Roman metallurgy

Last updated
Roman Chronology
Context for Metallurgy (Shepard 1993)
circa 753 BCFirst settlement in the Iron Age; see also founding of Rome.
600–524 BC

Etruscans control Italy.

550–500 BC

Carthaginian occupation of parts of Sardinia and Sicily.

Contents

509 BCCreation of the Roman Republic.
510–27 BCRoman Republic and beginning of Rome's expansion.
390 BC

Etruria becomes part of Rome.

264–146 BC

Punic Wars.

197 BC

Iberia becomes a Roman province.

197 BC

Athens becomes a Roman province.

146 BC

Carthage becomes a Roman province.

129 BC

Asia Minor becomes a Roman province.

58–52 BCRoman conquest of Gaul.
30 BC

Egypt becomes a Roman province.

27 BCThe institution of the Roman Empire begins with Emperor Augustus.
44 AD

Britannia becomes a Roman province.

106 AD

Dacia becomes a Roman province.

Roman silver ingot, Britain, 1st-4th centuries AD Roman, Silver ingot (FindID 768389).jpg
Roman silver ingot, Britain, 1st–4th centuries AD
Lead ingots from Roman Britain Lead ingots.JPG
Lead ingots from Roman Britain

Metals and metal working had been known to the people of modern Italy since the Bronze Age. By 53 BC, Rome had expanded to control an immense expanse of the Mediterranean. This included Italy and its islands, Spain, Macedonia, Africa, Asia Minor, Syria and Greece; by the end of the Emperor Trajan's reign, the Roman Empire had grown further to encompass parts of Britain, Egypt, all of modern Germany west of the Rhine, Dacia, Noricum, Judea, Armenia, Illyria, and Thrace (Shepard 1993). [1] As the empire grew, so did its need for metals.

Central Italy itself was not rich in metal ores, leading to necessary trade networks in order to meet the demand for metal. Early Italians had some access to metals in the northern regions of the peninsula in Tuscany and Cisalpine Gaul, as well as the islands Elba and Sardinia. With the conquest of Etruria in 275 BC and the subsequent acquisitions due to the Punic Wars, Rome had the ability to stretch further into Transalpine Gaul and Iberia, both areas rich in minerals. At the height of the Empire, Rome exploited mineral resources from Tingitana in north western Africa to Egypt, Arabia to North Armenia, Galatia to Germania, and Britannia to Iberia, encompassing all of the Mediterranean coast. Britannia, Iberia, Dacia, and Noricum were of special significance, as they were very rich in deposits and became major sites of resource exploitation (Shepard, 1993).

There is evidence that after the middle years of the Empire there was a sudden and steep decline in mineral extraction. This was mirrored in other trades and industries.

One of the most important Roman sources of information is the Naturalis Historia of Pliny the Elder. Several books (XXXIII–XXXVII) of his encyclopedia cover metals and metal ores, their occurrence, importance and development.

Types of metal used

Bronze statuette of Venus, dated to c. AD 118-136. Venus statuette, Roman, c. 118-136 AD, bronze - Middlebury College Museum of Art - Middlebury, VT - DSC07929.jpg
Bronze statuette of Venus, dated to c.AD 118–136.

Many of the first metal artifacts that archaeologists have identified have been tools or weapons, as well as objects used as ornaments such as jewellery. These early metal objects were made of the softer metals; copper, gold, and lead in particular, either as native metals or by thermal extraction from minerals, and softened by minimal heat (Craddock, 1995). While technology did advance to the point of creating surprisingly pure copper, most ancient metals are in fact alloys, the most important being bronze, an alloy of copper and tin. As metallurgical technology developed (hammering, melting, smelting, roasting, cupellation, moulding, smithing, etc.), more metals were intentionally included in the metallurgical repertoire.

By the height of the Roman Empire, metals in use included: silver, zinc, iron, mercury, arsenic, antimony, lead, gold, copper, tin (Healy 1978). As in the Bronze Age, metals were used based on many physical properties: aesthetics, hardness, colour, taste/smell (for cooking wares), timbre (instruments), resistance to corrosion, weight (i.e., density), and other factors. Many alloys were also possible, and were intentionally made in order to change the properties of the metal; e.g. the alloy of predominately tin with lead would harden the soft tin, to create pewter, which would prove its utility as cooking and tableware.

Sources of ore

Las Medulas, remains of the most important gold mine in the Roman Empire. The spectacular landscape resulted from the Ruina montium mining technique Panoramica de Las Medulas.jpg
Las Médulas, remains of the most important gold mine in the Roman Empire. The spectacular landscape resulted from the Ruina montium mining technique
Sources of ore
Ores and Origin (Healy 1978)

Gold

Iberia, Gaul, Cisalpine Gaul, Britannia, Noricum, Dalmatia, Moesia Superior, Arabia, India, Africa

Silver

Iberia, Gaul, Laurion (Greece), Asia Minor, Carmania, Midian, India, Bactria, Britannia, Cyprus

Copper

Iberia, Gaul,

Cisthene, Cyprus, Carmania, Arabia, Aleppo, Sinai, Meroe, Masaesyli , India, Britannia.

Tin

Iberia, Persia, Britannia

Lead

Iberia, Gaul, Sardinia, Sicily, Britannia

Iron

Iberia, Elba, Sardinia, Hallstatt, Noricum, Illyria, Macedonia, Dacia, Sinai, Meroe, Britannia

Zinc

Gaul, Gallia Transpadana, Campania, Germania, Andeira (in Asia Minor), Cyprus

Mercury

Iberia, Armani, Ethiopia

Arsenic

Phalagonia, Carmania

Antimony

Hypothesised:

Mytilene, Chios, around Smyrna, Transcaucasia, Persia, Tehran, Punjab, Britannia

Iberia (modern Spain and Portugal) was possibly the Roman province richest in mineral ore, containing deposits of gold, silver, copper, tin, lead, iron, and mercury. [2] From its acquisition after the Second Punic War to the Fall of Rome, Iberia continued to produce a significant amount of Roman metals. [3]

Britannia was also very rich in metals. Gold was mined at Dolaucothi in Wales, copper and tin in Cornwall, and lead in the Pennines, Mendip Hills and Wales. Significant studies have been made on the iron production of Roman Britain; iron use in Europe was intensified by the Romans, and was part of the exchange of ideas between the cultures through Roman occupation. [4] It was the importance placed on iron by the Romans throughout the Empire which completed the shift from the few cultures still using primarily bronze[ who? ] into the Iron Age.[ citation needed ]

Noricum (modern Austria) was exceedingly rich in gold and iron, Pliny, Strabo, and Ovid all lauded its bountiful deposits. Iron was its main commodity, but alluvial gold was also prospected. By 15 BC, Noricum was officially made a province of the Empire, and the metal trade saw prosperity well into the fifth century AD. [5] Some scholars believe that the art of iron forging was not necessarily created, but well developed in this area and it was the population of Noricum which reminded Romans of the usefulness of iron. [6] For example, of the three forms of iron (wrought iron, steel, and soft), the forms which were exported were of the wrought iron (containing a small percentage of uniformly distributed slag material) and steel (carbonised iron) categories, as pure iron is too soft to function like wrought or steel iron. [7]

Dacia, located in the area of Transylvania, was conquered in 107 AD in order to capture the resources of the region for Rome. The amount of gold that came into Roman possession actually brought down the value of gold. Iron was also of importance to the region. The difference between the mines of Noricum and Dacia was the presence of a slave population as a workforce. [8]

Technology

Roman ingots of lead from the mines of Cartagena, Spain, Archaeological Municipal Museum of Cartagena Lingotesplomocarthagonova.jpg
Roman ingots of lead from the mines of Cartagena, Spain, Archaeological Municipal Museum of Cartagena

The earliest metal manipulation was probably hammering (Craddock 1995, 1999), where copper ore was pounded into thin sheets. The ore (if there were large enough pieces of metal separate from mineral) could be beneficiated ('made better') before or after melting, where the prills of metal could be hand-picked from the cooled slag. Melting beneficiated metal also allowed early metallurgists to use moulds and casts to form shapes of molten metal (Craddock 1995). Many of the metallurgical skills developed in the Bronze Age were still in use during Roman times. Melting—the process of using heat to separate slag and metal, smelting—using a reduced oxygen heated environment to separate metal oxides into metal and carbon dioxide, roasting—process of using an oxygen rich environment to isolate sulphur oxide from metal oxide which can then be smelted, casting—pouring liquid metal into a mould to make an object, hammering—using blunt force to make a thin sheet which can be annealed or shaped, and cupellation—separating metal alloys to isolate a specific metal—were all techniques which were well understood (Zwicker 1985, Tylecote 1962, Craddock 1995). However, the Romans provided few new technological advances other than the use of iron and the cupellation and granulation in the separation of gold alloys (Tylecote 1962).

While native gold is common, the ore will sometimes contain small amounts of silver and copper. The Romans utilised a sophisticated system to separate these precious metals. The use of cupellation, a process developed before the rise of Rome, would extract copper from gold and silver, or an alloy called electrum. In order to separate the gold and silver, however, the Romans would granulate the alloy by pouring the liquid, molten metal into cold water, and then smelt the granules with salt, separating the gold from the chemically altered silver chloride (Tylecote 1962). They used a similar method to extract silver from lead.

While Roman production became standardised in many ways, the evidence for distinct unity of furnace types is not strong, alluding to a tendency of the peripheries continuing with their own past furnace technologies. In order to complete some of the more complex metallurgical techniques, there is a bare minimum of necessary components for Roman metallurgy: metallic ore, furnace of unspecified type with a form of oxygen source (assumed by Tylecote to be bellows) and a method of restricting said oxygen (a lid or cover), a source of fuel (charcoal from wood or occasionally peat), moulds and/or hammers and anvils for shaping, the use of crucibles for isolating metals (Zwicker 1985), and likewise cupellation hearths (Tylecote 1962).

Mechanisation

Drainage wheel from Rio Tinto mines Archscrew2.jpg
Drainage wheel from Rio Tinto mines

There is direct evidence that the Romans mechanised at least part of the extraction processes. They used water power from water wheels for grinding grains and sawing timber or stone, for example. A set of sixteen such overshot wheels is still visible at Barbegal near Arles and dates from the 1st century AD or possibly earlier, the water being supplied by the main aqueduct to Arles. It is likely that the mills supplied flour for Arles and other towns locally. Multiple grain mills also existed on the Janiculum hill in Rome.

Ausonius attests the use of a water mill for sawing stone in his poem Mosella from the 4th century AD. They could easily have adapted the technology to crush ore using tilt hammers, and just such is mentioned by Pliny the Elder in his Naturalis Historia dating to about 75 AD, and there is evidence for the method from Dolaucothi in South Wales. The Roman gold mines developed from c. 75 AD. The methods survived into the medieval period, as described and illustrated by Georgius Agricola in his De re metallica .

They also used reverse overshot water-wheels for draining mines, the parts being prefabricated and numbered for ease of assembly. Multiple set of such wheels have been found in Spain at the Rio Tinto copper mines and a fragment of a wheel at Dolaucothi. An incomplete wheel from Spain is now on public show in the British Museum.

Output

The invention and widespread application of hydraulic mining, namely hushing and ground-sluicing, aided by the ability of the Romans to plan and execute mining operations on a large scale, allowed various base and precious metals to be extracted on a proto-industrial scale only rarely matched until the Industrial Revolution. [9]

The most common fuel by far for smelting and forging operations, as well as heating purposes, was wood and particularly charcoal, which is nearly twice as efficient. [10] In addition, coal was mined in some regions to a fairly large extent: almost all major coalfields in Roman Britain were exploited by the late 2nd century AD, and a lively trade along the English North Sea coast developed, which extended to the continental Rhineland, where bituminous coal was already used for the smelting of iron ore. [11] The annual iron production at Populonia alone accounted for an estimated 2,000 [12] to 10,000 tons. [13]

Annual metal production in metric tons
Output per annumComment

Iron

82,500 t [14] Based on estimate of iron production at 1.5 kg per head in Roman Britain, extrapolated to population size of 55 million for entire empire [15]

Copper

15,000 t [16] Largest preindustrial producer. [17]

Lead

80,000 t [18] Largest preindustrial producer. [19]

Silver

200 t [20] At its peak around the mid-2nd century AD, Roman stock is estimated at 10,000 t, five to ten times larger than the combined silver mass of medieval Europe and the Caliphate around 800 AD. [21]

Gold

9 t [22] Production in Asturia, Callaecia, and Lusitania (all Iberian Peninsula) alone.

Production of objects

Mural in the House of the Vettii in Pompeii depicting cupids using the tools and techniques of Roman goldsmiths Casa Vettii - amorini.jpg
Mural in the House of the Vettii in Pompeii depicting cupids using the tools and techniques of Roman goldsmiths

Romans used many methods to create metal objects. Like Samian ware, moulds were created by making a model of the desired shape (whether through wood, wax, or metal), which would then be pressed into a clay mould. In the case of a metal or wax model, once dry, the ceramic could be heated and the wax or metal melted until it could be poured from the mould (this process utilising wax is called the “lost wax“ technique). By pouring metal into the aperture, exact copies of an object could be cast. This process made the creation of a line of objects quite uniform. This is not to suggest that the creativity of individual artisans did not continue; rather, unique handcrafted pieces were normally the work of small, rural metalworkers on the peripheries of Rome using local techniques (Tylecote 1962).

There is archaeological evidence throughout the Empire demonstrating the large scale excavations, smelting, and trade routes concerning metals. With the Romans came the concept of mass production; this is arguably the most important aspect of Roman influence in the study of metallurgy. Three particular objects produced en masse and seen in the archaeological record throughout the Roman Empire are brooches called fibulae, worn by both men and women (Bayley 2004), coins, and ingots (Hughes 1980). These cast objects can allow archaeologists to trace years of communication, trade, and even historic/stylistic changes throughout the centuries of Roman power.

Social ramifications

Slavery

When the cost of producing slaves became too high to justify slave labourers for the many mines throughout the empire around the second century, a system of indentured servitude was introduced for convicts. In 369 AD, a law was reinstated due to the closure of many deep mines; the emperor Hadrian had previously given the control of mines to private employers, so that workers were hired rather than working out of force. Through the institution of this system profits increased (Shepard 1993). In the case of Noricum, there is archaeological evidence of freemen labour in the metal trade and extraction through graffiti on mine walls. In this province, many men were given Roman citizenship for their efforts contributing to the procurement of metal for the empire. Both privately owned and government run mines were in operation simultaneously (Shepard 1993).

Economy

Roman trade routes, according to the Periplus Maris Erythraei 1st century CE Indo-Roman trade.jpg
Roman trade routes, according to the Periplus Maris Erythraei 1st century CE

From the formation of the Roman Empire, Rome was an almost completely closed economy, not reliant on imports although exotic goods from India and China (such as gems, silk and spices) were highly prized (Shepard 1993). Through the recovery of Roman coins and ingots throughout the ancient world (Hughes 1980), metallurgy has supplied the archaeologist with material culture through which to see the expanse of the Roman world.

See also

Related Research Articles

<span class="mw-page-title-main">Brass</span> Alloy of copper and zinc

Brass is an alloy of copper and zinc, in proportions which can be varied to achieve different colours and mechanical, electrical, acoustic and chemical properties, but copper typically has the larger proportion, generally 66% copper and 34% zinc. In use since prehistoric times, it is a substitutional alloy: atoms of the two constituents may replace each other within the same crystal structure.

<span class="mw-page-title-main">Metallurgy</span> Field of science that studies the physical and chemical behavior of metals

Metallurgy is a domain of materials science and engineering that studies the physical and chemical behavior of metallic elements, their inter-metallic compounds, and their mixtures, which are known as alloys.

<span class="mw-page-title-main">Smelting</span> Use of heat and a reducing agent to extract metal from ore

Smelting is a process of applying heat and a chemical reducing agent to an ore to extract a desired base metal product. It is a form of extractive metallurgy that is used to obtain many metals such as iron, copper, silver, tin, lead and zinc. Smelting uses heat and a chemical reducing agent to decompose the ore, driving off other elements as gases or slag and leaving the metal behind. The reducing agent is commonly a fossil fuel source of carbon, such as carbon monoxide from incomplete combustion of coke—or, in earlier times, of charcoal. The oxygen in the ore binds to carbon at high temperatures as the chemical potential energy of the bonds in carbon dioxide is lower than that of the bonds in the ore.

<span class="mw-page-title-main">Crucible</span> Container in which substances are heated

A crucible is a ceramic or metal container in which metals or other substances may be melted or subjected to very high temperatures. Although crucibles have historically tended to be made out of clay, they can be made from any material that withstands temperatures high enough to melt or otherwise alter its contents.

<span class="mw-page-title-main">Cupellation</span> Refining process in metallurgy

Cupellation is a refining process in metallurgy in which ores or alloyed metals are treated under very high temperatures and subjected to controlled operations to separate noble metals, like gold and silver, from base metals, like lead, copper, zinc, arsenic, antimony, or bismuth, present in the ore. The process is based on the principle that precious metals typically oxidise or react chemically at much higher temperatures than base metals. When they are heated at high temperatures, the precious metals remain apart, and the others react, forming slags or other compounds.

<span class="mw-page-title-main">Roman economy</span> Economy of ancient Rome

The study of the economies of the ancient city-state of Rome and its empire during the Republican and Imperial periods remains highly speculative. There are no surviving records of business and government accounts, such as detailed reports of tax revenues, and few literary sources regarding economic activity. Instead, the study of this ancient economy is today mainly based on the surviving archeological and literary evidence that allow researchers to form conjectures based on comparisons with other more recent pre-industrial economies.

<span class="mw-page-title-main">History of metallurgy in the Indian subcontinent</span>

The history of metallurgy in the Indian subcontinent began prior to the 3rd millennium BCE. Metals and related concepts were mentioned in various early Vedic age texts. The Rigveda already uses the Sanskrit term ayas. The Indian cultural and commercial contacts with the Near East and the Greco-Roman world enabled an exchange of metallurgic sciences. The advent of the Mughals further improved the established tradition of metallurgy and metal working in India. During the period of British rule in India, the metalworking industry in India stagnated due to various colonial policies, though efforts by industrialists led to the industry's revival during the 19th century.

<i>De re metallica</i> 1556 book on mining and metallurgy by Georgius Agricola

De re metallica is a book in Latin cataloguing the state of the art of mining, refining, and smelting metals, published a year posthumously in 1556 due to a delay in preparing woodcuts for the text. The author was Georg Bauer, whose pen name was the Latinized Georgius Agricola. The book remained the authoritative text on mining for 180 years after its publication. It was also an important chemistry text for the period and is significant in the history of chemistry.

Archaeometallurgy is the study of the past use and production of metals by humans. It is a sub-discipline of archaeology and archaeological science.

A doré bar is a semi-pure alloy of gold and silver. It is usually created at the site of a mine and then transported to a refinery for further purification.

In metallurgy, refining consists of purifying an impure metal. It is to be distinguished from other processes such as smelting and calcining in that those two involve a chemical change to the raw material, whereas in refining, the final material is usually identical chemically to the original one, only it is purer. The processes used are of many types, including pyrometallurgical and hydrometallurgical techniques.

<span class="mw-page-title-main">Mining in Roman Britain</span>

Mining was one of the most prosperous activities in Roman Britain. Britain was rich in resources such as copper, gold, iron, lead, salt, silver, and tin, materials in high demand in the Roman Empire. Sufficient supply of metals was needed to fulfil the demand for coinage and luxury artefacts by the elite. The Romans started panning and puddling for gold. The abundance of mineral resources in the British Isles was probably one of the reasons for the Roman conquest of Britain. They were able to use advanced technology to find, develop and extract valuable minerals on a scale unequaled until the Middle Ages.

Liquation is a metallurgical method for separating metals from an ore or alloy. The material must be heated until one of the metals starts to melt and drain away from the other and can be collected. This method was largely used to remove lead containing silver from copper, but it can also be used to remove antimony from ore minerals, and refine tin.

<span class="mw-page-title-main">Mining and metallurgy in medieval Europe</span> Medieval European history

During the Middle Ages, between the 5th and 16th century AD, Western Europe saw a period of growth in the mining industry. The first important mines were those at Goslar in the Harz mountains, taken into commission in the 10th century. Another notable mining town is Falun in Sweden where copper has been mined since at least the 10th century and possibly even earlier.

Mining archaeology is a specific field well-developed in the British Isles during recent decades. A reason of ongoing interest in this field is the particular bond between regional history and the exploitation of metals. References to mines in the area exist in Strabo's works. However the first accomplished study on the topic was attempted by Oliver Davies in 1935. Other momentous researches were that of geologist John S. Jackson about mines in Ireland and Lewis, Jones in Dolaucothi goldmine in Wales, and the pioneering work of Ronald F. Tylecote. Moreover, in the 1980s and 1990s a new generation of amateurs and scientists began investigations in different locations in the British Isles, including Duncan James on the Great Orme's Head, Simon Timberlake with the Early Mines Research Group at sites in Wales and William O'Brien in Ireland.

<span class="mw-page-title-main">Metallurgical assay</span> Compositional analysis of an ore, metal, or alloy

A metallurgical assay is a compositional analysis of an ore, metal, or alloy, usually performed in order to test for purity or quality.

Experimental archaeometallurgy is a subset of experimental archaeology that specifically involves past metallurgical processes most commonly involving the replication of copper and iron objects as well as testing the methodology behind the production of ancient metals and metal objects. Metals and elements used primarily as alloying materials, such as tin, lead, and arsenic, are also a part of experimental research.

<span class="mw-page-title-main">Conservation and restoration of copper-based objects</span>

The conservation and restoration of copper and copper-alloy objects is the preservation and protection of objects of historical and personal value made from copper or copper alloy. When applied to items of cultural heritage, this activity is generally undertaken by a conservator-restorer.

<span class="mw-page-title-main">Lead smelting</span> Process of refining lead metal

Plants for the production of lead are generally referred to as lead smelters. Primary lead production begins with sintering. Concentrated lead ore is fed into a sintering machine with iron, silica, limestone fluxes, coke, soda ash, pyrite, zinc, caustics or pollution control particulates. Smelting uses suitable reducing substances that will combine with those oxidizing elements to free the metal. Reduction is the final, high-temperature step in smelting. It is here that the oxide becomes the elemental metal. A reducing environment pulls the final oxygen atoms from the raw metal.

<span class="mw-page-title-main">Non-ferrous extractive metallurgy</span> Metallurgy process

Non-ferrous extractive metallurgy is one of the two branches of extractive metallurgy which pertains to the processes of reducing valuable, non-iron metals from ores or raw material. Metals like zinc, copper, lead, aluminium as well as rare and noble metals are of particular interest in this field, while the more common metal, iron, is considered a major impurity. Like ferrous extraction, non-ferrous extraction primarily focuses on the economic optimization of extraction processes in separating qualitatively and quantitatively marketable metals from its impurities (gangue).

References

  1. F., Healy, J. (1978). Mining and metallurgy in the greek and roman world. Thames & Hudson. OCLC   463199001.{{cite book}}: CS1 maint: multiple names: authors list (link)
  2. Healy 1978
  3. Healy 1978, Shepard 1993
  4. Aitchison, 1960
  5. Shepard 1993, Healy 1978
  6. Aitchison, 1960
  7. Sim 1999, Aitchison 1960
  8. Shepard 1993)
  9. Wilson 2002, pp. 17–21, 25, 32
  10. Cech 2010, p. 20
  11. Smith 1997, pp. 322–324
  12. Ian Morris, Francoise Audouze, Cyprian Broodbank (1994): Classical Greece: Ancient Histories and Modern Archaeologies, Cambridge University Press, p. 102 ISBN   978-0-521-45678-4
  13. Wertime, Theodore A. (1983): "The Furnace versus the Goat: The Pyrotechnologic Industries and Mediterranean Deforestation in Antiquity", Journal of Field Archaeology, Vol. 10, No. 4, pp. 445–452 (451); Williams, Joey (2009): "The Environmental Effects of Populonia's Metallurgical Industry: Current Evidence and Future Directions", Etruscan and Italic Studies, Vol. 12, No. 1, pp. 131–150 (134f.)
  14. Craddock 2008, p. 108; Sim, Ridge 2002, p. 23; Healy 1978, p. 196
  15. Sim, Ridge 2002, p. 23; Healy 1978, p. 196
  16. World output, the large bulk of which is attributed to Roman mining and smelting activities (mainly in Spain, Cyprus and Central Europe): Hong, Candelone, Patterson, Boutron 1996, p. 247; Callataÿ 2005, pp. 366–369; cf. also Wilson 2002, pp. 25–29
  17. Hong, Candelone, Patterson, Boutron 1996, p. 247, fig. 1 & 2; 248, table 1; Callataÿ 2005, pp. 366–369
  18. World output, the large bulk of which is attributed to Roman silver mining and smelting activities (in Central Europe, Britain, the Balkans, Greece, Asia Minor and, above all, Spain, with a 40% share in world production alone): Hong, Candelone, Patterson, Boutron 1994, p. 1841–1843; Callataÿ 2005, pp. 361–365; Settle, Patterson 1980, pp. 1170f.; cf. also Wilson 2002, pp. 25–29
  19. Hong, Candelone, Patterson, Boutron 1994, p. 1841–1843; Settle, Patterson 1980, pp. 1170f.; Callataÿ 2005, pp. 361–365 follows the aforementioned authors, but cautions that the Greco-Roman levels may have already been surpassed by the end of the Middle Ages (p. 365).
  20. Patterson 1972, p. 228, table 6; Callataÿ 2005, pp. 365f.; cf. also Wilson 2002, pp. 25–29
  21. Patterson 1972, p. 216, table 2; Callataÿ 2005, pp. 365f.
  22. Pliny the Elder: Naturalis Historia, 33.21.78, in: Wilson 2002, p. 27

Sources

General
Output

Further reading