ChicagoNewman, Richard, and Glenn Alan Gates. “The Matter of Madder in the Ancient World.” In Mummy Portraits of Roman Egypt: Emerging Research from the APPEAR Project, edited by Marie Svoboda and Caroline R. Cartwright.
Los Angeles: J. Paul Getty Museum, 2020. https://www.getty.edu/publications/mummyportraits/part-one/3/.
MLANewman, Richard, and Glenn Alan Gates. “The Matter of Madder in the Ancient World.” Mummy Portraits of Roman Egypt: Emerging Research from the APPEAR Project, edited by
Marie Svoboda and Caroline R. Cartwright,
J. Paul Getty Museum, 2020. https://www.getty.edu/publications/mummyportraits/part-one/3/. Accessed DD Mon. YYYY.
MadderCitation: Madder. A dyestuff derived from the root of the madder plant (Rubia tinctorum), which is native to the eastern Mediterranean and Persia. Likely introduced to Egypt by the Greeks or Romans, madder was used throughout antiquity for coloring textiles and as a pigment. Chemical name: Alizarin (1,2-dihydroxyanthraquinone), Purpurin (1,2,4-trihydroxyanthraquinone), a lakeCitation: Lake. A pigment manufactured by precipitating a dye onto an inorganic substrate/mordant (such as the metallic ions aluminum or calcium).pigmentCitation: Pigment. A colorant either derived from natural sources—mineral, plant, or insect—or produced synthetically. Typically, pigments are crushed into a fine powder and mixed with a binder, resulting in a suspension that becomes insoluble when dry; a dye produces a lake pigment when attached to an inorganic substrate or mordant., may be one of the more common colorants used in Egyptian mummy portraits. As a pure pigment, it is pink or red, sometimes slightly purplish, and is most often noted as the major coloring in red and purple drapery and claviCitation: Clavus (pl. clavi). A vertical stripe or ribbonlike ornament, placed in pairs, that adorned the shoulders of a tunic. In Rome some clavi of specific width and/or color distinguished members of particular rank or status, but the significance of the clavus in an Egyptian context remains undetermined. (figs. 3.1 and 3.2). Madder has been found mixed with a blue pigment in purple drapery, and it is commonly noted on lips and highlights on faces.
This paper will briefly review the nature of madder as well as methods by which it can be identified by noninvasive procedures and analyses of samples; a few specific instances of its use in mummy portraits will be described. Madder has been the subject of extensive research, from many points of view, and only a very few selected publications from this rich literature can be cited here.
The Rubiaceae family of plants, comprising more than thirteen thousand species in 617 genera,1 includes many from which red colorants can be extracted. Colorants from Rubiaceae have been utilized in many parts of the world as textile dyes and pigments. In Europe and western Asia, perhaps a half dozen Galium species were likely to have been used as textile dyes.2 In Europe, the dyes from this genus are commonly known as bedstraws and woodruffs; a related dye, Asperula tinctoria L. (often called dyer’s woodruff), was also probably important. Although several Rubia species are found in many of the same regions as Galium, their range does not extend as far north as that of Galium, and they grow in regions farther south. Rubia tinctorum (often referred to as common madder or dyer’s madder) is found in central and southern Europe, northern Africa, and central Asia. Rubia peregrina (wild madder) has a range mostly restricted to central Europe. Rubia cordifolia (Indian madder or munjeet) is found in central and eastern Asia, eastern Africa, and parts of Australia. Pliny the Elder suggested that by the first century AD at least, R. tinctorum was widely grown throughout Italy and the eastern Mediterranean, and it was a ubiquitous source of red dyes for textiles.3 There is less certainty about the relative availability of the other madders.
The word madder is usually restricted to species in the Rubia genus. Because many of the same chemical compounds occur in Rubia, Galium, and Asperula, for the purposes of this paper madder will refer to red colorants extracted from any of these botanical sources.
The colorants are found in the cores of roots and rhizomes (underground stems). At certain times of the growing season, the roots exhibit a pink to strong red color, and it is easy to understand how they would have been attractive as potential color sources (figs. 3.3 and 3.4). Extracting the colorants is simple: dried roots are crushed and soaked in hot water. Simple water extracts, if used alone, would not have been fast to moisture if used as stains or dyes, and it seems likely that early in the historical use of madders as colorants, they were combined with inorganic mordants to attach them to textiles’ fibers.4 Inorganic ions, such as aluminum, form complexes with natural compounds found in the roots that facilitate strong bonds to textile fibers and also typically turn the roots’ colors more reddish shades. For use as pigments, the extracts were probably prepared as lakes, as discussed below.
The compounds responsible for the color are hydroxyanthraquinones (HAs). HAs found in Rubiaceae are derivatives of 9,10-anthraquinone, with hydroxyl (and sometimes other groups) substituted on ring A. In roots, many of these are present mostly in the form of glycosides, in which a HA (or aglycone) is bonded to a sugar molecule (a monosaccharide or disaccharide). Some common glycosides are primeverosides, in which the HA is bonded to the disaccharide primeverose. About three dozen different HAs have been identified in the roots of Rubia tinctorum, the most extensively studied of the madder-producing plants.5 Structures of some common aglycones and glycosides in madders are shown in figure 3.5.
Air-drying of fresh roots leads to hydrolysis of the native glycosides by endogenous enzymes, although glycosides may be preserved if fresh roots are strongly heated, which destroys the enzymes.6 Glycosides can also be easily broken down during other steps by which plant extracts are prepared and attached as dyes to fibers or prepared for use as pigments. As a result, samples of pigments or dyes extracted from textiles usually contain only aglycones. There can be considerable variations in the specific aglycones and their relative amounts found in the end products of their use—be it textile or pigments—even if they were prepared from only one species of plant.7
HAs can also be extracted from a number of types of scale insects, and at least four major types have or may have served as important historical sources of dyes in the ancient world: kermesCitation: Kermes. An insect-derived ancient red dye/colorant and source of the word crimson. Early Egyptians made this red dye from the dried bodies of a female wingless scale insect—either Kermes ilices or Kermes vermilio, both of which live on certain species of Mediterranean oaks and produce a powerful, permanent scarlet dye and organic colorant. Chemical formula: kermesic and flavokermesic acid, C16H10O8 (from Kermes vermilio), Armenian cochineal (Porphyrophora hamelii), Polish cochineal (Porphyrophora polonica), and lac (Kerria lacca).8 The HAs in these sources, which contain substituents on rings A and C, do not overlap with any found in Rubiaceae.
Identification of specific HAs has been carried out by several techniques. Of them, the one that in principal requires the smallest sample is surface-enhanced Raman spectroscopyCitation: Raman spectroscopy. An analytical technique used to observe the vibrational, rotational, and other low-frequency molecular modes of a material. When excited by monochromatic light (visible, near infrared, or near ultraviolet) from a laser beam, the collected inelastic scattered light collected with a spectrometer produces spectra that are specific to the chemical bonds and symmetry of specific molecules. Comparing reference spectral databases allows for the identification of materials. (SERS). It has been shown that three of the major HAs found in many Rubiaceae (alizarin, purpurin, and pseudopurpurin) can be identified by SERS; however, if more than one HA is present in a given sample, SERS may detect only one of them.9
Typically, techniques that can confidently identify two or more HAs simultaneously require larger samples than does SERS. (The actual sample size or weight required for any analytical technique depends on many factors, including the concentration of the compounds of interest in the sample.) Liquid chromatography (LC) is a proven workhorse, either with diode array detectors (LC/DAD), mass spectrometer detectors (LC/MS), or both (LC/DAD/MSCitation: Liquid chromatography with diode array detection and mass spectrometry (LC/DAD/MS). Many natural dyes consist of more than one chemical compound, and LC is a technique by which the compounds can be separated and then individually identified by DAD and MS detectors.). Successful applications of LC to Rubiaceae-based dyes and pigments are extensive.10 Other mass spectrometric techniques, not combined with chromatography, have been less commonly utilized to date.11 Noninvasive methods by which the presence of madder can be established or at least hypothesized are discussed later.
Alizarin, purpurin, and/or pseudopurpurin are the HAs most commonly detected in chromatographic analyses of madder pigments. Many researchers have used the types and relative amounts of these HAs, as indicated by peak heights or areas in chromatograms, to hypothesize the botanical source of madder in specific samples.12 Of the various genera and species noted earlier, alizarin is the predominant HA only in certain extracts from R. tinctorum. If alizarin is absent or negligible, then Galium, R. cordifolia, or R. peregrina are the more likely sources. Conclusions can be problematic, as profiles of HAs are profoundly affected by processing of the raw materials, procedures used to prepare a dye or lake, and preparation of samples for analysis. In some instances, certain other HAs, present in minor amounts, may suggest specific botanical sources, but there is no general agreement on how useful the presence or absence of such minor components is for this purpose.13
The earliest published occurrence of a madder dye in a cultural artifact is in some (now unavailable or lost) fragments from Mohenjo-daro, an Indus valley site dated to about 3000 BC.14 However, the analysis may not have been correct given techniques available at that time. The earliest confirmed identification of madder (from an analysis carried out by SERS) is on an Egyptian leather quiver fragment dated to the Middle Kingdom, Dynasty 11 (ca. 2124–1981 BC), excavated at Thebes.15 The dye was rich in purpurin.
Some textiles excavated at a late second millennium BC site in the Tarim basin of western China were likely dyed with R. tinctorum.16 This species has also been identified as the source of red dye for some textiles, dated to the eleventh and tenth centuries BC, excavated at Timna, Israel.17
Analyses of textiles, apparently discarded between the first and third centuries by Roman garrisons at eastern desert sites in Egypt, identified two “types” of madder,18 likely from more than one type of plant. Dyers must have been aware of how different plants and manipulations in extraction and preparation procedures could affect the tint of the final product.
Textile dyeing and the making of lake pigments were separate operations, but lakes may have been made in some of the same workshops in which dyed textiles were produced. In later European history, the colorants used to make lakes were at least at times extracted from scraps of dyed textiles, and lakes in the ancient world could also have been made from such scraps.19
The earliest examples in a recent compendium of madder-based pigments in the ancient world are from Cypriot pottery of the eighth and seventh centuries BC; these works reportedly contained alizarin, purpurin, and possibly pseudopurpurin.20 Some paint samples contained significant alizarin, while others contained little or no alizarin. A small bowl of pseudopurpurin-rich madder lake was among several that were excavated at HawaraCitation: Hawara. A Roman site in Egypt located in the Fayum basin. The necropolis at this site is well known for the systematic and well-documented excavations by British Egyptologist Sir Flinders Petrie., the region in which many mummy portraits are thought to have been created.21
Analysis by LC/MS of purple pigments from a Hellenistic sculpture identified an alizarin-poor madder mixed with Egyptian blueCitation: Egyptian blue (cuprorivaite). A pigment that was manufactured and used by Egyptians possibly as early as 3100 BC. Considered to be the first synthetic pigment, Egyptian blue was made by mixing a calcium and copper compound with silica/quartz and a flux, heating the mixture to a very high temperature (900°C), and then grinding the glassy product to a powder. Chemical formula: Calcium copper silicate, CaCuSi4O10 or CaOCuO(SiO2)4, but traces of insect dyes were also found: a type of cochineal and lac.22 Lakes made exclusively from insect dyes may have been utilized in the ancient world, but there are few examples as yet.
Little is known about how lake pigments were made in the ancient world. Typical later procedures in Europe involved mixing the root extracts with a soluble aluminum sulfate salt (such as alum), then adding an alkali (such as plant ash) to precipitate the lake. The alkali can be used first, followed by addition of the soluble aluminum salt.23 Particles of lakes created by these procedures consist of HAs in complexes with aluminum (possibly also with smaller amounts of other elements, such as calcium) within particles of aluminum sulfate and mostly amorphous aluminum hydroxide.24Calcium carbonateCitation: Calcium carbonate (chalk, lime, calcite). A chemical compound used to create a stable white pigment with limited hiding power (opacity); this pigment is used to make grounds (preparation layers) for painting. Chemical formula: CaCO3, if used as an alkali, can result in a lake that contains calcium sulfate, which would be precipitated during the manufacturing process. After drying, a lake pigment can be ground like a mineral pigment, then mixed with the paint medium.
The substrates of ancient madder-type lakes have been identified as aluminum rich, clay containing, or based on calcium carbonate or calcium sulfate.25 The possible presence of extenders (or white pigments) in samples makes precise identification of the lake substrate difficult, since most published analyses have not been carried out on isolated lake particles.
When paint cross sections containing lake particles are available, the substrates of the lake can be studied by scanning electron microscopy / energy-dispersive X-ray spectrometry (SEM/EDS) with little interference from other compounds present in the paint sample. Figure 3.6 is an image of a cross section from a purple clavus in a mummy portrait (see fig. 3.1). The purple paint contains gypsumCitation: Gypsum (calcium sulfate dihydrate). A soft sulfate-based mineral found in nature. Often mixed with water to form plaster, it is used in the preparation of substrates, such as wood panels for painting. Also used as a white pigment, gypsum was identified in Tutankhamen’s paint box. Chemical formula: CaSO4·2H2O, natrojarosite, indigoCitation: Indigo. A natural blue dye derived from the plant Indigofera tinctoria and related species growing in the Mediterranean, India, and Asia, among other locations. It is believed that originally the dye woad (Isatistinctoria), rather than indigo, was used in antiquity. Chemical formula: C16H10N2O2, and a red lake that fluoresces orange. The SEM/EDS spectrum from the largest lake particle in the image (fig. 3.7) is quite similar to spectra from madder lakes in later European paintings: aluminum is the major element, with some sulfur and smaller amounts of several other elements associated with the raw materials that had been used to make the lake. The particle encompasses several very small grains of lead whiteCitation: Lead white. A white pigment, both found as a naturally occurring mineral known as hydrocerussite and produced synthetically by exposing metallic lead to an acid (e.g., vinegar). Lead white has been widely used in antiquity and in Egypt since around 400 BC. Chemical formula: Basic lead (II) carbonate, 2PbCO3·Pb(OH)2, perhaps an unintentional component of the lake. Lake-containing samples from mummy portraits examined at the Museum of Fine Arts often contain angular particles of calcium sulfate (usually gypsum) and finer-grained material that is rich in aluminum. The latter regions exhibit orange fluorescence. A chip of one pink paint sample is shown in figure 3.8. In this instance, the gypsum was likely added to the paints as a ground mineral after the lake was manufactured, rather than precipitating during manufacture.