Late Roman Glass at the University of Pennsylvania Museum

A Photo Essay

By: Stuart J. Fleming

Originally Published in 1997

View PDF

The Roman glassmaking industry took root in the late 1st century BC and became established through the 2nd century AD. Business-minded Romans adopted the earlier Hellenistic technology of casting and the novel eastern Mediterranean technologies of free-blowing and mold-blowing to fully develop the com­mercial potential of glass. The growth of the industry was facilitated by the far-reaching Roman trade network which played a vital role in moving not only glassware itself but also glass craftsmen and their ideas the length and breadth of the Empire.

The same forces guided the evolution of Roman glassmaking during the subsequent two centuries, but with some important qualifications. Some broad-impact changes in glassware decoration can still be attributed to innovations that were “traded out” of the eastern Mediterranean and adopted by the major western work­shops (Figs. 1, 2). However, as the provincial cities of the western Empire grew and local leaders usurped more political power, regional productivity and trade flourished and the glassmaking industry tended to frag­ment. In the first half of the 1st century AD there are many examples of glass vessel types with distributions that were Empire-wide, sometimes far beyond. By the end of that same century, however, there are new types made in the Rhineland or north-central France that are simply not found at sites south of the Alps (Fig. 3) (Price 1978).

In subsequent decades, a host of smaller inde­pendent glassmaking workshops sprang up, each thriv­ing well enough by servicing local domestic needs with utilitarian tableware and storage vessels. The hub of the local market would have been either a small military set­tlement or a self-sufficient villa, the likes of which pro­liferated in the agricultural landscape of the western Empire in the early 3rd century AD (Branigan and Miles 1989 King 1990). When, in the wake of Alexander Severus’s death in AD 235, the Roman world fell into a 50-year period of political confusion and fiscally crip­pling wars, glassmaking shed some of its industrialized character. Thereafter, native stylistic traditions re­emerged, region to region.

The story of change for the eastern Mediter­ranean provinces over this same period is rather differ­ent. The economic strength of the Hellenistic world in Egypt and Asia Minor, and along the Syrian-Judean coastline, had provided for the stable growth of crafts such as glassmaking in the 2nd and 1st centuries BC. These provinces did not experience the intensity of the Augustan industrial revolution that transformed the landscape of Gaul and Spain. So the kind of smaller workshop that developed only from the early 2nd centu­ry AD onwards in the western Empire was already firmly established in the eastern Empire by that time. These workshops, like their western counterparts, drew their trade from nearby Roman military settlements and native market towns.

Events of the last decades of the 3rd century AD did, however, put at risk the stability of every craft and industry throughout the Empire. Reports of Germanic raiders appearing deep in the heartland of Gaul and of streams of refugees pouring into the cities in search of protection set off a sense of panic and doom, even though the lives of very few Romans were actually touched by such events. Meanwhile, measures taken to combat economic inflation by coin debasement back­fired and left many a craft and industry—glassmaking included—struggling to define realistic market prices for their products amid uncertain, but usually escalating costs for raw materials (Howgego 1995).

Emperor Valerian’s capture by the Sasanians of Shaper I at Edessa in AD 260 raised the specter of invasion at the eastern frontiers as well. So, despite the price controls on goods and services that Emperor Diocletian tried to enact in his Edict of AD 301, the economies of the eastern provinces became quite volatile (Duncan-Jones 1982; Shelton 1988). Only some three decades later, when Constantine had founded Constantinople as the new administrative heart of the Roman world, did stability return to the region.

I have drawn illustrations for the following vignettes solely from glassware of the eastern Mediter­ranean that dates to the time of Constantine and his suc­cessors. This is from necessity rather than choice, as the Museum’s collections do not contain vessels from the western Empire dating to the Late Roman period. To follow the story of glassmaking in the northwestern and western provinces over this period, the reader might consult a general book on the cultural history of those times—e.g., Cornell and Matthews’ Atlas of the Roman World (1982)—in tandem with one focusing specifically on glassware, such as Harden’s Glass of the Caesars (1987).

Since the individual vignettes are oriented pri­marily toward describing how the various glass vessels were used, let me make some separate observations here about their decoration. To my eye, it displays a remark­able consistency, being based time and again on just four distinct design motifs:

  • A pair of pinched-tooth trails draped down either side of the body of the vessel (Fig. 4).
  • A zigzag meander comprising a single thread of glass wrapped around the vessel’s body (Fig. 5).
  • Two, sometimes three, thicker coils wound around the vessel just once, at various points (Fig. 6). Such coils usually served as a colorful accent to a rim or neck, but sometimes one was applied functionally, as a base.
  • A tightly wound spiral placed close to the top of the vessel’s neck (Fig. 7). A loosely wound spiral was sometimes used on its own (Fig. 8), but I have not treated this motif here, since it has antecedents at least as far back as the 1st centu­ry AD, no doubt because of the simplicity of the technique.

All these motifs are used singly in the 4th cen turfy AD. If there were any regional distinctions in their original use, they soon disappeared in the trade of ideas that flowed almost as easily as the glassware itself along the major roads and caravan routes of the East. Within a century or so the motifs were mixed together in all sorts of ways (Fig. 9).

These motifs were definitely the product of prevailing fashions, regional or otherwise: each was applied to vessels of many different shapes and sizes. Above, I illustrated the zigzag meander as applied to an oil lamp, but it was used to decorate everything from unguentaria to wine goblets (Fig. 10a–e). The persis­tence of these motifs over time—the tightly wound spi­ral extended well into the 6th century AD (see Fig. 26)—demonstrates how the material products of a craft can mirror stability in the political climate surrounding it.

“Xi Rho” Bottle

3 2 – 1 5 – 5 8

The conversion of Constantine in ADC 314 was the catalyst for Christianity’s swift rise to primacy in the Roman world; however, pagan values continued to exert much social influence for at least a century thereafter. Even in Rome, where the Church thrived on the strength of the martyrdoms of Peter and Paul by Nero, the mystery cults of Isis and Bacchus still flourished, and some of the city’s wealthiest families aggressively promot­ed traditional deities such as Ceres and Cybele. And in the Rhineland, scenes engraved on glass and depicted in relief on Roman metalwork were as often classical myths as they were biblical miracles (Harden 1987). Even within families, some members embraced the new faith, others rejected it (Toynbee 1971).

Away from the cities with their fine new churches, in rural areas where the primary concerns were crop and human fertility, the local deities protecting such matters received due worship. For example, at Karanis in the Fayum region of Egypt, where pagan temples were closed in the 3rd century AD, figurines of the deities Isis and Serapis and of cultic animals such as the crocodile (for the god Sobek) are found a century or more later, among all the normal domestic furnish­ings of village houses (Higashi 1990). Full adherence to Christian beliefs about life after death negates the need for any “afterworid” tomb furnishings, yet most Romans continued to be buried with the everyday necessities of toiletries and table-wares (Fleming 1997).

The bottle featured here (Fig. 11) comes from a grave in the massive northern cemetery of Beth Shean in Israel. On the basis of several associated coins, it dates to or just after the reign of Constantinius II (AD 337-361) (Fitzgerald 1931). The grave also contained a pottery oil lamp decorated with a cross (Fig. 12), so we know that the deceased was a Christian. The xi rho impressed on the base of the bottle makes sense in that con­text. Yet, as far as I can determine, it is the only item of non-secular glassware in the 4th and 5th century parts of the cemetery, which consist of about 60 tombs, many of them with multiple burials, at least six of them Christian. Contemporary graves within the same tomb contain only the typical kinds of g1assware-juglets, cups, ointment jars, and cosmetic flasks (Figs. 13-16)—that were then in fashion. Christians probably congre­gated in one another’s houses and used some glass bowls and jugs in blessing rituals at a makeshift altar. But I am inclined to believe that these vessels were domestic first, religious second, special more for their contents than themselves in those moments when their owners reaffirmed their faith.

Scholars are sometimes inclined to assign a name to an artifact that implies a confident understanding of its function. Thus the label “conical beaker” for a vessel of the kind featured here (Fig. 17) goes beyond the obvious aspect of its shape and categorizes it as an item of tableware. To a point, such a label is valid. There are a number of ancient artistic representations of beakers of exactly this conical form being used for wine (Fig. 18). And an engraving on a glass platter found in a 4th century AD catacomb at Beth Shearim, in Syria, depicts such a vessel alongside a pitcher and a two-handled bottle that may be associated with a Talmudic ritual blessing over wine (Avigad 1976). Several domestic contexts of that same date in the farming village of Karanis in Egypt have yielded conical vessels stored among everyday flasks, plates, and bowls (Gazda 1983).

It is also Karanis, however, that provides circumstantial evidence for an alternative label: “conical lamp” (Higashi 1990). During the first half of the 4th century AD, village-wide building activity turned many ground-floor rooms into subterranean grain storerooms. Trapdoors were added for access and to provide ventilation, but these rooms were now windowless and dark. It is surely no coincidence that the archaeological record of these events is marked by the deposition of many pottery oil lamps and a rich scatter of fragments of conical glass vessels. Someone could find his way about such a storeroom with a pottery lamp, but would work there by the light of several longer- and brighter-burning glass lamps set in ring brackets on the walls. Three centuries later, this simple rural lamp had evolved into the common bowl-and-stem type that fit neatly into the holes of a polycandelon and shed light down into a Byzantine church or an Islamic mosque (Figs. 19, 20a). It also may have played some part in the development of the kind of bucket-shaped lamp with an interior wick-tube (Fig. 20b) that became common in eastern mosques from the 12th century AD onwards.

Multiple Flasks with Spatula

MS 5105

Facial beauty was something that every well-to-do Roman woman strived for, and a daily ritual of bathing and cosmetic toiletry preceded her round of social engagements. One maid would pluck out offending gray hairs; others would then pile the hair high and arrange the ringlets to current taste. Yet another maid would powder and rouge the cheeks and outline the eyes in black (see Scarborough Fig. 2, this issue).

The lead sulfide mineral galena was popular in the ancient Mediterranean world for the last of these tasks (Paszthory 1990). Because lead is highly toxic, it is a precarious substance to spread on one’s face, or for a maid to crush and handle as part of her routine duties. It seems, however, that health concerns were outweighed by a love of the subtle silvery luster that characterizes galena (Fig. 22).

For the most part, it is pure conjecture how galena was prepared as a cosmetic; my guess is that the powdered mineral was mixed into an ointment in a small glass pot (Fig. 23), then applied in fine smears off the maid’s fingertip. (Most likely, many other dressings and cosmetic substances—balms for bee stings, tinted pastes to hide skin blemishes—were handled this way as well.) During the 4th centu­ry AD, however, there seems to be a specific association between cosmetic galena and a novel twin-chambered flask that became, over the subsequent century or so, one of the most common items among tomb goods in the east­ern Empire. The flask featured here (Fig. 21) still contains the bronze spatula that was used to portion out the galena. Other flasks of this kind have been found with a roll-tipped applicator, suggesting that one chamber was used to store the mineral, while the other was where it was turned into a paste.

Though many of these flasks are decorated with two of the usual eastern motifs of the time—the pinched-tooth trail and the zigzag meander—others depart strongly from this norm (Fig. 24). Many have a pair of added handles. These handles were not of much intrinsic significance perhaps, except that sometimes they greatly complicated access to the chambers beneath (Fig. 25). We lack any precious metal versions of these flasks that would have been in the cosmetic kits of wealthier ladies, so we cannot invoke mimicry as an explanation of their huge popularity. And that popularity apparently did not extend beyond the eastern Mediterranean; I know of no instance of such a flask (or a form derived from it) being found in the western Empire. Within the broad history of Roman glassmaking, this type of flask is the enigma of its day.


Un Guentarium

MS 5 0 2 3

Pliny the Elder, the 1st century AD encyclopedist, bemoaned the way that the Roman elite squandered their wealth on exotic incense for ostentatious funeral rituals. In a similar vein, his contemporary, the satirist Martial, ridiculed many Roman public figures, women and men, for smothering their hair and bodies with costly scented oils (Donato and Seefried 1989). The anonymous authors of Srriptores Historiae Augustae expressed shock at the eccentricity of the Emperor Elagabalus (reigned AD 218-222) who filled an entire pool with perfumed wine for him and his courtiers to bathe in (Mackie 1982). Yet a pas­sion for fine cosmetics remained one of the constants of Roman daily life in all parts of the Empire (Fig. 27).

Slender-necked unguentaria that became fashion­able during the 6th century AD are among the most aes­thetically appealing products of the glassmaker’s craft. They were functionally effective, too, since the narrowness and length of the neck discouraged evaporation of the ves­sel’s contents. The unguentariurn featured here (Fig. 26) also has a heavily recessed base which, as the scented oil was being poured out, would have held back coarser per­fume ingredients in much the same way as the base dimple of a wine bottle entraps fermentation sediments.

The tightly wound spiral at the neck is a decora­tive motif popular for some two centuries by the time it is applied here (see Fig. 7 and Fleming 1997).At reoccurs among the various flasks and uglets that were used daily in the preparation and mixing of cosmetic ingredients in the home (Figs. 28, 29). The motif seems to simulate a length of cord binding down a temporary seal of leather or cloth.


The vessels featured in the previous vignette are somewhat later than the “Late Roman” scope of the article overall—they are strictly early Byzantine. Including them gives me a chance to look at how far the glassmaking industry had evolved after its emergence late in the 1st century BC. The visual impact of such a retrospective is quite dramatic when it harks back to another unguentaria, the small bulbous vessel with which Roman glassmakers first explored the potential of free-blowing for mass production (e.g., Fig. 30). The linking of the two unguentaria through a sequence of shapes that are typical for intervening times (Fig. 31) allows us to appreciate how societal taste must have instigated change time and again, just as it does today. It is search for an understanding of the cultural dynamics of such change that is the challenge and pleasure of modern archaeological research.

Cite This Article

Fleming, Stuart J.. "Late Roman Glass at the University of Pennsylvania Museum." Expedition Magazine 39, no. 2 (July, 1997): -. Accessed June 15, 2024.

This digitized article is presented here as a historical reference and may not reflect the current views of the Penn Museum.

Report problems and issues to