eighteenth century all depended upon the intensive employment of gunpowder.

Equally obscure is the evolution of the whim from a simple hoist, worked at times by hand, to a powerful machine serviced by teams of six to eight horses or mules and capable of hauling up loads weighing 1,250 pounds. Bakewell refers to their operation at Zacatecas in 1679. Gemelli Carreri found some sixteen at work at Pachuca and Real del Monte. More important, he recorded the installation of two whims in one shaft. By the close of the eighteenth century the Valenciana employed no less than eight whims in one great central shaft, cut in octagonal form to facilitate their use. Elsewhere, the installation of four or five machines became common. According to an informant from Real del Monte, the haulage capacity of whims in that camp quadrupled during the course of the century.¹⁵

During the Bourbon epoch, the Count of Regla drove an adit 2,881 yards long beneath the Veta Vizcaína, a project which took over thirty years to complete. The Valenciana’s great shaft, with circumference of 32 yards and a depth of over 600 yards, cost a million pesos to construct—the equivalent of two churrigueresque churches.¹⁶ The new technology thus depended upon heavy capital investment.

In Peru, by contrast, deep-shaft mining remained relatively underdeveloped and in consequence whims do not appear in common use. The reasons for this difference were simple. At Potosí and possibly at other camps, the conical peak form of the area of ore deposits made adits the more practicable form of access and extraction. Though work began near the top of the cerro rico where the lodes broke the surface, as early as the 1560s depths of 200 and 250 yards were reached. At this point the first adit was cut by the Florentine miner Nicolás de Benino, designed to intersect the lode at a depth of 320 yards.¹⁷ By 1585 Capoche recorded one adit, presumably Benino’s, 250 yards long, and referred to another five under construction. These tunnels were six feet high and eight feet wide, measurements which allowed only two Indian pickmen to work on them at any one time.¹⁸ From these adits radiated a series of work shafts of the peak’s many mines. All ore was brought out on the backs of porters clambering up rough wooden ladders. The adits at this stage were used, of course, for entry and extraction rather than for drainage.

At Potosí the system of exploitation is difficult to understand. Individual mines only measured a few yards along a particular lode, and owners of adits permitted workers from other mines to use them. Presumably the farther the shafts or work tunnels were driven into the peak’s core, the greater was the need for new, longer adits to intersect the lode at lower levels. In 1640, Viceroy Chinchón related the construction of two new adits, one with a length of 400 yards. No doubt similar projects were initiated both earlier and later. During the eighteenth century the Crown sponsored the construction of a major adit with the purpose of draining the lower reaches of the peak. Work began in 1789, and by 1810 a tunnel six feet square had been driven 2,200 yards at a cost of 560,000 pesos.¹⁹ In its last epoch Potosí was thus still able to rival the Mexican mines in the scale of its undertakings.

Elsewhere in Peru information is lacking. At Huancavelica, the great mercury mine, it took nearly forty years, between 1605 and 1642, to construct an adit 434 yards long. Work here, however, was dangerous because of mercury fumes, and at times was suspended. Once completed, the adit permitted a safer and more rapid extraction of ore. As early as 1631 gunpowder was used for underground blasting at Huancavelica. Apparently damage to the mine was so extensive as to soon stop the practice, but it was reintroduced during the 1730s.²⁰

Other Peruvian camps remained technologically backward. At Cerro de Pasco at least two adits were cut during the eighteenth century. But the deepest shaft did not extend beyond 100 yards and was serviced by simple hand pumps and ladders rather than by whims.²¹ Whether this state of affairs sprang from the nature of the terrain, lack of capital, or technical ignorance is still unresolved.

III. Refining

In contrast to the paucity of information about excavation practices, an extensive literature surrounds the technique of amalgamation, and, to a lesser extent, smelting.²² As early as the 1530s a group of German miners introduced into New Spain both the stamp mill and lead smelting. Within twenty years, however, high fuel costs, combined with the falling quality of ores, had rendered this method uneconomic. The situation was transformed in 1554 when at Pachuca Bartolomé de Medina, a merchant from Seville, developed the amalgamation process by use of mercury and salt. Some controversy has raged about his right to claim the invention, but without doubt it was the Spaniard who carried out the experiments which made it an industrial reality. Despite the assertion that Medina added a catalyst called magistral (copper or iron pyrites), it seems likely that the surface oxidized ores at Pachuca already contained sufficient magistral,²³ Gómez de Cervantes did not mention this ingredient, and Mota y Escobar, writing about Zacatecas c. 1604, explicitly commented upon the recent introduction of the practice, made necessary by the increasing predominance of sulphides (negrillos).²⁴

Bargalló emphasizes that the first printed reference to amalgamation in an open patio was made by Gamboa in 1761. Both Gómez de Cervantes and Gemelli Carreri described a process carried out in large vats or containers.²⁵ The moment of transition remains obscure. Both West and Bakewell presume rather than prove that the patio system was the normal method throughout the colonial period.

Other changes are equally difficult to document. A Guanajuato tradition dates the invention of the arrastre, a stone-drag crushing device, to the early eighteenth century. Used in conjunction with the stamp mill, the arrastre, rather like amalgamation itself, was extra-ordinarily simple, cheap and efficient. Its reliance upon mule power reflected the lack of water in northern Mexico and the increased reliance upon the animal in New Spain.²⁶

By the close of the eighteenth century Mexican refining was notable for the size of its establishments and the high level of technical expertise. The Count of Regla’s mill at Real del Monte housed 24 arrastres and cost a half a million pesos to construct. At Sombrerete the Fagoaga mill had no less than 84 arrastres and 14 furnaces.²⁷ Similarly impressive was the diffusion of knowledge. Apart from Gamboa’s invaluable work, four other books were written about refining at this time, two of them by practicing miners.²⁸

At Potosí, the story, especially in its first stage, differs remarkably from that of New Spain. There Indians dominated the refining sector of the industry from discovery of the site in 1545 until 1572. They ground the ore in their simple stone mills called marays, and then smelted it in small clay furnaces fired by wind-fanned dried ichu grass or llama dung. At the height of this system over 6,000 of these guairas lined the hills surrounding the town. The only European element in this technology consisted in the use of lead or litharge to facilitate the separation of the silver.

By the late 1560s, however, production at Potosí fell dramatically, following the increasing depletion of high grade ores suitable for smelting. Two attempts to introduce amalgamation failed. But finally in 1572 Pedro Fernández de Velasco, backed by the Viceroy Toledo, conducted successful experiments. Immediately the structure of the industry was transformed. Spaniards took over the refining sector, constructing stamp mills, furnaces and vats.²⁹ Although some mills were powered by horses, the majority of these machines were driven by water wheels. To obtain a regulated flow of water, the town financed the construction of a series of dams, eventually creating about twenty artificial lakes. The total cost possibly amounted to over two million pesos.³⁰

The process introduced by Fernández de Velasco did not vary greatly from that described by Gómez de Cervantes in Mexico. The amalgamation took place in a stone vat or container, and was hastened by exposure to a low heat. But if production soared, so did costs. It was only towards the end of the next decade that refiners realized that the dark ores, the negrillos or sulphides, demanded the addition of a magistral, copper or iron pyrites, if an excessive consumption of mercury was to be avoided. By the end of the century they had worked out which ingredients should be added to the different types of ore. By then they found it cheaper and more efficient to conduct the process without fire. Henceforward no significant changes occurred. In Peru amalgamation was always carried out in stone or tiled containers, never in an open patio. These cajones held 50 quintales of ore, compared to the 15-32 quintales of the Mexican montón. The separation of silver went considerably faster than in Mexico, lasting from ten to fourteen days, whereas in New Spain it often took up to six weeks.³¹

By the early seventeenth century most American silver was produced by amalgamation; only high grade ores and lead compounds were smelted. The deeper or older the mine, the greater would be the dependence on mercury. The only rival to these two methods was the caso process developed by Padre Barba, in which the ore was smelted with mercury in a copper cauldron.³² The great drawback here was the tendency of impurities in the copper to enter the reaction. In 1786 the Hungarian Baron de Born invented a variant of this process, using copper-lined barrels. The Spanish Crown commissioned German experts to teach this method to colonial miners. In the upshot, however, the equipment proved to be expensive and complicated, and the practical results unimpressive; apparently it did not separate as much silver as amalgamation.³³ We may note that as late as 1870 some 71 per cent of all Mexican silver continued to be produced by this sixteenth-century technique.³⁴

Various problems confront the historian who attempts to come to a closer view of colonial refining. Perhaps the most important is an epistemological one. On many occasions both miners and royal officials claim that ore levels had fallen; they then provide an average figure, let us say, of one ounce of silver per hundredweight of mineral. The historian then has to decide: did these refiners know how much silver their ores really contained? To answer this question we must consider the basis of their knowledge. In the first place, colonial miners classed all ores according to their external characteristics, such as color, hardness and form. But equally important, these characteristics were related to the changes experienced by the different ores when smelted or amalgamated. In each camp a local tradition or expertise soon developed, and was fortified by the continuous application of generations of refiners. By the eighteenth century a miner at Potosí or Guanajuato could probably guess the quality of ore from its external appearance. The proof, of course, was in the cooking. Here too local tradition and experience determined the right balance of ingredients. No one understood the true nature of the chemical reactions, but all could judge its state by practical tests. In the nineteenth century Laur stressed that all attempts to improve the process from chemical theory invariably failed; the advice of the local azagueros was always to be preferred.

To assert the skill of the cooking is not to deny its failures. Laur took tests in various Mexican camps. He found that at Zacatecas about a quarter of the silver was left behind in the residue; at Atotonilco el Chico the proportion rose to over a third. By contrast, in Guanajuato, where the ores were noted for their tractability, the loss only amounted to 12.65 per cent.³⁵ He observed that the master workmen in charge of the process rarely assayed the residue, fearing their skill might be criticized. Earlier at Potosí, it was estimated that a good third of the silver was lost. Whatever the case, the historian should always guard against statements about ore general levels. By 1790 Cañete declared that at Potosí the average yield was ¾ ounce of silver per hundred-weight of mineral, a level considerably lower than the 2-ounce average claimed for Mexico.³⁶

A similar doubt surrounds the consumption of mercury. The usual ratio was two to one, i.e., a hundredweight of mercury produced one hundred marks (50 pounds) of silver. On the other hand, most treatises on amalgamation suggested that it was possible to refine a mark with a loss of only twelve ounces of mercury. In practice the situation varied from camp to camp according to the tractability of local ores. In New Spain the local treasuries demanded different ratios, ranging from 85 marks in Bolaños to 125 marks in Guanajuato per hundredweight of mercury distributed. For seventeenth-century Zacatecas, there were considerable variations, with 112-126 being the most frequent. For Potosí, Cañete estimated the ratio at 120-130 marks.³⁷ The range, therefore, was not extensive, a conclusion which has significant consequences for estimates of silver production.

Finally, we may note that nearly all the literature concentrates upon amalgamation to the virtual exclusion of smelting. Yet the mid-seventeenth century shortage of mercury drove the Zacatecas miners back to smelting to such a degree that in the years 1685-1705 just under half of all the camp’s silver was produced by that method. At nearby Sombrerete, newly discovered, all ore was smelted.³⁸

IV. Labor

No other internal aspect of colonial mining has attracted so much attention as its labor system. Since the denunciations of Las Casas, the royal levy of tributary service, called repartimiento in Mexico and mita in Peru, has been associated with the great question of the decline in Indian population. Here too the difference between the two viceroyalties is thrown into high relief. In Mexico forced labor ceased to be an important element among mine workers as early as the middle seventeenth century. In Peru by contrast the mita continued until its legal abolition in 1812. The reasons for this divergence are not entirely clear, but granted the uniformity of Spanish demand, it is tempting to locate such causes in the disparate development of the two Indian peoples and the different location of the principal sites respective to sedentary populations. In Mexico the formation of a class of professional miners has been taken for granted; in Peru the iniquities of the mita are equally assumed without question.³⁹

To inspect New Spain first, it is obvious that the twin themes which dominate its early history, the catastrophic decline in Indian population of the Center and the colonization of the North, governed the development of the Mexican mining industry. At Pachuca a weekly repartimiento of 1,108 Indians drafted from the surrounding villages in the years 1576-79 dwindled to a mere 57 men by 1661. The same process, no doubt, was at work in all the camps situated within the confines of the old Aztec Empire. Even as early as 1598 volunteer wage-earners outnumbered drafted workers by two to one. Moreover, the Chichimeca North mines always relied upon free labor, supplemented by an assortment of slaves, some Indian, but mainly African, suitable for work in the refining mills. At Zacatecas most mine-workers were Indian migrants from Michoacán and the central valleys. The town was soon surrounded by small villages or suburbs, inhabited by Tarascans, Mexicans, Tlaxcalans and Texcocans. Numbers here are important. Bakewell estimates that seventeenth-century Zacatecas, which produced about a third of all Mexican silver, never required more than about 5,000 workers. He therefore argues that labor supply was never a problem for the Mexican mining industry, since at most it needed 15,000 men.⁴⁰

By the eighteenth century, mine-workers constituted a labor aristocracy, characterized by lavish spending and geographical mobility. By this time the northern workers were mainly mestizo, mulatto, and acculturated Indians. Their earnings derived not so much from the basic daily wage of four reales, as from their partidos, a share of the ore which all pickmen received. Towards the end of the century, however, a change occurred, when in Guanajuato this practice was replaced by the payment of high daily wages of eight to ten reales. This unpopular decision probably did not extend to other camps; in part it sprang from the greater social power exercised by mine owners following the creation of the provincial militias. Whatever the system, labor still remained the chief cost in the mining sector of the industry, amounting to as much as 75 per cent of total expense.⁴¹

To pass from Mexico to Peru is to enter a vale of tears. The Peruvian mita, as first set up in 1574, was surely one of the most remarkable economic institutions devised by the wit of man.⁴² Each year the Indians of the region stretching from Potosí to Cuzco had to send roughly a seventh of all adult males to work in the mines and refining mills of Potosí, a number estimated at about 13,500 men. It took the Cuzco contingents some two months to cover the 600-mile journey across the Andean altiplano. An identical arrangement supplied Huancavelica with 3,280 Indians recruited from the surrounding districts. Some lesser minefields were also consigned small levies.

It should be observed that the mitayos were only obliged to work one week in three, their chiefs dividing the men into three shifts upon arrival. Similarly, of these 4,500 men, only a third entered the mines, the rest were being sent to the refining mills and other tasks. By all accounts, most Indians hired themselves out as free laborers, mingados, during their two free weeks. Not merely did they then receive different wages according to their status, but also, as free pickmen, they were entitled to a small share of the ore they cut.⁴³ The mita therefore supplied Potosí with an enforced labor contingent of about 4,000 men, together with the possibility of employing double that number as “free” workers.

The mita was then and later the object of fierce attack. At least three separate questions are involved. Did work at Potosí and Huancavelica actually kill Indians? Did the mita contribute to depopulation? Did it result in flight and social disruption?

A comparison between Mexico and Peru may throw light on this thorny subject. In New Spain a rather small number of Indians left their villages and lands in Michoacán and the central valleys to settle in the North to become professional mine-workers. In the Andes, so we are told, Indians similarly abandoned their lands and villages, but in this case it was to avoid working in the mines of Potosí. The difference in behavior is striking and inexplicable. Why should Potosí appear more hateful than Pachuca or Zacatecas? Was it so dangerous as to drive a man to desert his ayllu and abandon his lands? How does the class of permanent workers at Potosí fit into the picture? These are questions colonial historians must one day examine.

Whatever the case, the contrast between the two viceroyalties does illumine the economic significance of the mita. In fact, many Andean Indians did settle at Potosí and Huancavelica to work voluntarily in these mines. Even before Toledo organized the mita, the chiefs of Chucuito, a province in Puno, sent 500 men to Potosí to earn sufficient money to pay for all the district’s tributes.⁴⁴ Thus Peruvian mining could have obtained labor without the mita in much the same fashion as in Mexico, through the slow development of a highly paid permanent body of workers. What the mita represented was a massive input of cheap labor. The difference lay in the scale and the price. It was the mita which permitted production at Potosí alone to soar far beyond the contemporary total for all New Spain.

Further light is thrown on this question by comparison with the eighteenth century. By then all Peruvian camps, with the exceptions of Potosí and, to a much lesser extent, Huancavelica, had followed the Mexican pattern. They all relied on free labor. By then the mita, much reduced to no more than 3,280 men, acted chiefly as a subsidy to maintain Potosí in existence. Without this source of cheap labor the cerro rico, with all its rich ores long since exhausted, could not have continued to work minerals which on average yielded no more than ¾ ounce of silver per hundredweight. Whereas it took most Mexican mines over two hundred years to push through the enriched middle zones of their lodes, at Potosí the mita both created the first rapid boom and then subsidized continued production. The same was true at Huancavelica, where the Indians of the nearby districts, rather than working, paid a mita tax which enabled the miners to meet their labor bills.⁴⁵

Concentration upon the negative aspects of the mita has obscured the fact that by the eighteenth century most Peruvian silver was produced by workers who had freely chosen to become miners. We still lack any characterization of their provenance, earnings or social type. Yet surely Potosí’s impact upon Andean culture must have been all-pervasive. Indians from a vast region were thrown together; they were put into immediate contact with Spaniards, mestizos and Africans; they received the ministrations of a numerous clergy; they could entertain themselves in the equally numerous taverns of the town. Hitherto, emphasis upon the undoubted human suffering involved has inhibited any study of the social and cultural effects of this great annual migration.

V. The Crown

No consideration of colonial mining can afford to neglect the pervasive role of the Spanish Crown. In addition to the issuance of a legal code to govern the physical exploitation of mines, the Crown levied a heavy tax on all silver produced, monopolized the production and distribution of mercury, and greatly profited from its control of the mints. The justification for such extensive intrusion lay in the traditional legal doctrine that all natural deposits of precious metals belonged to the royal patrimony. The code drawn up by Toledo applied in all Spanish colonies until 1783 when Gálvez issued new ordinances for Mexico.⁴⁶ The commentators all emphasized that the Crown only conceded working rights to the colonial miner; failure to exploit a mine entailed the lapse of all proprietary claim. In the case of the great mercury deposits at Almadén and Huancavelica the principle of direct royal ownership was preserved, the mines being worked by leaseholders or under contract.

The customary tax charged by the Crown amounted to a fifth of production. Once his silver was cast in bar, the American miner was legally obliged to present it at the nearest treasury for stamping and payment of the quinto. Another one and a half per cent was also deducted for assay costs.⁴⁷ In New Spain, as early as 1548, miners obtained a reduction in this tax to a tenth. This concession, at first extended only to miners, as distinct from refiners, workers or silver merchants, later became general throughout the industry. In Zacatecas, the proportion of silver taxed at the old 20 per cent rate fell slowly in an uneven curve between 1559 and 1626, from about a quarter to a seventh of the total. After the late 1620s hardly any further quintos were collected. Evidence for other camps in New Spain is lacking until the 1660s, when the Viceroy Mancera reported that the tenth was indeed the usual rate, with only insignificant quantities of silver paying the fifth. This situation was finally legalized in 1723 when it was decreed that merchants and refiners should pay the same tax as miners.⁴⁸

At Potosí, by contrast, the full fifth was demanded throughout the sixteenth and seventeenth centuries. Thus the cerro rico not merely produced more silver than all camps of New Spain combined, it also yielded the Crown far more direct profit. Naturally, once the rich ores were exhausted, miners found this fiscal burden increasingly onerous, to the point where many, abetted by official laxity, smuggled their silver out through Buenos Aires. Despite various petitions and the prolonged decline in production, it was not until 1736 that the Crown relented and lowered the rate to the tenth. Already several other Peruvian camps had obtained this concession. In 1621, the miners of Castrovirreyna paid only the diezmo, a rate which by 1640 fell to 7½ per cent. Similarly, in this latter year, 10 per cent was collected at New Potosí.⁴⁹ How far the reductions became general is difficult to know; they suggest that, as always, Potosí constituted the greater exception and that other Peruvian camps resembled their Mexican counterparts.

After the Galvez Visitation (1765-71) several leading Mexican miners, all engaged on risky or costly renovations of old mines, received outright exemptions from the silver tithe until their initial investment cost was reimbursed. By the close of the century nearly all the leading enterprises at Zacatecas had secured such concessions.⁵⁰ This policy largely applied to camps with low grade ores; at Guanajuato and Catorce no one was granted an exemption. Whether this liberal fiscal program was ever introduced into Peru remains to be investigated.

The second great source of royal income derived from the Crown’s monopoly over the production and distribution of mercury. During the colonial period this indispensable ingredient was known in sizeable deposits in only three places: at Almadén in Spain, at Idria in modern Slovenia; and at Huancavelica in Peru.⁵¹ Idria, owned by the Austrian emperor, was by far the smallest, and only sporadically entered the American market. The resulting pattern was simple. Huancavelica generally supplied Peru and Almadén supplied New Spain. In both cases failure to produce sufficient mercury could have catastrophic effects upon the local silver mines. At the same time, changes in mercury price could easily alter the curve of silver production. It was this intimate connection which bound the Mexican silver miner to Europe.

Distribution in New Spain remained a royal monopoly managed by treasury officials and licensed magistrates. In the first years, Mexico City auction prices ranged from 132 to 236 pesos a hundredweight. Only slowly, in deference to local protest, did the Crown lower the price from 113 pesos in 1590 to 82½ pesos in 1627, the standard rate for the next century and a half. Northern miners, however, had to pay freight costs from Mexico City. During the later seventeenth century prices rose somewhat. At Zacatecas in the period 1680-1700 the average cost varied between 90 and 100 pesos. This upward trend in price reflected prevalent scarcity. After 1640 Almadén’s production plummeted far beneath the average 4,000-5,000 hundredweight needed by New Spain. During the 1670s Huancavelica sent shipments, but Peruvian mercury cost the Mexican miner 110 pesos, and at best served to supplement rather than replace Almadén as the chief source of supply.⁵²

Although during the seventeenth century Huancavelica produced far more mercury than Almadén, the nature of Andean topography diminished Pern’s advantage as far as prices were concerned. The cheapest route between Huancavelica and Potosí was by sea, via the ports of Chincha and Arica. The Crown contracted with local miners for fixed amounts; by 1600 the usual price it paid was 64 pesos a hundredweight, from which was then deducted the royal fifth. Transport expenses were high, so that cost price at Potosí was estimated at 85 pesos. It was then sold to miners for 102½ pesos. Supply, however, was abundant, since at the beginning of the seventeenth century Huancavelica on average produced 6,000-8,000 hundredweight a year, more than sufficient for the Peruvian industry.⁵³

Under the Bourbons, Almadén took the lead over its Peruvian rival, and especially after 1700 experienced an unparalleled increase, with an annual production which at times exceeded 18,000 quintales. More efficient management cut costs and permitted the Crown to lower the sale price. In two steps, in 1767 and 1776, the Mexican price was cut by half, from 82½ pesos to 41 pesos.⁵⁴ The great Mexican boom in silver production in great measure depended upon this twofold achievement in more than doubling supply at half the former price.

In Peru the same success was not repeated. Huancavelica suffered a decline in the first decades of the century, which, though interrupted by a revival during the 1760s, became permanent following 1779, when Areche terminated the old contract system. Costs remained high, and the Crown could only reduce the sale price to 79 pesos in 1779 and 71 pesos in 1787, a rate far above that charged in Mexico. Huancavelica’s inadequate supply of 2,000 quintales a year had to be complemented by shipments from Almadén. After the creation of the Buenos Aires viceroyalty in 1776, Potosí received most of its mercury from Europe. During the years 1786-96, the Crown obtained over 6,000 hundredweight a year from Idria, a sign of the strain caused by Mexico’s soaring production and Huancavelica’s debility.⁵⁵

Control of the mints provided the Crown with its third source of mining profit. The most obvious difference between the two vice-royalties lay in the location of their mints. Throughout the colonial period until the Wars of Independence, New Spain possessed only one mint, situated in Mexico City. In Peru the Viceroy Toledo established a mint at Potosí in 1572, and in consequence the office in Lima dwindled in importance until 1683, when it was refounded on a new basis. In the same year the Crown decreed that all silver, be it private or royal, must enter the mints for coinage.⁵⁶ Until then much bullion had been exported in bar.

From their initiation until 1728, all three mints followed much the same system: they all cut 69 reales from each mark of silver and returned 65 to the miner. The official grade or quality was 11 dineros 4 granos, or 930.5 milésimos. Most coins of this period, especially those emanating from Potosí, were very crude, resembling stamped pieces of silver rather than true milled coins.⁵⁷ Responsibility for this inferior performance can be attributed in part to the Hapsburg expedient of sale of office. In effect the Crown lost control of the operation; all mint officials bought their positions. Moreover, it was silver merchants rather than the mint itself who purchased silver bar and financed the coining costs.⁵⁸

In 1728 the Mexican mint was brought back under direct royal administration. New ordinances introduced a number of reforms. With new machinery the appearance of coins was vastly improved, even if their quality was reduced to 11 dineros or 916.6 milésimos. The mint continued to cut 69 reales from a mark, but now only returned 64 reales 2 maravedís to the miner. The official exchange ratio of silver to gold was pegged at 16:1, a level considerably lower than the prevailing European rate of just over 14:1.⁵⁹ The construction of a magnificent new building and the creation of a circulating fund to purchase silver upon presentation completed this set of reforms. Henceforth the mint was to yield high profit to the Crown.

In Peru these reforms were instituted somewhat later. The Lima mint was put under direct administration in 1748 and its new building completed only in 1761. Similarly, at Potosí, the veritable palace which housed the mint was constructed in the years 1753-71. Presumably it was then that the Crown appointed salaried officials. However, the creation of the Banco de San Carlos, responsible for the purchase of silver, eliminated the need for a circulating capital fund. Finally, it may be noted that by the close of the century, the Crown had further debased the coinage to 10 dineros 18 granos or 895.8 milésimos.⁶⁰

VI. Capital Structure

Whereas colonial texts deal at some length with the problems of labor supply and technology, they rarely comment upon the organization of the industry or its sources of capital and credit. The problems which confront the historian relate to the size and value of the units of production in both the mining and the refining sectors, the degree of vertical integration, and the precise role of the merchant-financiers to whom occasional brief reference is made. Needless to say, in all these areas great changes occurred in the course of two and half centuries.

Potosí especially remains an enigma. Almost nothing is known about the capital structure behind the massive exploitation of the cerro rico. By 1585 no fewer than 612 mines and shafts honeycombed the peak’s many lodes. Capoche lists about 500 registered mine owners, at the time none of whom employed more than 50 of the 1,369 mita Indians allocated for work within the peak. This system of multiple claims and small individual shafts clearly threw the industry into the hands of azogueros, the owners of the refining mills. True, some of these men, who in all numbered about 75, owned mines within the peak, but the majority obviously bought ore from the miners, or soldados as they were then called. It was the guild of azogueros which dominated the industry and dealt in its name with the Crown. Behind them stood a third group, 12 in number, the silver merchants, who bought silver bar and introduced it into the mint.⁶¹ Possibly, as elsewhere, they also advanced cash on credit to the refiners.

For the moment no further evidence is available concerning Potosí. Any number of questions present themselves. As the cerro rico declined, did the mines consolidate into large enterprises? Did the azogueros take over the peak and establish integrated firms? Or did the miners grow steadily more impoverished and more dependent on the refiners? By the middle of the eighteenth century much of the peak seems to have abandoned to Indian workers, the famous capchas, since by 1774, according to the Viceroy Amat, they accounted for half of Potosí’s production. The other important eighteenth-century development was the creation of the Banco de San Carlos in 1751.⁶² This institution bought bar for cash and hence eliminated the profits of the silver merchants; it also financed the purchase of mining supplies.

For the sixteenth century the picture in New Spain is equally dim. In Zacatecas, a process of concentration was at work. By the middle seventeenth century the earlier multiplicity of claims was succeeded by a group of large enterprises. One miner united four shafts to form one great mine. Also of consequence was the growing reliance upon the silver merchants of Mexico City. Acting through local intermediaries, these finance houses sent cash and supplies to the northern camp in return for bar which they had minted.⁶³ During the next century both of these tendencies reached their culmination.

For the Bourbon epoch the emphasis is on the role of mercantile capital and the tendency for single enterprises to control the exploitation of an entire lode or camp. By then there were silver banks and a handful of great mines with a capital of over a million pesos each. The structure was complex and varied from town to town. In the smaller camps, such as Sombrerete or Bolaños, one great vertically integrated enterprise dominated the industry. In Guanajuato or Catorce, by contrast, the two sectors of the industry remained distinct.⁶⁴ The effect of the Bourbon reforms was to drive mercantile capital into direct investment in mining ventures.

Contemporaries were aware of the great differences between the industries of Peru and New Spain. Cañete lamented that there were no miners in Peru who possessed the capital resources of their Mexican counterparts. It was for this reason that the local mines were so small; their owners did not command sufficient means to exploit even the 200 yards of lode to which they were legally entitled.⁶⁵ Certainly, few references to a mining aristocracy such as developed in Mexico appear in the standard descriptions of Peru.

VII. Production

There are three official sources from which the curve of American silver production can be reconstructed: the royal fifth or tenth; mintage records; and registered silver shipments to Spain.⁶⁶

The traditional view of colonial silver mining during the Hapsburg era has been largely determined by the impressive series of tables and graphs about transatlantic commerce compiled by Hamilton and Chaunu. Hamilton demonstrated that the level of bullion imports registered at Seville sharply increased after the 1550s to reach an early peak by the 1590s. The period of maximum production spanning the years 1580-1630 was followed by a severe depression, to the point where, by the 1650s, imports fell to the level of a century before. The curve of Atlantic shipping and trade drawn by Chaunu largely coincided with this cycle of high boom and sharp depression.⁶⁷

Recently, however, Lynch has criticized the entire notion of a crisis which affected the Peninsula and colonies alike. He argues that “Spain’s recession was America’s growth.” The failure in transatlantic commerce did not necessarily entail an internal depression; instead, it may well have stimulated local industry and trade between colonies. A greater proportion of royal revenue was spent in the colonies. Then again, Bakewell, in his key study on Zacatecas, showed that the camp reached its peak in the 1620s.⁶⁸ The subsequent crisis was severe— production fell by half—but in no sense was as catastrophic as Hamilton’s import figures suggested.

The key to this controversy can be found in the curve of mercury supply. As we have demonstrated above, a straightforward ratio existed between mercury consumption and silver production, a ratio which on average ranged between 100 and 125 marks of silver for every quintal or hundredweight of mercury. Since nearly all the mercury produced by Almadén and Huancavelica went to refine silver in America, a simple calculation will yield estimates of minimum silver output. The maximum, of course, depended upon the percentage of ore which was smelted, a proportion which could vary from 13 to 30 per cent. The chief value of this exercise is that it offers the one reliable check upon the veracity of the fiscal data. All the remaining figures—the quintos, the mint records, the registered shipments—were collected for taxation purposes. Mercury, by contrast, was an industrial supply. Naturally some was lost in transit; a small part was employed for other purposes; but used with caution, the statistics of mercury production and export throw new light upon an old problem.⁶⁹

Here then we encounter our first theme of discussion. What were the true proportions of the seventeenth-century crisis? Did Peru and Mexico vary greatly? How extensive was the first boom? What were the causes for this cycle of boom and depression? Naturally our conclusions are tentative; they are offered as hypotheses to explain the available data. Our innovation consists largely in the comparisons drawn between Peru and Mexico and the correlation of mercury consumption with silver production.

The first significant spurt in silver imports recorded by Hamilton occurred during the 1550s. It sprang from the recent discoveries of Potosí and Zacatecas and the invention in New Spain of the amalgamation process. The second jump forward took place after 1580, following the introduction into Potosí of amalgamation and the construction of refining mills. During the subsequent boom in bullion exports, Peru accounted for 65 per cent of all American silver shipped to Seville.⁷⁰ During this last quarter of the sixteenth century the Andean industry consumed about 70 per cent of all mercury produced. (See Graph III.)

The engines of this first silver boom were the great mines of Huancavelica and Potosí, “the two poles which maintain these Kingdoms and Spain.”⁷¹ Its author was don Francisco de Toledo. This viceroy stamped an impress upon the Andean industry which was not erased until the Wars of Independence. He sponsored the amalgamation experiments at Potosí. He opened up Huancavelica and set up the first contracts. He organized vast mita contingents for both these mines. He established the mint at Potosí. The results of this structural reorganization were impressive and immediate.⁷² Production at Potosí, in the doldrums in 1570, soared to unprecedented heights, and during the ensuing quarter century the cerro rico accounted for at least 70 per cent of all Peruvian silver production, or, to put the case in wider perspective, about half of all American silver.⁷³ The Toledan achievement depended upon a plentiful supply of mercury and a massive input of state-recruited labor. Its advantage—and danger—lay in the extreme physical concentration which permitted economies of scale and easy access to capital. For a brief moment Potosí acted as the magnet for the entire Atlantic economy.

The exceptional quality of the Toledan system is best assessed from the standpoint of New Spain. There, the mining industry expanded at a slower pace. It depended upon Almadén for its mercury. Its work force, by 1598 already dominated by free wage-earners, was sufficient but not ample. The mines were scattered through vast distances from Pachuca to Sonora. The leading camp, Zacatecas, at most accounted for 40 per cent of all production.⁷⁴ These characteristics of geographical dispersion, free labor and dependence upon Almadén prevented the Mexican industry from any rapid emulation of Potosí. If Zacatecas be taken as a guide, then during the 1590s New Spains production did not exceed 4 million pesos, compared to the 10 million pesos of all Peru and the 7 million of the cerro rico.

However, in the three decades which followed, from 1605 to 1635, New Spain’s industry experienced a steady expansion, whereas Potosí declined. A series of contrasts explain the difference. Huancavelica suffered a series of technical difficulties which cut back its output, whereas Almadén, heavily assisted by Idria, shipped increasing quantities of mercury across the Atlantic. In Peru the number of Indians recruited by the mita rapidly fell, so that by mid-century only half the original number appeared for service.⁷⁵ In New Spain the increase in free wage-earners more than offset the disappearance of the repartimiento laborers. Finally, the very rapidity with which the cerro rico had been exploited soon led to the exhaustion of its high grade ores, whereas in New Spain the rich middle zones of the lodes had barely been tapped.

All the long-term trends thus favored Mexico. Yet in the decades after 1635, it was the northern viceroyalty, and not Peru, which suffered the severest depression. The causes of this paradox were absurdly simple. In response to a passing setback at Huancavelica, the Spanish Crown diverted European mercury to Peru (see Graph III). After 1630, quite suddenly, mercury shipments to Mexico dropped by a half. Within five years, when all remaining stocks were exhausted, silver production followed suit.⁷⁶ Moreover, although this preferential treatment for Peru did not extend beyond 1645, by then production at Almadén, following the termination of the Fugger lease, fell from 4,000 hundredweight to just over 2,000 hundredweight, a level insufficient to satisfy Mexican demand (see Graph I). By the 1680s the miners of Zacatecas smelted up to half their ore. Thus we can distinguish short- and long-term causes at work in the Mexican depression. The royal decision to divert mercury shipments to Peru, taken presumably since that viceroyalty paid the quinto whereas Mexico only paid the diezmo, provoked an immediate crisis. But it was the failure at Almadén which forbade any permanent recovery until the next century. Then in the years after 1705 both mercury and Mexican silver production slowly moved upwards.

Granted the “artificial” occasion of the Mexican crisis, there is no reason to expect a similar accident at Potosí. There, throughout the seventeenth century, the curve of total production as measured by quinto payments drifted slowly downwards, from a yearly peak of 7.5 million in the decade 1585-95 to an average 3.2 million in the years 1670-90. Only then did production fall more rapidly, to the point where for the years 1710-30 the quintos registered little more than 1.2 million a year, a mere sixth of the former peak (see Graph II). It was an almost natural process, inherent in the slow exhaustion of the ore deposits, the reduction in labor supply attendant upon the wider phenomenon of Andean population decline and the marked fall in Huancavelica output after 1690.⁷⁷ It represented the slow withering of the Toledan system.

The silver mining industry of both viceroyalties thus experienced a severe depression, but whereas in Mexico this crisis lasted from 1635 to 1689, in Peru its sharpest phase only began in 1680. During the desperate years of 1630-45 when the Spanish monarchy lurched from one financial expedient to another, Peru consumed more mercury than ever before. It follows therefore that Hamilton’s import figures simply reflect official inefficiency and corruption; they bore little relation to the realities of silver production. Similarly, the downward curve registered by the Potosí quintos during these years was possibly false. Even as late as the decade 1671-80, the treasury at Potosí received two-thirds of all mercury distributed in the viceroyalty.⁷⁸ It cannot be argued, therefore, that at this period the decline at the cerro rico was appreciably offset by discoveries elsewhere. Whatever the case, a comparison of mercury consumption with the quinto figures and Hamilton’s series reveals that the major drop in Peruvian silver production probably did not begin until after 1680, and that an overall peak may well have been obtained during the years 1625-40 (see Graph IV and Table I). Mexico’s loss was Peru’s gain.

The same blind search for revenue which had sacrificed Mexican mining drove the Crown to order frequent seizures of private silver consignments, compensating the owners with rapidly depreciating copper coin and treasury notes. In consequence colonial merchants sought extralegal outlets for their silver. In Peru contraband soon reached the epidemic proportions attested to by Acarete du Biscay in his Voyage up the River de la Plata.⁷⁹ The seventeenth-century crisis was first and foremost a Spanish crisis. The catastrophic decline in silver imports was a Spanish rather than a European phenomenon. Throughout the seventeenth century the mines of Mexico and Peru continued to produce great quantities of silver. Possibly more stayed in the New World than before; possibly a greater amount left the hemisphere in the Manila galleon; but without doubt the overwhelming bulk of this bullion went to Europe. Combined American production reached a probable peak in the 1620s, and only slowly declined in subsequent years. The trough occurred not in the 1650s, but in the last twenty years of the century—precisely at the time when the Mexican industry had barely recovered and Peruvian output was falling fast.⁸⁰ It was of course this continuous American production of silver which permitted Europe to export bullion to Asia without experiencing any marked shortage of specie. Colonial silver not only paid for Spanish American imports, it also financed Europe’s trade with Asia.

During the Bourbon epoch, New Spain emerged as the favored possession of the metropolis. Mintage quadrupled from 6 million in 1706 to 24 million in 1798. By then, despite the emergence of Chile and New Granada as silver producers and the revival of the Peruvian industry, Mexico accounted for 67 per cent of the American total. The one camp of Guanajuato equalled the mintage of the entire viceroyalty of Peru, or that of La Plata.⁸¹ This great increase, moreover, despite interruptions caused by war with Great Britain or droughts, followed a consistent upward curve throughout the century, with an especially sharp leap forward in the 1770s. This outstanding achievement rested upon the long-term in-built tendencies of the period prior to 1630. In the first place, Almadén’s production suddenly surged forward from its 2,000 hundredweight of the years 1680-98 to nearly 5,000 hundred-weight and then after 1760 climbed to an unparalleled peak of over 20,000 hundredweight (see Graph I). Without this flood of mercury the great increase in silver production would not have been possible. The second prerequisite was the formation of a permanent, probably hereditary, class of mine workers. Behind the abundant labor supply of this period lay the natural increase of the Mexican population at large, which more than doubled in number between 1740 and 1810. The third element is to be found in the continued geographical dispersion of the Mexican industry. New discoveries joined the older centers, with no camp contributing more than a quarter of all production. The tardy pace of the seventeenth-century cycle had of course left vast ore deposits untapped even in the oldest minefields.

The abundant or adequate supply of mercury, labor and ore deposits formed without doubt indispensable prerequisites of the eighteenth-century boom, but it required the union of technology, capital and government policy with entrepreneurial talent to engineer it. The great shafts and adits of the period represented a great advance in technical expertise, and sprang from extensive capital investment. Bourbon fiscal policy, more enlightened than its Hapsburg equivalent, halved the price of mercury and granted individual tax exemptions for renovations of high risk or cost. Finally the sheer skill and enterprise of such men as José de la Borda and the Count of Regla should not be overlooked: they acted as the pacemakers of the industry.⁸² The Mexican boom, especially after 1770, sprang from as complex a group of elements as the Toledan achievement of the 1570s.

That government policy alone could not revive an industry is demonstrated by the case of Peru. Whereas by 1800 Mexico minted nearly five times as much silver as in 1632, the combined viceroyalties of Peru and Buenos Aires did not regain their 1600 level of 10 million pesos until the 1790s. This contrast began during the later seventeenth century. By the 1690s each colony produced about 5 million pesos, but by the 1720s Mexican mintage rose to nearly 8 million, whereas that of Peru fell still further to an average 3 million.

Peru’s slow recovery took its rise from a modest revival of the Toledan system together with the development of a relatively new industry. The two old centers, Potosí and Huancavelica, after a parallel depression in the years 1680-1724, both picked up after 1730. Huancavelica augmented production from 2,500 hundredweight c. 1710 to an average 6,000 during the 1760s, then collapsed after 1780, never again to yield more than 4,000 hundredweight (see Graph I). Thus the Andean viceroyalties were denied that plentiful supply of mercury which formed the natural prerequisite for any great expansion in silver production. Almadén and Idria were called upon to fill the deficit, but by this time New Spain enjoyed priority. Then again, it appears likely that the Andean population did not experience a natural increase equal to that of Mexico. Moreover, Potosí, and, to a much lesser extent, Huancavelica, still continued to rely upon the dwindling and inefficient mita for cheap labor. Other camps hired free wage-earners, a class about whom, however, little information is available.

The third prerequisite, the discovery of abundant ore deposits, appears to have been satisfied. By 1774 Potosí accounted for only about 40 per cent of the total production of 6½ million.⁸³ The former concentration had been replaced by a Mexican pattern of geographical dispersion. Cerro de Pasco averaged about 2 million pesos at this time. By the 1790s the viceroyalty of Peru minted 6 million pesos compared to the 4½ million of the region, later to be called Bolivia, which under the Hapsburgs had been the chief center of Andean mining. Of that amount Potosí accounted for 3½ million, although about a third derived from mines situated in its district rather than from the cerro rico itself.⁸⁴

Granted the paucity of data, little can be said concerning the role of technology, capital investment or entrepreneurial talent in eighteenth-century Peru. Cañete believed that the viceroyalty trailed behind its Mexican rival in nearly all these aspects of the industry. Royal policy was most successfully embodied in a series of measures to cut costs: the tax reduction of the fifth to a tenth in 1736; the lowering of the mercury price to 71 pesos in 1784; the creation of the Banco de San Carlos to hasten the exchange of silver bar for specie; and finally the construction of a major adit at Potosí. Whatever the precise weight we ascribe to the various elements, it is clear that Crown policy alone could not promote a major boom. Part of the problem lay in the very survival of the Toledan system, with its concentration upon Potosí and Huancavelica, and its reliance upon the mita. By then a wearisome anachronism, the mita virtually subsidized the maintenance of production at Potosí. But the chief force behind the eighteenth-century revival lay in the development of new centers staffed by free wage-earners. In the long term the Mexican structure of production had proved more successful than the Toledan system, and Peru perforce followed in the wake of its northern counterpart.

Author notes