Transcript
23 Paint it red: vermilion manufacture in the Middle Ages Stefanos Kroustallis and Rocio Bruquetas ABSTRACT Natural cinnabar and artificial vermilion were the first opaque bright red pigments available to artists. In recent years a growing interest in their manufacture and their use in painting techniques has given a new impulse to the debate about the invention and manufacture of artificial vermilion, both in the field of art technology and in the transmission of technical and scientific knowledge. Tis paper highlights and explains some particular technical aspects of the ‘dry method’ for manufacturing vermilion in the Middle Ages in Europe. It also describes an unusual Spanish ‘dry method’. Tis study is based on treatises including recipes, contemporaneous documentary information and later printed scientific works (8th–17th century). The invention of vermilion Initially, artificial cinnabar or vermilion 1 was prepared by heating together mercury and sulphur (the so-called ‘dry method’). During the last quarter of the 17th century, a ‘wet method’ was introduced that consisted of heating the aethi-ops mineral (metacinnabar 2 ) in a solution of a strong alkali. From that time, the wet method overtook the dry method in popularity because it considerably reduced the labour, time and cost of vermilion manufacture. Moreover, it produced a fine, uniform grain size of red pigment (Gettens et al . 1993; Ball 2008: 293).he idea that the procedure for making vermilion with the dry method was brought into Europe by the Arabs is quite widespread. his is because the reaction of sulphur with mercury to produce a red compound (Wisniak 2008: 148) had already been described by early Arab writers such as Geber ( c .721–815 CE) or Rhazes (865–925 CE). Nevertheless, the work of both these authors was not known in Europe before the 12th century, and the first translations in Latin date from that period. In Western Europe, the first known recipe describing the production of vermilion by heating mercury with sulphur was consigned to an art technological treatise, the Compositiones ad tingenda (Lucca, Biblioteca Capitolare, Cod. 490, f. 223v) dated from the end of the 8th century. Such recipes were also described within the earliest examples of the treatises of the Mappae clavicula tradition. hus, as stated by Wallert (1990: 155), the technical process for the preparation of vermilion was already known in Europe before the dissemination of Arabic scientific works. Moreover, an undated letter from Frotharius, bishop of oul ( c .813–847), to Aglemar, archbishop of Sens, demonstrates that this knowledge was neither isolated nor a mere literary reference, but common craft practice. In this letter, Frotharius requests some pigments for the decoration of the walls of the new cathedral and notably asks for mercury (McKitterick 1995: 146). he context is unambiguous as the only pigment obtained from mercury is vermilion. his leads to the asser-tion that the making of vermilion was known and practised in early medieval Europe.he fact that the manufacture of vermilion is not a development from Arabic science and technology also makes it impossible to relate to the alchemical mercury-sulphur theory of the generation of metals, proposed by Geber and disseminated in Europe by later Arabic and Latin alchemical texts. As the mercury-sulphur theory appears to have been unknown to Greek alchemists, the making of vermilion also predates alchemical transmutation processes. As Smith stated (2010: 41): ‘one of the most pervasive and enduring theories of matter and its transformation in metallurgy – the sulphur mercury theory of metals – may have emerged from the making of materials’. he idea that the invention of vermilion relates to alchemical theory is due to the perception that both pigment making and artisanal practice lack any theoretical background. hus alchemy was the closest possibility for ascribing a common knowledge – the theory of workshop practice – to art technology. Vermilion was not part of the ancient Greco-Roman paint-er’s palette, neither was it used in ancient medicine as a remedy, unlike mineral cinnabar. Mercury had been known in Indian and Chinese medicine and alchemy since antiquity, but it was not until the 3rd–4th century CE that mercury and sulphur were mixed to obtain a red compound (Needham 1976: 74, 78–9; Liu et al . 2008: 810–12; Wisniak 2008: 148). In the same period, the Alexandrian alchemist Zosimos of Panopolis (active at the end of the 3rd century to the beginning of the 4th century CE) described in a difficult-to-understand text the reaction of sulphur with mercury for producing a red substance. Chinese MTA-04-Kroustallis-v2.indd 2308/09/2014 12:57 24 STEFANOS KROUSTALLIS AND ROCIO BRUQUETAS Table 1 Vermilion manufacture recipes (8th–17th century) with references to stoichiometry, type of vessel and temperature details. CenturyManuscriptStoichiometry Hg:S Vesselemperature8th Compositiones ad tingenda (Caffaro 2003)2:1 (parts)9th–10th Mappae clavicula Sélestat MS (Smith and Hawthorne 1974)2:1 (parts)2:1 (parts) + urine Moderate fire Glassworker furnace11th–12th De diversis artibus (Dodwell 1961)2:1 (parts)Glass jar covered with clay Live coals12th De mixtionibus (Smith and Hawthorne 1974)1:2 (parts)Glass jar covered with clay Moderate fire12th–13th Liber sacerdotum (Berthelot 1963)2:1 (parts)Moderate fire13th Libellus de alchimia (Heines 1958)2:1 (parts)Glass vessel with narrow neckLow fire for half a day and strong fire for another day 13th–14th De coloribus naturalia excripta et colecta (Tompson 1935)2:1 (parts)Glass bottle covered with clay Live coals14th Il aquino Antonelli (orresi 1993)2:1 (parts)Also 1:2 or 1:1 recipesGlass jarLive coals14th Secretum philosophorum (Clarke 2009)1:2 (parts)Glass jar covered with clay Moderate fire14th De coloribus illuminatorum sive pictorum (Tompson 1926)1:2 (parts)Glass jarModerate fire14th Fondo Palatino della BNCF (Pomaro 1991)100 pounds: 33 pounds and 4 ounces.Glass jarLow fire first;then strong fire for 20–24 hours15th Libro de como se fazem as cores (Melo and Miguel 2012)5 pounds:1 poundGlazed potModerate fire15th Liber diversarum arcium (Clarke 2011)2:1 also 1:2 recipeGlass jar covered with clay Moderate fire 15th ractatus de coloribus (Edgerton 1963)1: ½ or 1:1/3 2:1Glazed ceramic pot;or sublimation vase; or glass jarGlass jar covered with clay Low fire and then strong fire15thMS. of the Newberry Library (Bommarito 1985)2:1 15th De coloribus (Torndike 1960)2:1Glass jar covered with clay Low fire and then strong fire15th Libro secondo di diversi colori (Wallert 1995)1:1Glass jarModerate fire;tile cover15th De coloribus faciendis (Van Acker 1972)1:2Glass jarLow fire15th Recipes from Fra Bartolomeo di Siena (ossati Soldano 1978)1:2Glass jarLow fire15th Incipit liber de diversarum operum nobilissimorum (Malaguzzi Valeri 1896)1 pound:2 poundsGlass jar with narrow neck covered with clay 3 hours low fire3 hours moderate fire3 h strong fire15th Fondo Palatino della BNCF (Pomaro 1991)8:2Glazed fire3 times in fire2:1 (parts)1 pound:1 pound and 4 ounces + tin1 pound:3 pounds1:2 (parts)Strong fireModerate fireModerate fireCharcoal fire 15thBolognese MS (Merrifield 1967)1:2 (parts)½ pound:4 parts + leadGlass jar covered withclay 14 hours15th De colorum diversitate tractaturi (Torndike 1960)1: 2 (parts)1:2 (parts) + 1/7 salt1:1Glass jarGlass jarGlass jarModerate fireLow fire16th Fondo Palatino della BNCF (Pomaro 1991)1:2 (parts)Glass jar16th De’secreti (Piamontese 1561)9:2 (parts)16th Arte de los metales (Barba 1640) 2:1 (parts)17th Fondo Palatino della BNCF (Pomaro 1991)3 pounds:8 ouncesSublimate MTA-04-Kroustallis-v2.indd 2408/09/2014 12:57 25 PAINT IT RED: VERMILION MANUFACTURE IN THE MIDDLE AGES and Hellenistic texts that quote the mixture of mercury and sulphur to give a red substance are nearly contemporary, there-fore vermilion making could not have srcinated in Chinese alchemy, then been passed on to Hellenistic alchemy, before being carried by the Arabs to the West (Gettens et al . 1993: 47). Greek alchemy was strongly influenced by Greek philosophy, Gnostic ideas and technical traditions in Mesopotamia and Egypt. Although Greek alchemy shares some principles with Chinese and Indian alchemy (Levere 2001: 3–6) it seems that each was developed independently and with different purposes.his paper argues that the manufacture of vermilion was invented independently in the late Roman Empire. It is quite probable that a direct consequence of the exploitation of the Almadén mines in Spain was the commercialisation and availability of mercury in the Mediterranean world. In addition, the development of Greek alchemy in Alexandria (which relied on a strong technological base of applied crafts and sciences involving chemistry such as metallurgy, the manufacture of glass, dyes, pigments, cosmetics and drugs) provided the necessary technological (and even theoretical) background for the dissemination of the artificial vermilion-making process. As mentioned above, Zosimos, one of the principal Alexandrian practical alchemists, was skilled in laboratory manipulations and the first in the West to mention the reaction of mercury with sulphur to obtain a red substance. he role of Byzantium was also very important in the formation and dissemination of a Mediterranean science. Around the 7th century, a revitalised interest in alchemical writings in Byzantium led to the compilation and organisa-tion of the so-called corpus of Greek Alchemists, structured in texts that were theoretical and dogmatic, and technical recipes. he latter were extensively reworked by compil-ers, interested either in alchemical practice (thus preparing alchemical treatises) or in art and craft practice (elaborating art technology treatises). hese art and craft manuals were at the root of the transmission of technical knowledge to the West and the dissemination of collections of instructions or recipes, such as the 8th-century Compositiones ad tingenda , characterised by a strong Greek background, which contains the first recipe for the making of vermilion. The manufacture of vermilion in the Middle Ages After the Compositiones ad tingenda vermilion recipe, tech-nical instruction appeared more frequently in treatises on art technology, notably in the manuscripts of the Mappae clavicula tradition (9th–12th century) or in the treatise by heophilus (11th–12th century). Except within art techno-logical texts, 3 no detailed technical description of the making of vermilion was known before the 13th century, although vague references to the process might be found in encyclo-pedic, medical or alchemical works. From the 14th century and onwards, recipes appear more frequently and become common knowledge in artistic, alchemical or pharmaceutical literature. For example, both Cennini in his Il libro dell´arte and the anonymous author of De arte illuminandi stated that vermilion was readily available commercially (hompson 1960: 24; Brunello 1992: 49). Cennini also stated that anyone who decided to ‘take the trouble’ to prepare vermilion would waste his time over the many variations of procedure (hompson 1960: 24). However, at that time the only process known for the making of vermilion was the dry method. Cennini probably referred to problems linked to the use of different proportions of ingredients and to the heating time. Indeed this is still a thorny question since technical descriptions were not stand-ardised in art technological medieval treatises, and there are significant variations not only in stoichiometry but also in the way reactants were employed and in the heating methods.Some technical aspects of vermilion making will be explained below in order to clarify historical references towards further analysis or reconstruction attempts. he main textual sources are medieval treatises on art techno-logy, but also include alchemical and medical treatises and documentary sources (8th–15th century). Later sources have also been considered (16th–19th century), as they comprise the technical material necessary for understanding earlier recipes. 4 Proportions A proportional system of measurements for ingredients (parts) is commonly used in the technical recipes in order to standardise and facilitate the transmission of the technical process described over time and through different work-shops. his system was also employed in the oldest recipes for artificial vermilion (able 1). he comparison of vermilion recipes in art technological treatises reveals that, at least until the 14th century, two types of mass (or weight) proportions were mainly used. 5 In the oldest recipes (8th–12th century) the ratio of 2:1 for mercury to sulphur is specified. his pro-portion can be found in the principal treatises such as the Compositiones ad tingenda , the Mappae clavicula family, the De diversis artibus of heophilu s , and the Montpellier Manuscript. he second ratio is exactly the opposite 1:2 (Hg:S). his ratio was first suggested in the 11th- or 12th-century treatise De coloribus et mixtionibus that preceded the Mappae clavicula of the Phillipps-Corning Manuscript (Smith and Hawthorne 1974: 26), and appeared frequently in later treatises until the 15th century (able 1).From the 14th century on the situation began to change, and references to specific weights became more frequent. When weights were mentioned, ratios ranged from 1:3 (Hg:S) in a recipe from the 14th-century Codex Pal . 949 (Pomaro 1991: 95–6) to 5:1 (Hg:S) in the recipe from the 15th-century he Book on How to Make Colours (Melo and Miguel 2012: 184–5).As Melo and Miguel stated (2012: 183), the reaction stoichiometry in order to obtain vermilion as a good yield is 1 part of sulphur to 6.2 parts of mercury by weight. 6 his ratio corresponds grosso modo to the proportions found in the natural reaction for the formation of cinnabar as mineral. Apparently, most medieval recipes would never have obtained good yields of vermilion as the quantity of mercury was far lower than required. Of course we need to keep in mind that vermilion and sulphur ratios are always relative data since no medieval recipe states the amount of product obtained MTA-04-Kroustallis-v2.indd 2508/09/2014 12:57 26 STEFANOS KROUSTALLIS AND ROCIO BRUQUETAS in the reaction and the possible losses. herefore, a later, relevant, historical source is used here in order to address this issue: the so-called Pekstok papers. Willem Pekstok was a well-known 17th-century pigment manufacturer and seller from Amsterdam, who wrote down the instructions for the manufacture of vermilion in his workshop. hus the practi-cal value of the information he delivered is unquestionable. Pekstok prepared vermilion with 80 pounds of mercury and 25 pounds of sulphur (Van Schendel 1972: 77–8), amounts that give a ratio of 3.2 parts of mercury to 1 part of sulphur. Pekstok’s ratio represents an excess of sulphur, which could serve to improve the reaction by preventing losses and, con-sequently, reduce the time required to obtain the pigment.Pekstok’s ratio is quite close to some mentioned in late medieval recipes, notably that from the 14th-century Codex Pal. 949 (Pomaro 1991: 95–6) with a ratio of 4:1 (Hg:S); that from the 15th-century treatise he Book on How to Make Colours (Melo and Miguel 2012: 184–5), with a ratio of 5:1 (Hg:S); and that of the 15th-century Cod. Pal. 916 (Pomaro 1991: 92) with a ratio of 4:1 (Hg:S). All these recipes gave instructions for the large-scale production of artificial ver-milion, probably for commercial use. his is confirmed by referring to a later source: in De’secreti , Alessio Piemontese stated that in Italy manufacturers did not know how to make vermilion on a large scale, and that normally it was prepared in small quantities of approximately 1 or 2 pounds (Piemontese 1561: 86–7). According to Piemontese’s instructions, vermil-ion was prepared by mixing 9 parts of mercury and 2 parts of sulphur, 7 which gives a ratio 4.5:1, very close to that of the late medieval recipes. On the basis of this information it can be concluded that recipes for a large-scale vermilion production first appear in the 14th century, and that it became a common practice in the following centuries. his large-scale vermilion production was a two-step dry process with, according to textual sources, a ratio varying from 3:1 to 5:1 (Hg:S). his is not fortuitous: it coincides with the commercialisation of mercury from the Almadén mines in Spain by Catalan, Genovese and Venetian merchants after the 13th century, which increased the exportation, distribution and availability of this product throughout Europe. As previously pointed out, the most common ratio in the early medieval period was 2:1 (Hg:S). Bearing in mind that Pekstok’s ratio 3.2:1 (Hg:S) could provide vermilion in profit-able yields, it could be assumed that the ratio 2:1 is close to a good ratio for vermilion. he difference in the amount of mercury might be explicable technically. A higher percentage of mercury could improve the efficiency of the overall process by compensating for possible losses in the heating process, even if the production costs were higher. Another explana-tion can be proposed on the basis of recommendations by Piemontese, who stated that for the practice of painting, the quantity of sulphur employed is not harmful as it produces a more intense colour, but for any other uses it was better to use a ratio of 3:1 or 2:1 (Hg:S). Piemontese was probably referring to the use of vermilion as a remedy in medicine and as an alchemical substance. he importance of mercury and sulphur in medieval alchemy is well known (Clarke 2012). Pliny and Celsus mentioned the use of mercury in medical applications. rachoma and venereal diseases were treated with natural cinnabar, and quicksilver ointments were used by Arabs to treat skin diseases. But, except for the recipe for artificial vermilion in the alchemical work Libellus de alchimia attributed to Albertus Magnus (13th century) and the recipe of the materia medica treatise acuinus Antonelli (14th century), we do not have any concrete data for the preparation and use of artificial vermilion in any of these domains. Again, later sources can shed light on this aspect. Esteban de erreros y Pando (1786: 431), a Spanish philologist and lexicographer, pointed out that artificial cinnabar was prepared with 3 parts of mercury and 1 part of sulphur, but ‘medicine see this remedy with caution, as it is difficult to eliminate the malignancy of sulphur’. Hernández de Gregorio, a Spanish royal chemist and apothecary, was more explicit when he stated that artificial cinnabar could be made with 2 parts mercury and 1 part sulphur, a proven and standardised ratio, as the most appropriate remedy in medicine and, more-over, as a purer substance than the natural one (Hernández de Gregorio 1803: 294). Is it possible that the oldest mentioned ratio of 2:1 (Hg:S) in textual sources is directly influenced by the medicinal and alchemical use of vermilion? 8 he mutual influence of vernacular science (artistic technology) and official science (medicine, alchemy) was favoured by the common environment of the apothecary, where vermilion was prepared and sold as both pigment and drug. his inter-action was also expressed through copying and dissemination of similar technical instructions in artistic, alchemical and medical treatises. Finally, another aspect to consider is vermilion recipes with the ratio of 1:2 (Hg:S). As mentioned above, this ratio appeared for the first time in the De coloribus et mixtionibus , an 11th- or 12th-century compilation that included recipes for preparing, tempering and mixing pigments. he De col-oribus et mixtionibus differs from the other treatises whose contents are usually vague: this short, specific and precise compilation related to the art of painting and illuminating also includes technical details. hese characteristics suggest that the author was a practitioner and that the instructions he recorded were close to contemporary artistic practice. In fact, these technical instructions can be linked to several recipe books dated from the 13th to the 15th century (ossati Soldano 1978: 55–8). A ratio of 1:2 (Hg:S) means that 6.2 g of mercury reacts with 2 g of sulphur. his reaction takes place with excess of sulphur, but this is not a problem as part of the sulphur can disappear by evaporation in the process and unreacted sulphur can be separated from vermilion in the final washing of the pigment. his means that the recipes with a ratio of 1:2 (Hg:S) can give vermilion in good yields, as the excess of sulphur could improve the reaction yield by ensuring the reaction of all the mercury, avoiding possible losses in the heating process and decreasing the time needed to obtain the pigment. Since the cheapest ingredient is in excess, the production costs would not increase. Additives Within the recipes describing the manufacture of vermil-ion, the use of additives remains an exception, although lead, tin and salt were sometimes mentioned (Merrifield 1967: 478, 480; horndike 1960: 64). Urine was also cited MTA-04-Kroustallis-v2.indd 2608/09/2014 12:57 27 PAINT IT RED: VERMILION MANUFACTURE IN THE MIDDLE AGES as an additive in one recipe of the Mappae clavicula from the Sélestat Manuscript dated from the 9th or 10th century (Sélestat, Bibliothèque Humaniste ms. 17, f. 213v; Smith and Hawthorne 1974: 61, 93). his recipe is identical to a text from a Greek alchemical manuscript, the Codex Marcianus Gr . 299 (f. 106r, Biblioteca Nazionale Marciana) dating from the 10th or 11th century. his observation might suggest a common Byzantine srcin. 9 The mixing of mercury and sulphur he way mercury and sulphur were employed was an impor-tant aspect of the whole process. Within the sources under scrutiny, two methods are described. In the first method, both ingredients were added directly to the pot and then put in the fire. 10 he second method was a two-step process in which metacinnabar (i.e. the aethiops mineral, black β-HgS) was first prepared using a room temperature process by stirring the mixture, 11 or by a hot process by heating and grinding the mixture and then putting it in the fire in order to obtain ver-milion (red α -HgS). his two-step method tended to replace the first one in the recipes from the 14th century onwards.here is no clear relation between the different ratios of mercury and sulphur and the use of one or another method. However, when the ratio 2:1 (Hg:S) was used there was a tendency to employ the two-step method and when the ratio 1:2 (Hg:S) was used it was more common to find both compounds directly mixed in the pot. In that case, some sources advise adding first sulphur and then mercury, prob-ably in order to improve the effectiveness of the reaction by minimising possible losses of sulphur in the heating process. Equipment he vessel employed was almost always a glass bottle or jar with a narrow neck. Usually, the bottle was covered with clay to avoid direct contact with the fire but also to protect the bottle from breaking and to keep the temperature constant. Glazed ceramic pots were less frequently mentioned than expected; when they were it was stated that they should be ‘like bottles, broad below and narrow above’. 12 It is quite prob-able that a bottle-shaped pot could facilitate the reaction as a narrow neck would hamper the evaporation of sulphur and mercury, and would be easier to cover. On the other hand, it is very difficult to take vermilion out of a vessel with a narrow neck: according to some sources, it is necessary to break it to do so. Perhaps we should assume that this was a more common practice than mentioned. Finally, in only in one case was sublimation apparatus (Edgerton 1963: 183) and a glass-worker’s furnace mentioned (Smith and Hawthorne 1974: 61). he reference to a cracking noise inside the vessel was only mentioned in recipes involving a bottle completely covered in clay, the mouth included, ‘so no smoke could escape’. his noise indicated the completion of the process and could be linked to a low concentration of oxygen during the heating process. However, a misunderstanding in the transmission of the recipe is also possible: the first reference to such a noise came from the Sélestat Manuscript recipe where urine was used as an additive and from the alchemical manuscript Marcianus 299 (Sélestat, Bibliothèque Humaniste ms. 17, f. 213v (Venice, Biblioteca Nazionale Marciana, Marcianus 299, f. 106r). Allusion to the cracking noise from the vessel was made only if the fire was too strong. Temperature he heating method and the temperatures reached are very important. he control of temperature and time was based exclusively on the experience and skills of the craftsman. It is thus usual to only find imprecise information about this. In the manufacture of vermilion, the most frequent refer-ences are to the use of a ‘moderate fire’ or ‘live charcoals’. he temperature reached by a heap of red-hot ‘live’ charcoal is approximately 300 °C (Brunet 2009: 62). Charcoal has the advantage of burning at a constant temperature for a longer period of time than wood. If mercury and sulphur were added directly to the pot, some sources recommend start-ing with a moderate fire, possibly to enable the formation of metacinnabar in the same pot, and later to increase the tem-perature (Pomaro 1991: 95–6). Usually, the time of heating was not mentioned, and only indirect indicators were given for the conclusion of the process such as the abovementioned cracking noise inside or the emission of red smoke. In a very few cases a specific period of time of nine hours to a day and a half is mentioned, mostly when weights were given. Melo and Miguel suggested that the production of red mercury sulphide is more a solid-state reaction than a subli-mation process, as a black product is formed at temperatures higher than 580 °C. hey state that when vermilion is mixed with sulphur, a black mercury sulphide is formed as a kinetic product and it is necessary to overcome this kinetic barrier to convert metacinnabar to vermilion. 13 he transformation temperature has been experimentally established to be in the range 335‒386 °C, but research showed that the transforma-tion of metacinnabar to cinnabar is quite slow and it can even take place at lower temperatures (Melo and Miguel 2012: 186–7, 194). Once again both the temperature and prepara-tion time described in medieval treatises of art technology are very close to those reported in modern scientific experiments (able 1). The Spanish ‘dry method’ for preparing artificial vermilion As pointed out above, from the 8th to the 15th century, all documentary sources suggested only the dry method for pre-paring vermilion. Most probably, this was also the method used in the Iberian Peninsula, although until now we have lacked information about the process used by Arabs in the exploitation of the Almadén mining district. 14 However, we have some direct references concerning the artificial manu-facture of vermilion with the dry method immediately after the re-conquest of the area by Christian kings in the mid-13th century. At that time, the mine was granted to the Order of Calatrava to ‘manufacture vermilion from the mercury of the same mines’ and to commercialise it inside and outside the kingdom (Matilla ascón 1958: 13). MTA-04-Kroustallis-v2.indd 2708/09/2014 12:57
