The history of brass is complex, and it defies popular ideas about metallurgy – assuming that alloys are made by melting metals together. Metallic zinc was little known in Europe until the early modern period, but its alloy with copper – brass – was in high demand in antiquity and the middle ages. Recipes typically call for powdered minerals or furnace accretions that (unknown to the ancients) contained zinc, and use a “cementation” process to combine the metals – very different to melting the metals together. Instead, solid or molten copper is exposed to zinc vapour inside a sealed vessel, heated to a high temperature. Various names are used for the ingredients of these recipes, and where I’ve struggled to understand them, I have referred to book five of Dioscoridesii, which was written much earlier in history, but provides a useful glossary.
Mappae Clavicula is a large collection of recipes from the middle ages, and earlier. The translation that I use comes from a 12thC manuscript, but many of the recipes were several hundred years old when that was written, and some even older (similar recipes appear in the Leyden/Stockholmiii papyrii, written in the 3rd century AD). As well as a shift in time, there is a shift in location, and the recipes of Egypt are joined by northern European runes, and even a few hints of Anglo-Saxon ingredients. In short, it’s a mess. There is some very technical information in this document, but it has to be teased out from amongst the repetitions and transcription errors of the scribes who collated and cared for the Mappae manuscripts over hundreds of years. If we want an unambiguous brass recipe, there is only one:
74. [Making brass]: “Take ductile copper of the kind which is called caldarium, or fired copper, and hammer it out and make sheets of it. Spread these sheets out on the bottom [of a pot] and sprinkle on them thoroughly ground white cadmia. This is engendered in Dalmatia, where coppersmiths use it. Lute the furnace thoroughly with potter’s clay so that it cannot breath for a day. Then open up and if it is good, use it; if not, cook it with cadmia a second time, as above. If it comes out well, the caldarium copper is permeated throughout with gold [color].”
There’s plenty of information to unpack in this recipe. First, the type of copper is called caldarium – broadly speaking, a copper used for making cauldrons, which could be refined copper, or could be a range of alloys containing tin, and other metals. It is specified as ductile, emphasised by the instruction to hammer it out, so it probably doesn’t contain any lead. What we can be sure of is that it doesn’t contain much zinc at the beginning of the description – the purpose of this recipe is getting zinc into the alloy, which will enhance it’s colour, and make the alloy less prone to soaking up oxygen when molten (a problem with pure copper, as it has an affinity for oxygen).
The caldarium is hammered out to increase the available surface area of the metal, which speeds up the process considerably. The sheets are spread out “on the bottom [of a pot]”, and sprinkled with “white cadmia”. In this context, cadmia is clearly a zinc-bearing compound, but not all descriptions of cadmia are so reliable, as we shall see later. There is then the instruction to “lute the furnace thoroughly”, which is a little confusing. Typically, you would would expect the copper and cadmia to be placed into a vessel, which is sealed with lute (perhaps a clay slip that fills gaps between the vessel and it’s lid, making it airtight). There may be some confusion here over what is being sealed – the vessel or the furnace. It’s worth noting that there is no discussion of fuel in this recipe. It is usual to put charcoal or some other organic material inside the sealed vessel, to create a reducing atmosphere, and more fuel around the vessel, so that it can be heated to an appropriate temperature.
After a day, the furnace is opened up, and the metal is checked. Ideally, there will be a golden coloured alloy inside, either in sheet form, or as a melted mass, depending on the temperature that the furnace reached. If the colour is not good enough, the process begins again. This recipe calls for a second attempt, but some of the other texts (eg. Theophilusiv) call for three or four. It’s hard to tell if that’s a matter of efficiency, or expectation. Zinc content can approach 30% in archaeological alloys, but it can be much lower. When colour is the main requirement, a high percentage is needed, but even a tiny amount of zinc will inhibit the oxidation of molten copper. Although #74 is the only recipe for brass production, Mappae suggests a more complex role for brass and zinc in the medieval workshop,
For example, #18. According to Dioscorides, armenium is similar to chrysocolla, which is usually a copper compound, and the translators suggest copper carbonate. Zonitodos is less clear. Dioscorides does use the name zonitos for a zinc compound, and in any case, cadmia is mentioned here in its own right. This could, potentially, be a sort of smelting process, intended to reduce the compounds to a metallic state. The reference to copper is more suggestive of alloy production.
18. Making proven gold: “Two parts of armenium, 1 part of zonitidos. Grind them all, add a forth part of bull’s dung and an equal part of cadmia. Melt it and it will be rather heavy. Do the same thing also in copper.”
Cadmia itself isn’t that reliable as a term, unfortunately. #125 definately isn’t a recipe for brass, but the description of cadmia is clear – “from the upper part of the furnace”, which is consistent with accounts of artificially produced zinc compounds. However, it exposes the confusing nature of cadmia, which could simply be a name for anything scraped from the inside of the furnace. Lead compounds seem likely in this case, and Dioscorides mentions lead compounds as an alternative for the zinc oxide that he calls spodium.
125. Silver ore: “Prasinus is a green earth, from which ore silver flows. And this earth is engendered in rocky places where many ores of different colors are found. This rock when crushed has white veins, and when they are cooked, black stuff will emerge. It is tested like this. When it is broken up after it has been cooked, it shows colors as of silver insider; this stone is the one from which silver will emerge. Put this ore with cadmia from the upper part of the furnace and fill up with coals; and so, stratum super stratum with wood and charcoal on top, smelt it as I said above. Melt it for a day and let it cool in the same place. Next take up the lump and break it into tiny pieces, and put it back in the same furnace as before and with it [some] feminine lead: for a hundred-pound lump, 15 [pounds] of lead. Cook as before for 3 days. After this throw the lump out and break it up; put it in a furnace and melt for 2 hours.”
Another reference to cadmia appears in #9, which is a recipe for gold, apparently, despite describing 2 parts of copper for every one of gold. Zinc would improve the colour of the alloy (which would otherwise be very red). Here, it says “…cadmia or Trachian stone – it is yellow and laminar, ie. flat or milky…”), which seems to describe a mineral, with cadmia either as an alternative to it, or a synonym for it. I’ve been experimenting with modern 9ct red gold alloys recently (37.5% gold), and even a tiny quantity (<1%) of zinc improves the working properties of the resulting alloy, as well as shifting the colour towards yellow. The recipe follows on from #8, “A recipe for gold”.
9. Again: “Take as much gold as you want and twice as much imported misy and the same amount as the misy of thick pyriapian filings of good copper or Cyprian copper that has been melted. Mix them both together and make a gold tube into which you deposit the three preparations. And so melt it, cooling as necessary, and take it out of the furnace and wash it off. You will then find the greatest weight of gold; and when this has touched the fire, it will become better. Mix in some cadmia or Trachian stone – it is yellow and laminar, ie., flat or milky – until it appears to you perfect gold.”
A recipe with similar instructions is #28, but this time it seems to be enriching an existing brass alloy. However, without knowing the identity of demnas, which is the largest individual ingredient in the recipe, it is hard to be sure. With a mention of both brass and cadmia, zinc seem certain. This recipe follows on from #26, “A doubling of gold”.
28. Another way: “Prepare 1 part of prime brass filings, so that they can easily be melted, 8 minas of Samian cadmia, 8 parts of roasted misy, 12 of demnas. Melt carefully with this mixture.”
It might seem odd to talk about brass in connection with gold, but in many of these recipes, the notion of fakery is important. Mappae descends, in part, from the Leyden and Stockholm papyrii, which each provide a forger’s handbook, with exhortations to keep the secrets, and deceive the artisans who will buy the resulting alloys, which are called asem in the papyrii, though that word does not appear in Mappae. Thorough discussion of Leyden and Stockholm will have to wait for another article, but in brief, asem refers to gold, silver, electrum, and anything that looks like them. Passing various inspections, including the experienced eye of the artisan, was important. This reference to “proven gold” has the same character, and it is here that an alchemical link becomes apparent, because there do seem to be some recipes that turn base metals into precious metals. OK, so the “golden” colour of brass is the colour of a low-carat alloy, rather than the rich, deep (and almost nauseating) yellow of fine gold, but it’s enough to pass muster for most people.
Despite this supposed alchemical power of zinc, is every copper-to-gold recipe about brass? The practice of changing colour is an important aspect of medieval metalwork, whether providing a metallic layer (eg. Gilding and silvering), or faking it. As a metal, copper has a unique red colour, which can be modified by alloying it with other metals. Zinc, as we know, provides a golden colour. Tin gives a brown to white, depending on the copper-tin ratio. There is also arsenic (usually in the form of the compound orpiment in Mappae Clavicula) which produces a white alloy with copper.
In practice, ancient alloys tend to be mixtures of more than two metals, even if the trace elements are ignored. The characteristics of the resulting materials can vary considerably, and the artisan might need to change not just the colour, but also the working properties of the metal – how it melts, how it flows, how strong it is when cooled, and whether it can be forged or not. Various stages of refining the alloy or adding other metals could take place before the desired result was achieved (pictured is one of the experimental ceramics that Graham Taylor made for me, which are being used to recreate some of these operations).
In conclusion, brass and zinc compounds are regularly mentioned in Mappae Clavicula, but there is a only a single recipe for producing copper-zinc alloys. This is suggestive of a context in which brass production is known, and the resulting alloys and compounds are used as additives for later metallurgical operations. That said, the text reveals little depth, despite its origin in the eastern Mediterranean region, where brass was produced in quantity during antiquity. The instructions are garbled, and most of the recipes could be served by local markets rather than integrated metallurgical practice.
i Smith and Hawthorne, Mappae Clavicula – A Little Key to the World of Medieval Techniques.
ii Dioscorides, Osbaldeston, and Wood, Dioscorides – De Materia Medica.
iii Caley and Jensen, The Leyden and Stockholm Papyrii.
iv Presbyter, Smith, and Hawthorne, On Divers Arts.