CHAPTER 9 Naturam ars imitata: European Brassmaking between Craft and Science THILO REHREN AND MARCOS MARTINÓN-TORRES ABSTRACT This chapter presents a summary of analytical work on medieval brassmaking crucibles, spanning more than half a millennium and tracing what is believed to be a gradual development of increasing skill and efficiency of the craftspeople who used these crucibles. This summary is then contrasted to a similarly diachronic sequence of textual sources concerned with brassmaking, illuminating the discrepancy between the matter-of-fact practitioners’ reports and the somewhat befuddled attempts of philosophers of nature to understand and explain the essence of brass as opposed to copper. The different strands of knowledge generation and transfer, namely, observation and apprenticeship versus theoretical consideration and text, are placed into the changing scholarly environment from the High Middle Ages to the Renaissance. This exercise provides fresh insight not only into the development of brassmaking technology but also into the driving, or otherwise, forces in technological developments in general. The comparison of archaeological, scientific, and historical evidence is used to demonstrate the potential of such multisource studies. Introduction Brass has a more chequered history than most other alloys do, so much so that despite more than three millennia of brass use, it finally was acknowledged as an alloy only less than 500 years ago. For most of its history, brass caused wonder and stimulated the minds of people trying to understand what it really was – a process that in some sense continues to 167 168 CHAPTER 9 this day. The reason why it took so long for brass to be recognised as an alloy has been discussed elsewhere in length (de Ruette 1995; Zacharias 1989); it suffices here to summarise the key points. The term brass refers to an alloy of copper and zinc. Zinc concentrations in archaeological brass range from about 5 wt% at the lower end to about 30 wt% at the upper end. Practically, there is no lower limit for the zinc concentrations in copper: it is a semantic question when we start calling zinc-containing copper ‘brass’, and when we consider it an impure copper, or copper with just trace or low concentrations of zinc. Archaeologically, though, this is more than a linguistic question, as the term brass (rather than impure copper) strictly speaking implies an intentionally produced alloy, as opposed to an accidentally produced mixture of metals. The problem with zinc as a metal is that at temperatures above 907°C it exists only as a vapour, but most smelting furnaces operate at much higher temperatures. Thus, metallic zinc on its own was almost unknown in Antiquity in the West (but see Rehren 1996, and Rehren in Fellmann 1999, and literature therein), ruling out the traditional way of alloying by co-fusing existing metals, as done with copper and tin to produce bronze. Instead, brass was produced by a technique called cementation, which allowed the manufacture of the copper-zinc alloy several centuries before unalloyed zinc metal could be exploited. In this process, metallic copper, charcoal, and powdered zinc ore (usually carbonates or oxides) were heated together in a crucible, so that zinc vapour formed and reacted with the copper metal, thus producing brass. Significantly, and owing to the high vapour pressure of zinc already well below its boiling point, this process can proceed at temperatures as low as 800°C. What is important here is that, contrary to the more common alloying method, only one metal is seen to enter the reaction vessel (copper), and a different type of metal is produced (brass), hence the problems in the understanding of the process. To facilitate the reaction, the crucible had to be closed to force the zinc vapour into the copper, rather than it escaping with the fumes of the furnace. Brass with the typical Roman level of approximately 15 to 25 wt% zinc melts at 1050 to 950°C, which is well above the temperature necessary for the cementation process but still lower than the temperature required for smelting copper, of around 1100 to 1200°C. Thus, depending on the composition of the forming brass and the operating temperatures, the resulting alloy would be either solid or liquid. The upper zinc limit for cementation brass has been intensively disputed. Empirically, very few well-dated objects are known with more than about 32 wt% zinc, although experimental work indicates that higher zinc concentrations can be achieved by both classical and modern methods of brass production (Haedecke 1973; Newbury, Notis, and THILO REHREN AND MARCOS MARTINÓN-TORRES 169 Newbury 2005; Ullwer 2001; Welter 2003; Werner 1970; Zwicker et al. 1985). This, however, has no particular implications for the perception of the material, and we are not addressing this aspect here any further. This chapter looks at the changes in the perception of brass, as well as the understanding and explanation of the principles behind its production, taking place in Europe over the last millennium. In contrast to most papers on medieval and later brass, we do not focus on analytical data of objects themselves but juxtapose historical texts addressing the material, its properties and its production, with archaeological evidence for the latter. In doing so, we hope to identify and explain the transition from an intuitive process of improving brassmaking practice, based on tradition and practical replication with modification on the side of the practitioners, to a period of more conscious developments, based on a theoretically framed transcendence of the process on the side of scholars, and possibly involving experimental work. In addition, we attempt an insight into the evolving relationship between theory and practice in this arena, and the impact of the Renaissance rationality and empiricism in the making and understanding of brass. Here, we are able only to scratch the very surface of this topic, which is clearly much richer archaeologically and much more complex in its textual sources than can be dealt with in the space of a few thousand words. However, we hope to show some of the potential in this subject for in-depth study. Premedieval Brassmaking in the Old World By way of introduction, and as a backdrop against which European brassmaking is to be understood as fundamentally different from the other alloys inherited from Antiquity, we briefly outline three main phases of brass use, mainly in Europe, separated by periods for which we have little or no evidence for brass being made or widely known. The first phase of brass use covers the Bronze Age and Early Iron Age of the Near and Middle East. The number of brass artefacts now known from the eastern Mediterranean and Western Asia is so big that they can no longer be ignored as flukes of chance (Craddock and Eckstein 2003; Thornton and Ehlers 2003). This archaeological presence broadly correlates with textual reports in the archaic Greek literature of a particular and highly prized copper alloy called oreichalkos, or ‘mountain copper’, now widely believed to be brass (Craddock and Eckstein 2003:217). In the 1st century C.E., Pliny refers to aurichalcum (literally ‘golden copper’ but phonetically an adaptation of the Greek term) in a way suggesting that by his time this alloy was no longer being made (Hist. Nat. XXXIV:2–4; 170 CHAPTER 9 translated in Rackham 1952). In the absence of any published archaeological evidence for the production of this alloy in pre-Roman times, and given its typically low and rather variable zinc content, one may assume that it was produced by smelting a particular zinc-rich copper ore. This assumption is consistent with the use of a specific term for this metal, identifying it as a variant of copper but sufficiently different for it to not be included in the generic chalkos (aes in Latin) used for copper and bronze. It is also consistent with the archaeological disappearance of this alloy, possibly because of simple exhaustion of the particular ore deposit(s) from which it was smelted. At least during the final phase of this period, in the Iron Age, oreichalkos seems to have enjoyed a reputation as a rare and desirable material, worthy of use by the best of the time, mostly for jewellery and ornaments. The second phase in this history spans from the mid-1st century B.C.E. up to the late Roman period. It starts with the sudden appearance of small quantities of brass both in the East and to the north and west of the Roman empire, followed by a boom of brass use, first in military implements and other state-controlled contexts such as currency, during the late first century B.C.E. and the first century C.E. (Bayley 1998; Ponting 2002; Ponting and Segal 1998; Weeks 2004). This gradually develops into a period of more widely distributed, and literally also diluted, use, when brass and zinc-containing ternary alloys permeate into general domestic use. The supply of fresh brass during this later stage of the second phase seems to have dried up, as no or only very little high-zinc, high-quality brass is still being used then (Bayley and Butcher 2004; Dungworth 1996, 1997a, 1997b). There is a considerable body of literature on alloy compositions from this period (see references above), but only very few instances of brassmaking installations have been published so far. These have in common that they clearly point toward the conscious production of brass in closed crucibles using the cementation method, enabling relatively close control over the resulting alloy composition. Those few case studies reported so far paint a rather diverse picture, ranging from tiny purpose-made acorn-shaped (and -sized!) crucibles used in Colonial Ulpia Traiana/ Xanten (Rehren 1999a) to large amphorae-like vessels used for brassmaking in Lugdunum/Lyon (Picon, Le Nezet-Celestin, and Desbat 1995) and normal, fist-sized, metallurgical crucibles in Roman Britain (Bayley 1998) (Figure 9.1). These vessels are all conspicuously free of slag, suggesting that the process was probably done at a relatively low temperature, possibly involving only solid copper (but see Craddock and Eckstein 2003). The resulting brass, thus formed in the solid state, would probably be melted and further refined only when needed for casting. The cementation crucibles were lidded and tightly sealed, and the pressure released THILO REHREN AND MARCOS MARTINÓN-TORRES 171 Figure 9.1 • Formal diversity in Roman brassmaking crucibles; from left to right, examples of crucibles from Xanten (Germany); Culver Street in Colchester (United Kingdom) (© Justine Bayley); Palace Street in Canterbury (United Kingdom) (© Justine Bayley); and Lyon (France) (© Société Française d’Étude de la Céramique Antique en Gaule). Note the different scale of the Lyon crucible, which would otherwise appear even larger when compared to the rest; see text for references (all crucibles redrawn by M. Martinón-Torres). probably through the porous ceramic fabric and, in some cases, though a hole in the lids. Remarkable is the wide range in vessel sizes between the three more or less contemporary published production sites mentioned above, indicating that a degree of variability existed between different production centres, while the basic principle was the same for all three. The systematic and consistent production of brass of a certain quality clearly demonstrates full mastery of the process during the early part of this second phase of brass use, while textual references show that brass was understood as a derivative of copper. Unfortunately, Pliny does not discuss its production, although he details that ‘Livian’ copper from Gallia/Gaul is better suited to be treated with cadmea (zinc ore) than that from other sources and produces a metal almost as good as the traditional oreichalkos (Hist. Nat. XXXIV:2–4, 100–103; translated in Rackham 1952). Discussion of the reason(s) for the disappearance of freshly made brass during the later Roman Empire is beyond the scope of this paper (cf Dungworth 1996). Pliny’s failure to report any technical details may indicate that the practicals of brassmaking were not in the public domain and that it was a loss of know-how or state-controlled demand rather than exhaustion 172 CHAPTER 9 of raw materials that led to its demise. Hence, we have a fair idea of the production and perception of brass as a relatively cheap copper-related metal during the Roman period but no feeling whether the Roman or indigenous brassmakers considered their craft different from that of other alloy makers, nor why its production in the West seemingly ended well before the collapse of the Roman Empire. Medieval and Early Modern Brassmaking in Europe The third phase of brass use in Europe begins in the High Middle Ages and continues to this day. For this phase, we have the most comprehensive archaeological and textual evidence, and the emphasis of this chapter is very much on the first half of this third phase, that is, roughly from 1000 to 1600 C.E. Brass was again being made by cementation, soon replacing bronze as the alloy of choice. It is this classical method of brass production, prevailing in Europe and the Middle East until the Industrial Revolution, that led to so much confusion about the nature of brass. As noted above, metallic zinc occurs in this process only ever as a vapour phase within the reaction vessel, invisible to the metal worker of the time and unknown to the scholars interested in the materials of nature. Accordingly, the actual amount of zinc entering the copper was less well-controlled than in other alloys. Furthermore, this peculiarity raised serious doubts whether brass really qualified as an alloy or whether it was just a particular type of copper ‘coloured’ yellow. It is this enigmatic change in the properties of copper, similar to the change of properties known from ‘proper’ alloying when making bronze, but not linked to any known metal, that puzzled scholars for many hundreds of years. However, this uncertainty did not preclude generations of metal workers from mastering and perfecting the process, so much so that brass quickly became the dominant copper alloy of the medieval and the modern periods, extensively used for everyday life implements as much as for civil and ecclesiastical works of art – as widely documented in analytical studies (for example, Cameron 1974; Caple 1995; Hachenberg 2004; 2006; Mitchiner, Mortimer, and Pollard 1986; Pollard and Heron 1996; de Ruette 1996; Riederer 1988). The six centuries of brass use starting around 1000 C.E. are of particular interest here, since they provide both a unique sequence of archaeological evidence for brassmaking and a series of highly illuminating texts about the nature of this particular alloy, written by some of the brightest minds of their time. In the following text, we summarize the archaeological evidence for the gradual evolution of brassmaking in Central Europe, from the late 1st millennium C.E. up to the Renaissance. This summary will THILO REHREN AND MARCOS MARTINÓN-TORRES 173 then be contrasted with some of the contemporary sources discussing the nature of this alloy and its production, with the aim to understand the changing perception of brass. Eventually, we look at the relationship between the practical mastery and the theoretical understanding of brassmaking, in an attempt to judge the degree of mutual dependency of practitioners and theoreticians in the field. The Evolution of Brassmaking: Archaeological Evidence As noted above, the third phase in the history of European brassmaking starts after the mid-1st millennium C.E., with fresh brass emerging in Scandinavian and Viking contexts (Eremin, Graham-Campbell, & Wilthen 2002). By the time of the Carolingian revival, brass is the main copper alloy in use, and it will retain this position to this day. Thus, brass was and is clearly an alloy of huge economic significance, and its production and use played a central role in many aspects of the development of the modern world. Little, however, do we know of the early period of its production, and the source of the Viking-period brass remains enigmatic. The first archaeological evidence for brassmaking, from Dortmund in western Germany, dates to the late 1st millennium C.E., followed by a series of somewhat later instances excavated across Germany. The evidence from Soest and Schwerte, some 50 km to the east of Dortmund, dates to the first half of the 2nd millennium C.E.. The latest evidence of relevance here is from Zwickau in southeast Germany, dating to the late 15th century C.E.. Of interest to us is the evolution of the crucible design over the period of half a millennium, particularly in size and profile. On the contrary, we do not address in any details the zinc ores and their preparation, for which the existing information is even scantier and has been summarised elsewhere (Craddock and Eckstein 2003). It suffices to remember that two main types of zinc ores were used: on the one hand, ‘natural’ minerals in the form of calcined zinc carbonate (ZnCO3) or roasted sphalerite (ZnS) – generally called calamine, cadmia fossilis, galmey, and so on, depending on the authors; on the other hand, ‘artificial’ zinc oxiderich crusts scraped as byproducts from the inner walls of furnaces where zinc-bearing lead ores were smelted – named furnace calamine, cadmia fornacum, tutty, and others (cf de Ruette 1995). The Dortmund crucibles are short and cylindrical in shape, about 6 to 8 cm in diameter, and usually less than 10 cm high, corresponding to a volume of less than half a litre (Figure 9.2). Little if any evidence for a lid exists for these vessels, suggesting a rather wasteful process with a high proportion of the zinc vapour escaping from the reaction vessel instead 174 CHAPTER 9 Figure 9.2 • Medieval and Renaissance brassmaking crucibles from Germany; from left to right, in chronological sequence, sketch drawings of crucibles from Dortmund; Soest/Schwerte; and Zwickau; see text for references. of being absorbed by the copper to form brass (Rehren et al. 1993). The walls are around 1 cm thick, their fabric is dense, and, with over 27 wt% Al2O3 (normalised data after Rehren 1999b; Martinón-Torres 2001), much more alumina-rich and hence more refractory than the contemporary local domestic wares. The Soest and Schwerte vessels, roughly thought to be two to three centuries more recent, are considerably bigger than the Dortmund crucibles. They are tubular with a round bottom, an internal diameter of around 9 cm and a height of at least 15 to 20 cm (Figure 9.2). The ceramic is visually almost indistinguishable from the earlier Dortmund crucibles and, although less rich in alumina (19 wt% Al2O3; normalised data after Rehren 1999b) still sufficiently refractory to serve its purpose. Significantly, though, there are several ceramic disks among the assemblage from Soest which were tentatively identified as lids. These would have to some extent limited the zinc loss, even though they probably only loosely fitted the mouths of the crucibles. In all these cases, the evidence is for a cementation process as in the Roman examples, but invariably involving slag formation within the THILO REHREN AND MARCOS MARTINÓN-TORRES 175 crucibles. This suggests that the operating temperatures were now on the whole higher than in the Roman period and probably involved melting the brass in the crucibles (Rehren 1999b). Another several hundred years later, in the late 15th century C.E., brassmaking vessels had apparently moved up further in scale and technical design. At Zwickau, brassmaking vessels were by now holding about 20 kg of metal, having a height of at least 40 cm and a body with a bulbous profile reaching 28 cm in diameter, resulting in a volume of approximately 12.5 litres (Martinón-Torres and Rehren 2002) (Figure 9.2). They are made from a ceramic of only medium alumina content (around 17 wt%), making them less refractory than the earlier vessels from Dortmund and similar to those from Soest and Schwerte. However, the mechanical stability of these large vessels was ensured by walls that are rather thick, at around 2.5 cm, and further covered with an outer layer of even less refractory ceramic – an established method of improving performance of technical ceramics (Rehren 2003). Furthermore, the openings of the Zwickau vessels were covered with a sophisticated dome-shaped lid, itself comprising a smaller opening that was closed with a lump of clay if and when needed. While a thin film of slag appears inside some of these vessels, this is a reaction layer between the ceramic fabric, zinc oxide, and an iron oxide-rich gangue, but it appears that the brass did not melt during the cementation process, and there is evidence, within the same assemblage, for the melting and refining of fresh brass in different, triangular crucibles (Martinón-Torres and Rehren 2002). Thus, overall we see a growth in crucible volume from the small cups of Dortmund to the buckets of Zwickau, with the ceramic design being adapted to match the physical demands of the increased weights of the charge (Figure 9.2). In addition, the opening of the crucibles and their lids become much more sophisticated, obviously in an attempt to minimise loss of zinc vapour while still maintaining access to manipulate the charge with iron rods (see below). Finally, the later crucibles indicate a better control of the working temperatures, which would result in a more economic use of the raw materials and higher-quality brasses. Altogether, it is tempting to picture a technology being passed on from one generation of craftspeople to the next, whereby brassmaking improves by experience and through iteration but no significant innovation or breakthrough is noticeable. Scholarly Understanding of Brassmaking? Historical Evidence The Middle Ages The earliest detailed description of brassmaking is by Theophilus, dating to the 12th century C.E. and therefore roughly contemporary to the 176 CHAPTER 9 evidence from Soest und Schwerte. In his third book of On divers arts, he advises: And when the crucibles are red-hot take some calamine, about which I spoke above, that has been [calcined and] ground up very fine with charcoal, and put it into each of the crucibles until they are about one-sixth full, then fill them up completely with the above-mentioned copper, and cover them with charcoal . . . Now, when the copper is completely melted, take a slender, long, bent iron rod with a wooden handle and stir carefully so that the calamine is alloyed with the copper. Then with long tongs raise each crucible slightly and move them a little from their position so that they may not stick to the hearth. Put calamine in them all again as before and fill them with copper and cover them with charcoal. When it is once more completely melted, stir again very carefully and remove one crucible with the tongs and pour everything into little furrows cut in the ground. Then put the crucible back in its place. (Hawthorne and Smith 1979:143–44) The chronologically next source, discussing the nature of brass as well as its production, is the Book of Minerals by Albertus Magnus, written probably in the early second half of the 13th century. In Chapter 6 of the fourth book he explains the nature of copper as a result of the combination of the particular qualities and quantities of quicksilver and sulphur that make up all metals. For him, Quicksilver is good, not full of dross and dirt, but still not completely cleansed of extraneous moisture; and… the substance of the Sulphur is full of dross, burning hot and partly burnt, and in this condition it is mixed with the Quicksilver, both in substance and in quality. Then undoubtedly it changes the copper to a red colour; and because neither [the Sulphur nor the Quicksilver] is sufficiently subtle, they cannot be well mixed. And this will make copper, which is not at all well mixed, since much dross is separated from it, and it evaporates greatly in the fire. (Wyckoff 1967:223) He then goes on to explain how and why the copper is not well mixed and burns, concluding that: ‘Now, therefore, we understand the material of copper; it is a metal having rather more Quicksilver than it ought to have, which has been converted into a red form by mixture with burning Sulphur’ (Wyckoff 1967:224). Building on this understanding of copper, he then explains the making of brass as follows: But those who carry on much work with copper in our region – that is, in Paris and Cologne and other places where I have been and seen this tested by experience – convert copper into brass [aurichalcum] by means of a powder of a stone called calamine. And when this stone evaporates there still remains a dark lustre, approaching the appearance of gold. And to make it paler in colour, and so more like the yellow of gold, they mix in a little tin; THILO REHREN AND MARCOS MARTINÓN-TORRES 177 but because of this, brass loses the malleability of copper. And those who wish to deceive and to produce a lustre like gold ‘bind’ the stone so that it may remain longer in the copper on the fire, and not evaporate from it so quickly. And the ‘binding’ [is done] with ‘oil of glass’. They take fragments of glass, crushed and sprinkled into the crucible on the copper after the calamina is put in; and then the glass that has been put in floats on the top of the copper and does not allow the power of the stone to evaporate, but reflects the vapour of the stone down into the copper. And in this way the copper is thoroughly purified for a long time and the drossy material in it is burnt up. But after a while the oil of glass evaporates, and then the power of the stone evaporates, too; but the brass is made much more brilliant than it would have been without it . . . But Hermes says that if powdered tutty is mixed with molten copper – either white tutty or red – it changes the copper to the colour of gold. What tutty is will be explained in the following book, where ‘intermediates’ are treated. But it is enough [to say] here that the burning heat of tutty consumes the earthiness and purges the superfluous moisture out of the copper; and so then it will be more beautiful. But the power of tutty, too, evaporates if it stands for a long time on the fire; and therefore, unless some remedy is used, the tutty will evaporate and the copper will regain its original colour. (Wyckoff 1967:224–25) These two texts, separated by only about a century, show two radically different approaches to the topic. Theophilus gives a very matter-of-fact description, a recipe, of how to make brass; he does not seem to spend much thought on what actually happens to the copper in the process. Clearly, brass is an important material to him, and he spends several paragraphs on the best preparation of the furnace for brassmaking, as well as on the different qualities of brass: qualities that are of relevance if the metal subsequently is to be gilded or hammered rather than cast. Based on this text, any dexterous person can successfully make brass, regardless of whether they have a degree in science, or philosophy, or are barely literate enough to read his text. No word is spared on theoretical speculations. Thus it is not surprising that the contemporary archaeological evidence from brass workshops is much more easily explained in the light of this text. Albertus Magnus, in stark contrast, explains in much more detail the basic constituents that make up the various metals, manifested in their different degrees and qualities of coolness, humidity, and ‘burning heat’, and as embodied in the principles for the generation of all metals: quicksilver (mercury) and sulphur. He then places copper within this theoretical framework, which he carefully developed from the scriptures of the classical authors, before addressing brass. The relevant sentences here are an interesting mixture of plain observations of current practice in Paris and Cologne and attempts to accommodate what he noticed within his theoretical framework of varying ‘qualities’ of metal that have to be ‘purged’ of inferior aspects. 178 CHAPTER 9 It would be quite difficult to follow his description in practice, though. For instance, no mention is made by him of the crucial ingredient charcoal, necessary for the zinc ore to be reduced to metal vapour and without which the whole cementation process would not work. However, Albertus devotes many words to discuss the addition of crushed glass as a protective coating for the melt, which, from a present-day perspective, is a less important aspect of the process. The ensuing discussion demonstrates some form of practical understanding that the central component zinc is volatile and should be kept from evaporating. However, the agent that he describes to evaporate, and that he clearly links to the change in colour from red copper to yellow brass (and back to red copper through evaporation), is seen more as an immaterial quality, a ‘power’ of tutty, rather than substantially the tutty itself – which is, of course, zinc oxide. The apparent discrepancy could not be stronger, but it can be somewhat explained by considering the backgrounds and aims of both authors. Theophilus was a Benedictine artisan monk extensively experienced in arts and crafts, he collected practical recipes as a memory aid, and showed no concern with explanations. The result was a book that, from a technical viewpoint, was not superseded ‘until the books by Cennini (1437) on pain-ting, Månsson (ca 1520) on glass, and Biringuccio (1540) on almost everything but painting’ (Hawthorne and Smith 1979:xxxi). Albertus Magnus, widely acknowledged as one of the greatest thinkers and most prolific writers of the Middle Ages, held high office in the Dominican Order and later served as a bishop, and was eventually canonised in 1931. From a higher social rank and in a completely different academic pursuit, what he attempted was an explanation for the nature and formation of the natural world that could reconcile Aristotle’s four elements, Avicenna’s sulphur-quicksilver theory, further alchemical beliefs, and even Christian theology. Returning to brassmaking, his main objective was not to produce a beautiful metal of appropriate material properties but to investigate the causes behind its formation and, ideally, to extract relevant information that would aid the search of metallic transmutation. Medieval alchemists sought the ‘tincture’ or ‘elixir’ that, added to base metals, would give them the quality – that is, the actuality – of gold. It was in this manner that they sought to imitate natural gold with artifice. The sensorial qualities of artificial brass did indeed compare reasonably well those of natural gold, but, to the alchemists’ despair, the yellow colour of brass (that is, the zinc) evaporated on prolonged melting. In another text attributed to Albertus, several paragraphs insist that ‘the alchemical art requires that some means be discovered whereby the spirit does not escape, is not consumed or burned, but instead penetrates and THILO REHREN AND MARCOS MARTINÓN-TORRES 179 is mixed with all parts of the metal, so that it may tint with a permanent tincture’ (Kibre 1944:305). In these pages, tutia (zinc ore), marcasita (iron sulphide), and sale armoniaco (ammonium chloride) – all of which contain volatile elements – are mentioned as potential ‘tinctures’ that need to be processed and improved. It is probably within this context that Albertus calls ‘deceivers’ (see above) those who sprinkle broken glass in the crucibles to prevent evaporation, since this does not solve the problem of the ‘permanent tincture’. Avoiding colour losses was good enough for a metallurgist, but not acceptable for a demanding scholar. All in all, these written sources demarcate two distinct traditions: on the one hand, those such as Theophilus, concerned with the practicalities; on the other hand, those pursuing explanations, as was the case of Albertus Magnus. Not only Theophilus but also Albertus had some firsthand acquaintance with metallurgical practice – probably obtained in the Rammelsberg some 30 years before writing his book (Wyckoff 1967). His theoretical speculations bear some technical basis, but craftsmen and scholars are working in different spheres. Meaningfully, Albertus disregards Pliny because ‘he does not offer an intelligent explanation of the causes common to all stones’ (Wyckoff 1967:10), and nothing in The Book of Minerals suggests that Albertus, the great encyclopedist, ever read Theophilus. The Renaissance The 16th century saw a great expansion in printing technology, together with other technological and scientific achievements, and a fervent humanistic interest in understanding nature and recording knowledge. In this context, several major treatises on mineralogy and metallurgy appeared, where we find detailed accounts on brassmaking. Three crucial texts must be considered here: Vannoccio Biringuccio’s De la pirotechnia, first published in 1540, Georg Agricola’s De natura fossilium – notably the second edition of 1558 – and Lazarus Ercker’s 1580 Beschreibung aller fürnemisten Mineralischen Ertzt vnnd Berckwercksarten. How do they differ from the earlier texts? Biringuccio gives a brief description very similar to Theophilus’s, although he fails to mention the necessary charcoal charge: For this process they placed in each one of the vessels 25 pounds of German rosette copper broken in pieces, and they filled up the rest to within 2 dita of the rim with a powder of a mineral earth, yellowish in color and very heavy, that they called calamine. The rest of the empty space in the crucible they filled with powdered glass . . . Then they applied a melting fire for 24 hours and after this time they found the material entirely fused, and the 180 CHAPTER 9 copper, which was red before, had become a smooth and lovely yellow, almost like 24-carat gold in color. (Smith and Gnudi 1990:72) His approach to, and understanding of, the matter is clearly expressed in the introductory paragraph (Smith and Gnudi 1990:70): Since I have no knowledge other than that gained through my own eyes, I can tell you as a certainty that just as steel is iron converted by art into almost another kind of metal, so also brass is copper given a yellow color by art. Notably, he describes the furnaces used, and he also mentions that ‘little fitted clay shutters’ may be used on the crucibles (Smith and Gnudi 1990:72). Later he addresses the nature of this ‘earth that colors copper into brass’: I do not know that this earth serves any other purpose than coloring copper, because the mineral matter is of bad elemental mixture and poorly fixed . . . and not only does it color copper another color, but it increases its volume so much that the workman covers the cost of the copper and the expense of coloring it. (Smith and Gnudi 1990:75) Further, he elucidates: I believe that in its nature it is of a hot and dry quality like marcasite, as experience shows, because it does not melt alone by itself but burns and all its substance goes off in smoke. (Smith and Gnudi 1990:113) Soon after the publication of De la pirotechnia, in 1546, the first edition of Agricola’s De natura fossilium came out. Unlike his famous compendium De re metallica (Hoover and Hoover 1950), the former volume addressed brass production. This discussion was further expanded in the second edition, published in 1558 (text on brass translated and discussed in Martinón-Torres and Rehren 2002). Agricola starts his exposition by referring to Pliny and the old oreichalkos, before restating an old alchemical principle: naturam ars imitata (‘art imitates nature’) – it is thus possible to artificially dye copper. He then presents two alternative ways of making brass. The first one is again similar to Biringuccio’s, although he notes that the charge is best arranged in the crucible in layers, recaptures the possibility of adding crushed glass, and explicitly acknowledges that either mineral or furnace calamine may be used. He also insists on the need to carefully adjust the working temperatures. In the second method, the cementation principle remains the same, but he provides specific details such as the use of large pots, covered with THILO REHREN AND MARCOS MARTINÓN-TORRES 181 a perforated lid and externally coated with another clay layer – almost exactly the description of the archaeological crucibles from Zwickau. It is also worth noting that he quantifies the amounts of charge to be used and, particularly, the fact that the copper, ‘transformed into brass . . . has become much heavier’ (Martinón-Torres and Rehren 2002:102). A few decades later, Lazarus Ercker provides another very detailed account of brassmaking. He reports that the calamine is first roasted, then mixed with twice its volume of charcoal dust and dampened, with water, urine, or alum solution. After a blow-by-blow recipe of how to charge and fire the crucibles, he notes that the amount of copper charged in the crucibles increases in weight over the course of nine hours of firing, from 64 units to 90 units. Again, he judges the quality of the product by its ‘fine colour’, but all of his discussion centres on mundane issues such as the comparison between natural calamine from the ore deposits and artificial calamine collected from the furnaces near Goslar, and the lead content of the various types (Sisco and Smith 1951:255–56). His record is accurate and comprehensive, from the practice for the practice and seemingly with little regard for theoretical speculation and systematic ordering. ∗∗∗ All three accounts have much in common: for example, the use of colour as a measure for the quality of the metal produced, but also the recognition of the increase in mass – be it economically quantified by Biringuccio (‘covers the cost of the copper and the expense of coloring it’) or directly measured as weight gain by Agricola and Ercker. Some differences do occur, though these are more subtle than those identified above between Theophilus and Albertus. Agricola, on the one hand, was a foremost scholar of his time, with studies of philology, philosophy and science, and a degree in medicine. As a humanist, he wrote in Latin for a learned audience and filled his text with references to the classical authors. Biringuccio and Ercker, on the other hand, were primarily trained as practitioners – the former as a founder, the latter as an assayer – and wrote in their vernacular languages. However, they were also exposed to broader institutional and intellectual arenas: both of them spent their lives under the patronage of wealthy families and worked alongside scholars and scientists. Perhaps most remarkable is the fact that, by and large, these authors are providing clear and effective metallurgical recipes, but also some observations and thoughts on the nature of the elements involved and the process taking place. The emphasis certainly is on the practical side, but, as exemplified in Biringuccio’s lines on the calamine – and many others not reported here – they show an awareness of existing theoretical speculations, which they combine with their own observations – such 182 CHAPTER 9 as the increase in weight of brass. They are writing in the wake of the Renaissance: the world is being explored and conquered, information is published and discussed, knowledge about the natural world is keenly sought, and, increasingly, practical observations and theoretical abstractions are being linked. The boundaries between the scientist and the craftsman, between the alchemist and the metallurgist, between theory and practice, are more and more blurred (Beretta 1997; Long 1991, 2001; Martinón-Torres and Rehren 2005). This is epitomised by the ‘naturam ars imitata’, which inspires Agricola as much as it inspired Albertus and others. Of course, some are still concerned with perfecting the ‘permanent tincture’ for ‘brass, which is the Philosopher’s Gold, and that is true’ (att. Alfonso V King of Portugal 1651:69), but they all share a more direct connection among experiment, abstraction, and practice. To what extent did this Renaissance empiricism and rationality affect the evolution of brassmaking? Interestingly, Ercker not only describes brass production in close proximity to the smelting area of Goslar. In his Brief Report on the Rammelsberg, first published in 1565 (Beierlein 1968), he describes the occurrence of a strange white-metallic material that seeps out of some of the lead-smelting furnaces. This material, which he says is called Contrafeth, is most likely metallic zinc, condensing in cracks and crevices in the furnace walls. He suggests that much more of it could be made if one only spent some thought on it and laments that the workmen are not interested in new things, even if they would be useful (‘Such contrafeth could be made in quantity if effort and thought were put to it; but it is not valued, and the worker and smelter put no effort into new things, however much it would help’; translated in Beierlein 1968:252). He goes on to say that it is not regularly recovered and only ever collected when specifically requested by someone paying them a tip. Furthermore, he states that nothing can be made of this metal on its own, because it is too brittle, and proposes to alloy it with other metals to make it useable. Here we see, for the first time, the expression of the belief that a new material could be developed –through experiment and thought – into an economically interesting commodity: the beginning of industry-based research and development in the modern sense. It is also during this period that ingots of metallic zinc begin to arrive from China and India where, contrary to Europe, the distillation of zinc had been mastered centuries before. This puzzles scholars such as Andreas Libavius, who describe it as a ‘kind of tin’ (de Ruette 1995:195). Eventually, experimentation and active research led to ‘speltering’, the process in which metallic zinc is alloyed with copper to produce brass. This possibility seems to have been first realised by the 17th-century German chymist Johann Glauber, who wrote in 1657: THILO REHREN AND MARCOS MARTINÓN-TORRES 183 Zink is a volatile mineral, or a half ripe metal when it is drawn out of its ore. It is much clearer and brighter than tin, yet not so malleable and fluxile as tin is . . . We have it not much growing in Germany, but great quantity of it is every year brought us by the merchants out of the East Indies . . . It is a golden but an unripe mineral, it gives red copper a yellow colour and turns it into brass, as lapis calaminaris doth; and indeed that same stone is nothing else but unmeltable zink, and this zink may properly be called a susile lapis calamnaris (sic); for as much as both of them partake of one nature (translated in Packe 1689:319). The realisation was there. However, it would be another 200 years before the true nature of zinc as a metal was fully understood and its production sufficiently perfected to replace cementation as the basis of industrial brass production. Discussion It is tempting to see in the few examples of archaeological evidence for brassmaking briefly presented above a chronological trend line from small and primitive, unlidded vessels to successively larger and better closed vessels, representing an evolution of increasing productivity and improved procedures. Minor but significant design details, such as the wall thickness and the application of an outer wrap of less refractory clay, together with the better temperature adjustment in Zwickau, also point in this direction. They could well have been the result of semicontinuous minute changes in the practice, leading over 500 years to significant advances. This, however, is only the positivistic interpretation of the scenario. As seen from the three Roman examples mentioned above, it is as well possible that different crucible sizes coexisted (almost) simultaneously, possibly reflecting different economic settings, scales of production, or simply local technological styles rather than necessarily representing an evolutionary sequence. Only a much larger and better contextualised archaeological sample will shed clearer light on this issue. The textual evidence draws a more complex picture, as so often in this period. At a first glance, one sees a fault line running through the corpus of medieval texts addressing brass and brassmaking. On the one side stand scholars such as Albertus Magnus, who primarily try to understand fundamentally the material world using theoretical concepts inherited from the classical authors and further developed by their own thoughts. On the other side are writers more concerned with description rather than explanation, who carefully report their observations and convey recipes to facilitate the arts and crafts, rather than philosophy. In the Renaissance, the texts suggest a more systematic combination of practice and abstract thought that enabled a more practical understanding 184 CHAPTER 9 of the matter. This required, first, realising that the cementation process entailed a significant weight increase and not just a colour change. This observation is a major step forward from the old concept of a ‘power’ that colours copper an alien colour, shifting the balance from the ideas of hotness, inherent moisture, and so on to acknowledging that a physical substance is being added to the copper and effects this change. A second important realisation will be the fact that this extra substance is indeed zinc and that brass can therefore be made by simply alloying the two metals. Much needs to be said about the transmission of knowledge, both over time and between the two sides – practitioners and scholars. It would appear that the latter owe the former a lot of careful observation and experimentation; however, practitioners such as Biringuccio and Agricola explicitly credit ‘the alchemists’ with the invention of brassmaking. Whether this is a reference to a possible ‘Eastern’ origin of the revived medieval brassmaking or simply a generic attribution for everything enigmatic and artful is another question outside the remit of this chapter. Whatever the case, the historical record shows that, in theory, by the 17th century everything was ready for a Renaissance breakthrough in brassmaking that could have turned it into an even more efficient and profitable industry – hence Ercker’s encouragement of research and development. What happened in practice? Did these seeds of innovation grow? To cut the story short, it is enough to remember that a method for the industrial production of zinc by distillation was first patented only in the late 1730s (by William Champion in Bristol, England), and still this proved rudimentary and unprofitable, and in need of much development, before it could allow speltering as an established brassmaking method (Day 1973, 1991, 1995, 1998; Dungworth and White 2007). Not only for Britain but also for Bavaria (Priesner 2000) we have a rich historical and archaeological record evidencing that the industrial production of brass continued to rely on the traditional cementation method well into the 19th century, that is, over 200 years after the necessary discoveries for direct alloying had taken place. There is not enough space to elaborate on the reasons why this Renaissance wave of experimentation and discovery took so long to affect daily, industrial practice, that is, what we would now term the relationships between science and technology. In considering possible explanations, we are inclined to agree with Johann Glauber: in spite of his excitement about, and encouragement for, the exploitation of the new metal zinc, ‘which is most excellently excellent (sic)’, he was frustrated that ‘men are hardly drawn back from an old custom’ (Packe 1689:320). At a practical level, tradition was clearly stronger than the stimuli for innovation, hence the slow pace of industrial applied research. Surely, additional reasons THILO REHREN AND MARCOS MARTINÓN-TORRES 185 for this lie in the particular socioeconomic contexts, the balance between the perceived costs and benefits, competition between companies, and the flow of information between scholars and practitioners. As this chapter has tried to show, the full picture will be reconstructed only through combined work in archives and libraries, archaeological sites, and archaeometric laboratories. Acknowledgments This chapter grew out of many years of cooperation with many colleagues, allowing access to excavated finds as well as to their thoughts. We can mention only some of them, including N. Zieling, H. Brink-Kloke, and J. Beutmann for their archaeological expertise, and Chr. Bartels and M. Charlton for their thoughts and comments on ideas developed in this text. Any errors of judgment remain ours. The underlying analytical work, published elsewhere, was done first in the Institut für Archäometallurgie at the Deutsches Bergbau-Museum, Bochum, and later at the Wolfson Archaeological Science Laboratories at the UCL Institute of Archaeology, London. We are indebted to these institutions and their staff, particularly Andreas Ludwig and Kevin Reeves, for their unfailing support. We acknowledge with gratitude financial support given to MMT by the Barrié de la Maza Foundation (Spain) and the British Council (UK). References Alfonso V, King of Portugal. 1651. Five Treatises of the Philosopher’s Stone: Two of Alphonso King of Portugall, as It Was Written with His Own Hand, and Taken out of His Closset . . . London: Thomas Harper. Bayley, J. 1998. ‘The production of brass in Antiquity with particular reference to Roman Britain’, in P. T. Craddock (ed.), 2000 Years of Zinc and Brass. London: British Museum Occasional Paper 50, The British Museum, 7–26. Bayley, J., and S. Butcher. 2004. Roman Brooches in Britain: A Technological and Typological Study Based on the Richborough Collection. London: The Society of Antiquaries of London. Beierlein, P. R. 1968. Lazarus Ercker: Drei Schriften. Bochum: VFKK. Beretta, M. 1997. ‘Humanism and chemistry: The spread of Georgius Agricola’s metallurgical writings Nuncius: Annali di Storia della Scienza XII:17–47. Cameron, H. K. 1974. ‘Technical aspects of medieval monumental brasses’, Archaeological Journal 131:215–36. Caple, C. 1995. ‘Factors in the production of medieval and post-medieval brass pins’, in D. R. Hook and D. R. M. Gaimster (eds.), Trade and Discovery: The Scientific Study of Artefacts from Post-Medieval Europe and Beyond. British Museum Occasional Paper 109. London: The British Museum Department of Scientific Research, 221–34. Craddock, P. T., and K. Eckstein. 2003. ‘Production of brass in Antiquity by direct reduction’, in P. T. Craddock and J. Lang (eds.), Mining and Metal Production Through the Ages. London: The British Museum Press, 216–30. 186 CHAPTER 9 Day, J. 1973. Bristol Brass: A History of the Industry. Newton Abbot: David and Charles. ———. ‘Copper, zinc and brass production’, in J. Day and R. F. Tylecote (eds.), The Industrial Revolution in Metals. London: The Institute of Metals, 131–99. ———. 1995. ‘Trade and innovation in non-ferrous metals at Bristol’, in D. R. Hook and D. R. M. Gaimster (eds.), Trade and Discovery: The Scientific Study of Artefacts from Post-Medieval Europe and Beyond, British Museum Occasional Paper 109. London: The British Museum Department of Scientific Research, 205–20. ———. 1998. ‘Brass and zinc in Europe from the Middle Ages until the nineteenth century’, in P. T. Craddock (ed.), 2000 Years of Zinc and Brass, British Museum Occasional Paper 50. London: The British Museum, 133–58. de Ruette, M. 1995. ‘From conterfei and speauter to zinc: The development of the understanding of the nature of zinc and brass in post-medieval Europe’, in D. R. Hook and D. R. M. Gaimster (eds.), Trade and Discovery: The Scientific Study of Artefacts from Post-Medieval Europe and Beyond, British Museum Occasional Paper 109. London: The British Museum Department of Scientific Research, 195–203. ———. ‘Brass foundry workshops of the Southern Low Countries and the Principality of Liége: A technical approach’, Historical Metallurgy 30:116–20. Dungworth, D. 1996. ‘Caley’s “zinc decline” reconsidered’, Numismatic Chronicle 156:228–34. ———. 1997a. ‘Iron Age and Roman copper alloys from northern Britain’, Internet Archaeology 2, http://intarch.ac.uk/journal/issue2/dungworth/index.html. ———. 1997b. ‘Roman copper alloys: Analysis of artefacts from northern Britain’, Journal of Archaeological Science 24:901–10. Dungworth, D., and H. White. 2007. ‘Scientific examination of zinc-distillation remains from Warmley, Bristol’, Historical Metallurgy 41:77–83. Eremin, K., J. Graham-Campbell, and P. Wilthew. 2002. ‘Analysis of copper-alloy artefacts from pagan Norse graves’, in K. T. Biró and E. Jerem (eds.), Archaeometry 98: Proceedings of the 31st Symposium, Budapest, April 26–May 3 1998, 343–49. Fellmann, R. 1999. ‘Das Zinktäfelchen vom Thormebodewald auf der Engehalbinsel bei Bern und seine keltische Inschrift’, Archaeologie im Kanton Bern 4B:133–75. Hachenberg, K. F. 2004. ‘Der Werkstoff Messing im Musikinstrumentenbau vom 16. bis zum Ende des 18. Jahrhunderts’, in B. E. H. Schmuhl and M. Lustig (eds.), Jagdund Waldhörner: Geschichte und musikalische Nutzung. 25. MusikinstrumentenbauSymposium, Michaelstein, 8. bis 10. Oktober 2004, Michaelsteiner Konferenzberichte 70. Augsburg: Wißner. Hachenberg, K. F. 2006. ‘Die analytische Beschaffenheit des Nürnberger Messinggusses von 1450 bis um 1720’, Guss im Wandel der Zeit 39:11–9. Haedecke, K. 1973. ‘Gleichgewichtsverhältnisse bei der Messingherstellung nach dem Galmeiverfahren’, Erzmetall 26:229–33. Hawthorne, J. G., and C. S. Smith. 1979 [1963]. Theophilus: On Divers Arts. The Foremost Medieval Treatise on Painting, Glassmaking and Metalwork. New York: Dover. Hoover, H. C., and H. L. Hoover. 1950 [1912]. Georgius Agricola: De Re Metallica. Translated from the First Latin Edition of 1556. New York: Dover. Kibre, P. 1944. ‘An alchemical tract attributed to Albertus Magnus’, Isis 35:303–16. Long, P. O. 1991. ‘The openness of knowledge: An ideal and its context in 16thcentury writings on mining and metallurgy’, Technology and Culture: The International Quarterly of the Society for The History of Technology 32:318–55. ———. Openness, Secrecy, Authorship: Technical Arts and the Culture of Knowledge from Antiquity to the Renaissance. Baltimore: Johns Hopkins University Press. THILO REHREN AND MARCOS MARTINÓN-TORRES 187 Martinón-Torres, M. 2001. The Technology of Brass Production in Central Europe from the 10th to the 16th Century: Archaeometry and History, Dissertation for MSc in Technology and Analysis of Archaeological Materials. London: Institute of Archaeology, University College London. Martinón-Torres, M., and Rehren, T. 2002. ‘Agricola and Zwickau: Theory and practice of Renaissance brass production in SE Germany’, Historical Metallurgy 36:95–111. ———. 2005. ‘Alchemy, chemistry and metallurgy in Renaissance Europe: A wider context for fire assay remains’, Historical Metallurgy 39:14–31. Mitchiner, M. B., C. Mortimer, and A. M. Pollard. 1986. ‘Nuremberg and its jetons, c.1475 to 1888: Chemical compositions of the alloys’, Numismatic Chronicle 147:155. Newbury, B. D., M. R. Notis, and D. E. Newbury. 2005. ‘Revisiting the zinc composition limit of cementation brass’, Historical Metallurgy 39:75–81. Packe, C. 1689. The Works of the Highly Experienced and Famous Chymist, John Rudolph Glauber, Containing Great Variety of Choice Secrets in Medicine and Alchymy, in the Working of Metallick Mines, and the Separation of Metals . . . London: Thomas Milbourn. Picon, M. M. Le Nezet-Celestin, and A. Desbat. 1995. ‘Un type particulier de grands récipients en terre réfractaire utilisés pour la fabrication du laiton par cémentation’, in Actes du Congrès de Rouen de la Société Francaise d’Étude de la Céramique Antique en Gaule. Marseille: Société Francaise d’Étude de la Céramique Antique en Gaule, 207–15. Pollard, A. M., and C. Heron. 1996. Archaeological Chemistry. Cambridge: The Royal Society of Chemistry. Ponting, M. J. 2002. ‘Roman military copper-alloy artefacts from Israel: Questions of organization and ethnicity’, Archaeometry 44:555–71. Ponting, M. J., and I. Segal. 1998. ‘Inductively coupled plasma-atomic emission spectroscopy analyses of Roman military artefacts from the excavations at Masada, Israel’, Archaeometry 40:109–22. Priesner, C. 2000. ‘Ein Compositum von Natur und Kunst’: Zur Technologie der Messingfabrikation im 18. Jahrhundert’, Der Anschnitt 2000:130–41. Rackham, H. 1952. Pliny, Natural History. London: Heinemann. Rehren, T. 1996. ‘A Roman zinc tablet from Bern, Switzerland: Reconstruction of the manufacture’, in S. Demirci, A. M. Özer, and G. D. Summers (eds.), Archaeometry 1994. Ankara: Symposium on Archaeometry, 35–45. ———. 1999a. ‘Small size, large scale: Roman brass production in Germania Inferior’, Journal of Archaeological Science 26:1083–87. ———. 1999b. ‘The same . . . but different: A juxtaposition of Roman and Medieval brass making in Central Europe’, in S. M. M. Young, A. M. Pollard, P. Budd, and R. A. Ixer (eds.), Metals in Antiquity, BAR International Series 792). Oxford: Archaeopress, 252–57. ———. 2003. ‘Crucibles as reaction vessels in ancient metallurgy’, in P. T. Craddock and J. Lang (eds.), Mining and Metal Production through the Ages. London: The British Museum Press, 207–15. Rehren, T, E. Lietz, A. Hauptmann, and K. H. Deutmann. 1993. ‘Schlacken und Tiegel aus dem Adlerturm in Dortmund: Zeugen einer mittelalterlichen Messingproduktion’, in H. Steuer and U. Zimmermann (eds.), Montanarchäologie in Europa: Berichte zum Internationalen Kolloquium ‘Frühe Erzgewinnung und Verhüttung in Europa’ in Freiburg im Breisgau vom 4 bis 7 Oktober 1990. Sigmaringen: Jan Thorbecke Verlag, 303–14. 188 CHAPTER 9 Riederer, J. 1988. ‘Metallanalysen Augsburger Bronze- und Messingskulpturen des 16. Jahrhunderts’, Berliner Beiträge zur Archäometrie 10:85–95. Sisco, A. G., and C. S. Smith. 1951. Lazarus Ercker’s Treatise on Ores and Assaying, Translated from the German Edition of 1580. Chicago: The University of Chicago Press. Smith, C. S., and M. T. Gnudi. 1990 [1959]. The Pirotechnia of Vannoccio Biringuccio: The Classic Sixteenth-Century Treatise on Metals and Metallurgy. New York: Dover. Thornton, C. P., and C. B. Ehlers. 2003. ‘Early brass in the ancient Near East’, iams 23:3–8. Ullwer, H. 2001. ‘Messingherstellung nach dem altem Galmeiverfahren’, Erzmetall 54:319–26. Weeks, L. 2004. ‘An analysis of Late Pre-Islamic copper-base artefacts from Ed Dur, U.A.E.’, Arabian Archaeology and Epigraphy 15:240–52. Welter, J.-M. 2003. ‘The zinc content of brass: A chronological indicator?’, Techne: La science aus service de l’historie de l’art et des civilisations 18:27–36. Werner, O. 1970. ‘Über das Vorkommen von Zink und Messing im Altertum und im Mittelalter’, Erzmetall 23:259–69. Wyckoff, D. 1967. Albertus Magnus: Book of Minerals. Oxford: Clarendon. Zacharias, S. 1989. ‘Brass making in medieval western Europe’, in M. L. Wayman (ed.), All That Glitters: Readings in Historical Metallurgy. Montreal: The Metallurgical Society of the Canadian Institute of Mining and Metallurgy, 35–40. Zwicker, U. H. Greiner, K.-H. Hofmann, and M. Reithinger. 1985. ‘Smelting, refining and alloying of copper and copper alloys in crucible furnaces during prehistoric up to Roman times’, in P. T. Craddock and M. J. Hughes (eds.), Furnaces and Smelting Technology in Antiquity, British Museum Occasional Paper 48. London: The British Museum, 103–15.