Figural mosaic images on beads are among the most rare of ancient glass. Early Roman Egyptian face beads of approximately 100 B.C-A.D. 100 represent the most numerous of such ornaments, perhaps existing in the low thousands and were widely distributed through Europe and the Middle East, and are usually regarded as luxury goods. Recent finds place them in the Crimea, Yemen, Iran, Syria, Egypt, Nubia, and Croatia (Sidebotham et al., 2015), as well as Hungary (Liu, pers. comm.). Their most common imagery displays full-frontal faces of Medusa as a Gorgon with stylized snakes as hair, or less often of Medusa as a beautiful woman with luxurious long hair, a necklace and a bust. Late Roman face beads, of the fourth/fifth centuries B.C., number less than thirty, occur primarily in northern Europe and Russia, but have entirely different imagery, most likely emperors. The only other cultures that also produced figural mosaics are the Javanese (Jatim beads of fifth/sixth century A.D., Lankton and Bernbaum 2007), Thai at Klong Thom (tabular face beads that may not be mosaics, possibly first to seventh century A.D.), at Bara in Pakistan and potentially somewhere in Afghanistan, based upon two nearly identical beads with complex griffin and duck mosaics.
For those who study such mosaic face beads and glass with similar canes, there are major, problematic suppositions, since no workshops with such figural canes have ever been found, nor any face canes used for beads. Some Egyptian glass workshop sites did contain mosaic glass/canes (Stern and Schlick-Nolte 1994: 27). However, two larger cane slices depicting the most frequently used Gorgon and Medusa as a woman are in the collection of the Freer Gallery of Art (Liu 2008b: 62, 1.6 cm high and 2.2 cm diameter) and may be very rare examples of face canes before they were pulled or reduced to the diminutive sizes suitable for marvering onto or encasing in a face bead (page 38). Composite mosaic bars with half-face images for theater masks have been found, but these larger, more complex canes were never used for beads (Liu 2008b).
Because the Gorgon/Medusa canes are placed into so many different shapes of beads (spherical, tabular, barrel, and bulla-shaped), numbering between one for tabular, two to eight face canes per spherical bead and varying in how they are placed on the bead, with a great variety of surrounds and colors, in addition to other mosaic imagery, researchers have postulated that Medusa, Gorgon and other canes employed on face beads were most likely produced as a basic cane by skilled glassworkers in Ptolemaic-Roman Egypt, such as in Alexandria, Egypt, the Levant or elsewhere in the Middle East (Henderson 2013, Lankton et al., 2016). These face/Medusa/Gorgon canes, whether in large diameter form or reduced by pulling while hot into smaller sizes suitable for face beads, were then distributed to disparate end users, usually glass beadmakers. Thus the great individual variation found on early Roman face beads. But there has also never been confirmation that face canes were an item of trade, although glass tesserae for mosaics or beads have been found in such contexts (Andersen and Sode 2010, Henderson 2013, Neri et al., 2016).
While we do not know the size of ancient face canes, Brian Kerkvliet (Liu 1989), the first American to make mosaic face canes, used the layering or the hot-strip method for his murrine canes. In one continuous work session of about two hours, he ended with a piece of glass cane approximately three inches in diameter by four inches long (approximately 7.5 x 10.0 cm), before pulling the rod while hot to a size small enough for application to face beads, at approximately 0.35 cm diameter.
The iconography of the three Gorgon sisters, including Medusa, the only one who was mortal, is schematic but not overly rigid when portrayed in glass mosaics: Gorgons have variable number of hair striations or stylized snakes, seen as square/rectangular rods (Liu 2014). Most, but not all Gorgons on tabular beads, also have a red line on the lower portion of the face cane, indicating the blood from her severed head, sometimes misinterpreted as a beard. Very rarely do Gorgon face canes on spherical beads display this red arc of blood.
Medusa, shown with long black hair, a neck with necklace and bust, also shows considerable variation; she is known for her charms and beautiful hair, but in a number of her face canes, the overlying black hair barely covers the stylized snake hair, which protrude from the forehead as knobs, as seen below. Where Medusa is shown with long hair, achieved by hot layering black glass (actually purple), it varies considerably, as do the mosaic bars used to denote her necklace. The application of Gorgon and Medusa as a woman has never been seen on the same face bead, except on one extremely rare glass spindle whorl held by the Metropolitan Museum of Art (Liu 1976). All these images have apotropaic value, most likely adding to their value as luxury beads. Presumably, people contemporary with these face beads would have been able to decipher the meaning of all their features, but not modern viewers, who lack knowledge about their mythology and their iconography.
In 2014, I was able to view high resolution color photographs of the tabular mosaic face beads excavated in the 1920s by Reisner in Nubia from the Merotic culture. These early Roman face beads, besides being the largest cache from a known context, were unique for their lozenge shape, versus round or square for all extant tabular face beads. Like other early Roman mosaic face beads, those from Nubia probably date from about 100 B.C.-A.D. 100, although those from Meroë are more tightly dated to 40 B.C.-A.D. 114. In addition, Gorgon canes were used to represent both Gorgon and Medusa, without overlaid black hair for Medusa, but she had a neck, necklace and bust, and many faces and bodies of both types were in flesh-toned glass, also never before seen in extant face beads (Liu 2014). Gorgon used as a woman, i.e. Medusa, is a rare mosaic image; out of a database of about 220 face beads that I have scanned, only 38 present Medusa as a woman, usually seen as plaques on spherical face beads, even rarer on tabular face beads. Of these 38, only 7 appear to be made from Gorgon canes; none of these modified Gorgon faces had ever been applied to tabular beads. Since Nubians were adept with faience working, glazed stones, glassworking in the form of unique stratified eyebeads with gold bands, as well as enameling (Markowitz and Doxey 2014a), I thought at that time that these Nubian face beads could be an excellent example of local glassworkers modifying imported face canes. But without compositional studies on the glass of face canes and their surrounds, there is no way to prove this supposition. In recent email discussions with James Lankton, he might test some of these Nubian face fragments if granted permission.
With the advent of portable XRF spectrometers and their ability to undertake accurate, non-destructive compositional analyses (Lankton et al., 2016; Liu et al., 2012), museum curators and collectors should be much less loath to have their specimens tested. The Nubian face beads, while weathered to some extent, have mainly non-devitrified surfaces, so they should not skew results of XRF testing. If glass samples of the eyes/nose block, matrix of the face, their hair striations, hot-layered hair and surrounds were tested, as well as the matrices surrounding the cane slices, there should be enough information to compare with the compositional glass databases now being gathered (Henderson 2013; Lankton and Dussubieux 2006). If face beads from other known sites were also tested, the results would offer more for comparison. Large and varying differences in glass of the face canes versus their surrounds could suggest local production of the beads, with the mosaic canes as an import.
The face beads excavated by Reisner could have been imported by the Nubians from Roman Egypt, since there was extensive trade, including luxury items like glass, between gold and agricultural product rich Nubia and Egypt (Markowitz and Doxey 2014b; O’Connor 1993: 89). Our study of these Nubian face beads presented a compelling case for the supposition that all forms of facial images for early face beads were all derived from a Gorgon cane, adapted by beadmakers into Medusa and numerous other variations.
This past October, both Tom Holland and I were going to be in Boston, so I requested permission from Denise Doxey (one of the Museum of Fine Arts, Boston, Egyptian Curators) if we could study the Nubian glass finds, which included those seen previously as photographs and a few others mentioned in excavation reports. Sage and Tom Holland are among the few American glass artists who have replicated ancient glass beads (2003), so I was grateful to have his experienced eye in directly examining these face beads. Since we were not at the museum long, further shortened by an emergency evacuation, I sent a set of my photographs to Sage and Tom so they could analyze them further. Tom’s first observation was that all the Nubian tabular face beads were slices from a whole cane and not face canes marvered onto a bead. Like most tabular face beads, these were therefore not made on a mandrel but were a nipped slice of a cane, reheated for working into a more uniform shape, then hot-pierced, often called rod-pierced. Hot-piercing with an iron rod distorts the cane, as seen in the photographs shown of both broken Nubian and non-Nubian tabular face beads. Interestingly, while the interior glass is distorted, the surface face cane images themselves do not show this effect. In this type of piercing procedure, the glass is hot, while the iron rod is cold. In contemporary glassworking, a red-hot tungsten rod is used to pierce cold glass, the complete opposite of the ancient process.
When we examined the tabular Nubian face beads, both intact and fragmentary, on the surface of almost all Nubian-found tabular face beads, while very well preserved, there was deteriorated white, mottled glass surrounding the face canes, also noted by Reisner in his field sketches. Using macro photography and flash at the right angle, it become apparent that this is merely extensive cracking, crazing or crackling of the glass surrounding the face cane, indicating great incompatibility with the opaque murrine face cane glass. On some, there were also fractures (very apparent with transillumination), both in the glass and on the surface, as well as many pits on the glass surface. Not all tabular Nubian face beads displayed these additional features, nor did other non-Nubian tabular face beads with similar transparent green glass surrounds. The crazed glass appears to be a translucent brown glass, seen in a few examples that had not crazed too badly (page 39). The cracked glass framed the face cane to provide support for the facial components, to prevent surface tension from distorting these features when the mosaic cane was being hotworked. This type of incompatibility was never seen on any other extant tabular early Roman face beads.
Three other face beads from Nubia represented a round green classic Gorgon tabular bead, much degenerated so it was not possible to determine the original color of the face cane; a spherical bead with two faces of Medusa, with cane slices almost matching the green glass of the bead matrix. Rarely are face canes just marvered onto a matrix without different squares of color filling the intervening spaces, or of the face cane not being of a contrasting color to the matrix of the bead. The last was a green barrel bead with two somewhat battered and atypical faces of Gorgon (page 37); this face and that of Medusa were white, unlike the other flesh-toned Nubian tabular beads, and lack cracking/crazing of their glass, although they do show damage from burial. Thus, we strongly believe these three face beads were imports.
While Nubians certainly had skills working silicates such as faience and enameling, there is no evidence of glassworking (Markowitz and Doxey 2014b), although Lacovara (1998) comments on the importance of the Merotic glass industry. The approximately 35 intact, broken and fragmentary tabular face beads excavated from Meroë (W308 number 27, accession numbers 23.830a, b and c, Dunham 1963) show unique features: rhomboid or lozenge shape, versus round or square for all other extant tabular face beads; both Gorgon and Medusa images derived from Gorgon canes, without addition of overlaid hair on Medusa; flesh-colored face and body, versus white for other extant tabular beads; extreme incompatibility between the glass of the face canes and overlaid or surrounding glass, manifested in cracking, crazing and crackling of the glass surface. Despite all these dramatic differences, we cannot conclude whether these Merotic finds are locally altered or are imports. Whether locally adapted or imported from Roman Egypt, all the tabular Nubian beads are the product of the same workshop or beadmaker, and are unique in the totality of early Roman face beads. The other three face beads that were not found in W308 are most likely imports. Perhaps careful compositional testing will provide an answer to the origin and makers of this tantalizing group of Nubian tabular face beads.
Robert K. Liu thanks Denise M. Doxey, Yvonne J. Markowitz, Amelia Kantrovitz, and Carolyn Cruthirds of the Museum of Fine Arts, Boston, for providing both the photographs of the Nubian glass beads, for answering my numerous questions on their attribution, as well as permission to study, photograph and publish the Nubian material. He also thanks Jamey D. Allen for the yellow face bead image of Medusa and prior discussions on Roman mosaic face beads.
Alekseeva, E. M. 1971 Miniature mosaic in glass ornaments, first century B.C. second century A.D. [in Russian]. Sovetskaia Arkheologiia 4: 178-185.
—1978 Ancient Beads of the Northern Black Sea Littoral 2 [in Russian]. Sovetskaia Arkheologiia G1-12: 104 p., plates 20-34.
—1982 Ancient Beads of the Northern Black Sea Littoral 3 [in Russian]. Sovetskaia Arkheologiia G1-12: 85 p., plates 35-54.
Andersen, J. H. and T. Sode. 2010 The Glass Bead Material. In: Ribe Excavations 1970-76, Vol.6, edited by M. Bencard and H. Brinch Madsen. Jutland Archaeological Society, Hojbjerg, Denmark: 17-59.
Auth, S. H. 1999 Mosaic glass mask plaques and the ancient theater. Journal of Glass Studies 41: 51-72.
Bruhn. K.-S. 1995 Designs in miniature: The story of mosaic glass. Corning, The Corning Museum of Glass: 48 p.
Dunham, D. 1957 The Royal Cemeteries of Kush. Vol. IV. The Royal Tombs at Meroë and Barkal. Excavated by the late George A. Reisner, edited and compiled by Dows Dunham. Boston, The Museum of Fine Arts: 218 p.
—1963 The Royal Cemeteries of Kush: Vol. V. The West and South Cemeteries at Meroë. Excavated by the late George A. Reisner, edited and compiled by Dows Dunham. Boston, The Museum of Fine Arts: 466 p.
Engle, A. 1976 Readings in glass history 6-7. Jerusalem, Phoenix Publications: 142 p.
Erdrich, M. and H.-U. Voss. 1997 Die Perlen der Germanen des 1.-5. Jahrhunderts in Mecklenburg-Vorpommern, Schleswig-Holstein und Niedersachsen: 77-93. In: U. von Freeden and A. Wieczorek (eds) Perlen. Archäologie, Techniken, Analysen. Bonn, Dr. Rudolf Habelt GmbH: 386 p., 26 pls.
Ettinghausen, R. 1962 Ancient glass in the Freer Gallery of Art. Smithsonian Publication 4509: 44 p.
Friedman, F. D. (ed) 1998 Gifts of the Nile. Ancient Egyptian Faience. Thames and Hudson and Museum of Art, Rhode Island School of Design: 288 p.
Fukai, S. 1983 Persian glass beads. Kyoto, Tankosha: 238 p.
Goldstein, S. M. 1979. Pre-Roman and Early Roman Glass in The Corning Museum of Glass. Corning, The Corning Museum of Glass: 312 p.
—2001 Ancient glass. In: Ancient Glass, Miho Museum: 178-183. Shiga Prefecture, Miho Museum: 217 p.
Gore, R. 1984 The Dead Do Tell Tales at Vesuvius. National Geographic 165 (5): 557-613.
Harden, D.B. 1967. Some aspects of Pre-Roman mosaic glass. Annales du 4e Congres de l’Association Internationale pour l’Histoire du Verre. Liege: AIHV: 29-38.
Henderson, J. 2013 Ancient Glass. An Interdisciplinary Exploration. Cambridge University Press: 433 p.
Holland, S. and T. Holland. 2003 Master Class: Warring States Beads. Ornament 27(1): 46-51.
Karlin, E. Z. 2007 The historic use of the medusa image in jewelry and the decorative arts. Presented at the Second Annual Fall Jewelry Conference at FIT: A place in time: Jewelry within the context of the decorative arts. October 6, 2007, New York City.
Lankton, J.W. and M. Bernbaum. 2007 An Archaeological Approach to Understanding the Meaning of Beads Using the Example of Korean National Treasure 634, a Bead from a 5th/6th Century Royal Sila Tomb. Beads 19: 32-41.
—and L. Dussubieux. 2006 Early glass in Asian maritime trade: A review and an interpretation of compositional analyses. Journal of Glass Studies 48: 121-144.
—Diamanti, J. and J. M. Kenoyer. 2003 A bead timeline. Volume I: Prehistory to 1200 CE. Washington, D.C., The Bead Museum: 95 p.
—B. Gratuze, K. Tantrakan, Q.H. Li, and S. Liu. 2016 Scientific Analysis of Ancient Glass: Answering Questions and Questioning the Answers. In: F. Gan, L. Qinghui and J. Henderson (eds). Recent Advances in the Scientific Research on Ancient Glass and Glaze. Vol. 2, Ch. 14: 267-301. Hackensack, World Century Publishing Corp. and Singapore, World Scientific Publishing Co., Pte, Ltd: 572 p.
Lierke, R. 1992 Early history of lampwork—Some facts, findings and theories. Part 2. Fire or flame? Lampworking techniques in antiquity. Glastechnische Berichte 65 (12): 341-348.
Liu, R. K. 1974 Glass mosaic or millefiore beads. The Bead Journal 1 (1): 22-26.
—1976 Ancient glass ornaments with human facial images. The Bead Journal 2 (3): 27-32.
—1989 Collectibles: Mosaic Face Beads. Ornament 12 (3): 22-23.
—1995 Collectible Beads. A Universal Aesthetic. Vista, Ornament, Inc.: 256 p.
—2008a Jewelry of the Classical World. The Met’s New Greek & Roman Galleries. Ornament 31 (3): 36-39.
—2008b Roman Mosaic Face Plaques and Beads. Ornament 31 (5): 60-65.
—2012 Islamic Glass Beads. The Well-Traveled Ornament. Ornament 36 (1): 58-63, 70.
—2014 Nubian Mosaic Face Beads. The Enigma of Variations. Ornament 37 (5): 40-45.
Liu, S., Q.H. Li, F. Gan, P. Zhang, J.W. Lankton. 2012 Silk Road glass in Xinjiang, China: chemical compositional analysis and interpretation using a high-resolution portable XRF spectrometer. Journal of Archaeological Science 39 (7): 2128-2142.
Markowitz, Y.J. and D.M. Doxey. 2014a Gold and the Gods. Jewels of Ancient Nubia. Ornament 37 (4): 32-37.
—2014b Jewels of Ancient Nubia. MFA PUBLICATIONS, Museum of Fine Arts, Boston: 184 p.
Neri, E. et al., 2016 Late Roman and Byzantine mosaic opaque “glass-ceramics” tesserae (5th-9th century), Ceramics International (2016), http://dx.doi.org/10.1016/j.ceramint.2016.09.033i.
O’Connor, D. 1993 Ancient Nubia. Egypt’s Rival in Africa. The University Museum of Archaeology and Anthropology, University of Pennsylvania: 178 p.
Rütti, B. 1991 Die römischen Glãser aus Augst und Kaieraugust. Forschungen in Augst 13/1: 370 p., tables.
Sarpellon, G. 1990 miniature di vetro. murrine 1838 1924. Venezia, arsenale editrice: 194 p.
Sidebotham, S.E., I. Zych, J. K. Radkowska and M. Woxniak. 2015 Egypt. The Harbor Temenos in the Southwestern Bay. Polish Archaeology in the Mediterranean. XXIV/1: 306-319.
Selling, D. 1942 Mosaikpärlor med ansiksmasker. Fornvännen: 23-48.
Sode, T. 2004 The Glass Bead Material. In: Ribe Excavations 1970-76, Vol.5, edited by M. Bencard and H. Brinch Madsen. Jutland Archaeological Society, Hojbjerg, Denmark: 83-102.
Spaer, M. 2001 Ancient glass in the Israel Museum. Beads and other small objects. Jerusalem, Israel Museum: 384 p., 2 maps.
Stern, E. M. 2001 Roman, Byzantine, and Early Medieval Glass. 10 BCE-700 CE. Ernesto Wolf Collection. Ostfildern-Ruit, Verlag Gerd Hatje: 427 p.
—and B. Schlick-Nolte. 1994 Early glass of the ancient world. 1600 B.C.-A.D. 50. Ernesto Wolf Collection. Ostfildern, Verlag Gerd Hatje: 450 p.
Stout, A. M. 1985 Mosaic glass face beads: Their significance in northern Europe during the Later Roman Empire. Vols. I, II. Ph.D. dissertation, University of Minnesota: 359 p.
—1986 The archaeological context of Late Roman Period mosaic glass face beads. Ornament 9 (4): 58-61, 76-77.
Tempelmann-Maczynska, M. 1985 Die Perlen der römischen Kaiserzeit und der frühen Phase der Völkerwanderungszeit im mittereuropäischen Barbicum. Romisch-Germanische Forschungen 43: 339 p., 80 Tafe
Robert K. Liu is Coeditor of Ornament and for many years its in-house photographer. His recent book, The Photography of Personal Adornment, covers forty-plus years of shooting jewelry, clothing and events related to wearable art, both in and out of the Ornament studio. A frequent lecturer, some of his topics include precolumbian jewelry, prehistoric Southwest jewelry and ancient Egyptian jewelry. In this issue Liu revisits the unique early Roman mosaic face beads found in Nubia, and enlists the help of glass bead experts and replicators of ancient glass, Sage and Tom Holland.