The earliest glass vessel decorated with enameling dates from about 1425 B.C.E.¹ It successfully combines one of humankind’s oldest creative urges (the desire to draw on things) with one of the most advanced technologies of the ancient world (glassmaking). Today, essentially the same process continues to be widely used for the creation of permanent decoration on the surface of objects ranging from bottles produced by machines to original works of art fashioned by celebrated studio glass artists.

The purpose of this study is to reveal the surprising degree of complexity in a process that, when first considered, deceptively seems so straightforward, and to correct widely held misconceptions of how that process was carried out until relatively recently.² A more accurate understanding of the process should benefit both specialists and connoisseurs, and it will probably have an impact on such important issues as the dating and authenticating of objects.

Enamel decoration consists of colored glass (or a mixture of colorless glass and a coloring agent) that is pulverized and painted, in the form of fine granules suspended in a liquid, onto a cooled glass vessel, then permanently fused to its surface by intense heating. Enameling stands in sharp contrast to cold painting, which, when executed on a smooth, unprotected glass surface, is famously impermanent.

These two basic types of enamel preparations can be in the form of high-fire or low-fire enamels. High-fire processes can soften and possibly damage an object, while low-fire processes never soften an object to the point of imperiling it. High-fire processes can be carried out in a glass furnace or, after the 18th century, in a kiln, while low-fire processes always take place in a kiln.³

The Enamels

Enamel Types

The following classifications are based on how and when the enamels get their color. Chemical analyses of enamels (and the glasses to which they have been applied) from a broad range of historical objects have been widely published, along with useful interpretations of the data.⁴ For the purposes of this study, which will focus on the processes by which the objects were decorated rather than on how the enamels were made, only one distinction will be observed concerning the physical structure of enamels. The identification of these two distinct types of enamel, in conjunction with the traditional categorizations based strictly on chemical composition (e.g., soda-lime glasses versus lead-silicate glasses), may afford a more complete picture of these materials.⁵

Graph A
A schematic temperature/time profile for firing low-fire enamels in a kiln, using the low-fire process.

 

Premelted (or Prefritted) Enamel. This type of enamel consists of finely pulverized colored glass mixed with a liquid medium. The glass must be carefully formulated so as to provide an adequate intensity of color while being physically compatible with the vessel glass. Compatibility means that, when the vessel is cooled after the enamel is fired, the enamel must decrease in volume at about the same rate and to about the same degree as the glass to which it has been fused, or they will crack apart. Almost all modern commercial enamels are of this type.

Cold-Mixed (or Unfritted) Enamel. Here, an essentially colorless glass (often, it is the same glass as that from which the vessel itself was made) is pulverized and mixed with a coloring agent such as a metallic oxide (e.g., cobalt oxide for blue) or a colored mineral (e.g., hematite for red), in addition to a liquid medium. During the firing process, when a metallic oxide has been added, chemical reactions take place between the glass (flux) and the coloring agent (pigment) that cause the glass to become colored. When particles of a colored mineral have been added to a colorless glass and fired, they give the enamel the appearance of having been colored throughout. In both enamel types, during the firing process, the particles of glass fuse with one another and to the surface of the vessel, thus ensuring permanence.

The manner in which the enamel attains its color is distinctly different between the two types. The cold-mixed type acquires its color during the firing process, while the color of the premelted type was achieved by chemical reactions that took place long before the firing process, when the ingredients were melted in a crucible within a glassmaking furnace. When viewed in cross section under powerful magnification, these two types of enamel look quite different.⁶ The premelted enamel appears essentially homogeneous, since its ingredients were well mixed by convection in the crucible when they were “cooked.” By contrast, the cold-mixed enamel appears as a sea of colorless glass that is somewhat sparsely populated with colorant particles, each of which is surrounded by a veil of colored glass. The coloration of the glass is strictly limited to the immediate vicinity of the interface of the colorless glass and the colorant.

Enamel Categories

The following classifications are based on the firing temperature of the enamel in comparison with the temperature at which the vessel might begin to become distorted under its own weight.

High-Fire Enamels. For firing, these enamels require exposure to such a high temperature for such an extended period that without some intervention on the part of the glassworker, the vessel would begin to collapse under its own weight. High-fire enamels require either a furnace-firing procedure or the kiln-firing procedure known as flash firing (see below). These were the only enamels available for decorating glass until the 19th century.

Low-Fire Enamels. Low-fire enamels can be fired at a temperature that is well below the softening point of the vessel glass. Even if that temperature is held for an extended period, the form of the vessel is not in danger of changing because of the softening.

The making of low-fire enamels requires a markedly greater sophistication in glassmaking ability. For success, three criteria must be met, compared with two for high-fire enamels. The color and the compatibility must be managed correctly, and the temperature/viscosity characteristics of the enamel must be precisely adjusted and controlled. Apparently, this considerable technological feat was not accomplished for a wide range of colors until sometime in the 19th century.

Firing Processes

The following classifications are based on the type of appliance used and the temperature attained during the firing process.

Kiln Firing. The firing of enamels on a glass vessel within a kiln has apparently been possible only since the 19th century. Although both low-fire and high-fire enamels can be fired in a kiln, it is low-fire enamels that make kiln firing an easy procedure that continues to be used almost universally today. Two very different strategies are employed when firing objects in a kiln: the low-fire method and the flash-firing method.

The low-fire method (graph A) is by far the easier of the two kiln-firing methods. A blank decorated with low-fire enamels is placed in a kiln and slowly heated to a temperature that is sufficient to fire the enamels. Because the firing temperature is well below the point at which the vessel glass might begin to become distorted, the form of the object is not endangered. The completed object is then annealed (slowly cooled to room temperature).

Graph B
A schematic temperature/time profile for firing high-fire enamels in a kiln, using the flash-firing process.

 

The flash-firing method (graph B) has long been employed in top-fired kilns for the firing of flat glass decorated with high-fire enamels, but its use in the firing of vessels is comparatively recent. This is because a kiln suitable for holding vessels must be capable of great, sudden, and highly controlled temperature changes. These are best accomplished in a modern gas or electrically heated appliance. In addition, the operator must be both highly knowledgeable and extremely vigilant during the process.

In flash firing, the decorated object is gradually heated to a temperature just below the annealing point of the glass (about 850°F or 455°C for common soda-lime glass). The temperature is then suddenly “spiked” to a level that is sufficient to fire the enamels quickly (perhaps 1,200°F or 649°C). After a short interval, perhaps one minute, the operator opens the kiln door to confirm that the firing has succeeded (by observing that the enamels are glowing orange and shining) and to begin a precipitous reduction of the temperature inside the kiln. A moment later, the door is closed and the temperature is stabilized at the annealing point (about 900°F or 482°C for common soda-lime glass). The temperature is then gradually reduced to anneal the object.

Flash firing is possible because of the exceptionally poor thermal conductivity of glass compared with, say, almost any common metal.⁷ When the temperature in the kiln suddenly spikes well above the softening point of the glass, the vessel is not deformed because only its surface initially becomes soft; within the vessel wall, the glass is still stiff. By the time the firing is completed and the inside of the wall is beginning to soften, the kiln door has been opened and the surface of the vessel begins to stiffen with the sudden drop in temperature, thus ensuring its integrity. Because of the great thermal agility of modern kilns and the precision of the computers that control them, this flash-firing method is used in mass production today nearly as often as the low-fire method with low-fire enamels. This is because high-fire enamels invariably have markedly higher durability than low-fire enamels, which is an important asset in such industries as the manufacture of returnable bottles.

Furnace Firing (graph C). Until the 19th century, a glassworker’s only option for firing enamels on a glass vessel involved holding the decorated blank on the end of a long tool and reintroducing it into the intensely hot glassworking furnace. The decorated blank was very gradually reheated to a temperature slightly above the annealing point of the glass (about 950°F or 510°C for modern soda-lime glasses). When the blank was hot (but not sufficiently hot to fire the enamels), it was attached to a tool and removed from the kiln in which it had been preheated. From the Middle Ages onward, the pontil was used to hold the blank. Earlier, the blank had been held by the metal rod used in the coreforming process or the blowpipe that had been employed to blow the blank.

Graph C
A schematic temperature/time profile for firing high-fire enamels, using the furnace-firing process.

 

Next, the decorated blank was immediately thrust into the glass furnace (or into another similarly high-temperature—about 2,000°F or 1,093°C—furnace) and held there until the enamels began to glow orange and shine. To the glassworker, this visual cue indicated that the enamels had fully fused, or fired. In Renaissance Venice, the very hot blank would periodically have been removed from the furnace in order to allow the shaping process to be resumed. When the firing process had been completed, the finished object was broken free of the pontil and placed in an annealing oven.

A Chronological Survey of Historical Objects

The following survey of about 60 enameled objects is intended to describe and illustrate all variations of the enameling process as it was carried out between the 15th century B.C.E. and the 18th century. The reconstructed processes are based on a close hands-on examination of historical objects, extensive practical experiments, and a variety of period descriptions of decorating processes. The most useful and detailed of these accounts is a 15th-century manuscript in the Library of San Salvatore in Bologna. It was brought to my attention by Hugh Tait in 1982, and it became the seed from which this study sprouted. That text says:

To paint glass, that is to say, cups or any other works in glass with smalti or any colour you please, take the smalti you wish to use, and let them be soft and fusible, and pound them upon marble or porphyry in the same way that the goldsmiths do. Then wash the powder and apply it upon your glass as you please and let the colour dry thoroughly; then put the glass upon the rim of the chamber in which glasses are cooled, on the side from which the glasses are taken out cold, and gradually introduce it into the chamber towards the fire which comes out of the furnace and take care you do not push too fast lest the heat should split it, and when you see that it is thoroughly heated, take it up with the pontello and fix it to the pontello and put it in the mouth of the furnace, heating it and introducing it gradually. When you see the smalti shine and that they have flowed well, take the glass out and put it in the chamber to cool, and it is done. . . .⁸

Egypt (15th Century B.C.E.)

The practice of creating enameled decoration on glass vessels began very soon after such vessels were initially made. A small turquoise-colored jug in The British Museum (Fig. 1.1) was probably found in the tomb of Thutmose III at Thebes, and it is thought to have contained sacred oils used in the burial ritual. The presence of the pharaoh’s cartouche on this vessel indicates an 18th-Dynasty date of about 1425 B.C.E. Like other glass vessels associated with Thutmose III, the jug was made by core forming. Before this technique was displaced by glassblowing in the first years of the Common Era, it was the most efficient way to produce small, usually closed vessel forms. Core forming was the dominant method of making glass vessels for some 1,500 years.⁹

Throughout this period, the technique was virtually unchanged. A mixture of materials—most likely dung, clay, and water—was kneaded together, then formed to take the shape of what was to become the vessel’s interior. A long metal rod was inserted partway down its length to be used as a handle. The core was then thoroughly dried and gradually heated to a temperature sufficient to both fire the clay and burn away all of the organic matter. The heated core, now sturdy, was covered with glass and decorated, typically by trailing contrasting colors of glass around the body, then combing or feathering them. Next, the object was annealed to avoid cracking. Finally, the rod was pulled out, and the core, now crushable and brittle, was picked and scraped out, leaving a cavity in the vessel.

Recent investigations of this ancient process have focused on the exact method by which the core was covered with glass.¹⁰ The compositions of ancient cores are less uncertain: surviving fragments have been analyzed, and contemporary experiments have demonstrated that combining the above-mentioned materials in a fairly wide range of percentages usually works successfully.

In the Thutmose III jug, the turquoise glasses of the vessel body and the handle are equally dense and homogeneous, without the granular, seemingly aerated quality that is often observed in core-formed vessels, particularly the later examples. While a convincing argument can be made that the walls of some of these objects were built up by fusing finely crushed glass to the core, in this instance the core seems to have been covered by dipping it in a crucible of molten turquoise glass.

The process of enameling and firing this jug was surprisingly complicated, and it bespeaks a highly inventive craftsman. Because it was closely associated with Egyptian royal funerary practices, we can be sure that both the workers who produced such burial accouterments and their customers were acutely concerned with permanence. It is remarkable that, at a moment so early in the history of glass technology, it was accurately understood that cold painting on smooth glass would not last indefinitely. Because the workers were apparently already convinced of the stone-like durability of the glass, it was entirely logical for them to invent a way of decorating it with a paint that was itself glass, and then to fuse the two together, fixing the decoration by firing.

The maker (or makers) of the Thutmose III jug used a three-stage process for creating permanent enameled decoration on this glass vessel. In principle, as we will see, this process remained unchanged until the end of the 18th century, some 3,200 years later.

First, a blank was made by the core-forming process. The bottom was left rounded, and neither the decorative yellow wraps nor the handle that we see in the finished object was attached. The core and the rod were left in place after the completed blank had been annealed (see Figure 2.1, in which the undecorated blank is shown at the right on the table).

In the second stage of the process, the enamel was painted on the cooled blank. The enamel consisted of a mixture of finely ground yellow glass (this glass was undoubtedly the same as that employed for the yellow wraps) in a liquid medium that was perhaps made of gum arabic and water or a light oil (Fig. 2.1–2.3). We can be certain that the decoration was applied in liquid form because there is vestigial evidence of “flow” in the longer lines and of the migration of finer particles to the edge of some droplets. This is particularly visible at the rounded tips of some of the leaves and the ends of the stems (Fig. 1.1 and 1.2).

In the third stage, the object was fired in a series of critical steps, almost all of which imperiled it. The decorated blank was reheated very slowly to a point at which the glass would soon have become distorted under its own weight if left unattended (about 950°F or 510°C for common modern soda-lime glass).

Reheating glass, especially vessels, invites cracking. Three factors that are now well understood are involved here. First, as glass is heated from room temperature to the vicinity of its annealing temperature, it expands considerably. Second, as we have seen, glass is an exceptionally poor conductor of heat. Third, because of its varying shape and wall thickness, a glass vessel is much more difficult to heat evenly than, say, a piece of plate glass. When a glass vessel is reheated, different parts inevitably heat and thus expand unevenly, and some areas are exposed to thermal stress. If these localized forces are sufficiently concentrated in a small enough area to overcome the strength of the glass, a crack will form and the object will break. Even today, reheating a glass vessel in a computer-controlled electric kiln remains a tense undertaking. In the time of Thutmose III, it must have been positively nerve-racking.

In any case, after the object had been heated well above its annealing temperature, it was held by its temporary handle in the glassworking furnace, probably until the enameled decoration glowed and shone. The shine would clearly have signaled the worker that the particles of yellow glass had softened and fused, both to each other and to the vessel (Fig. 2.4). At this point, work resumed on finishing the vessel (Fig. 2.5–2.8). Part of the handle sits atop the upper yellow band, so we can be sure that the handle was the last addition. Less certain is the order in which the yellow bands were added. It is interesting to note that, as the flat bottom was created, the lower third or so of the vessel was soft. This is indicated by the subtle spiraling of the lowest part of the stems of the palm leaves and by the linear breakup of the larger, less well fused enamel particles (Fig. 1.1 and 1.3). This twisting and stretching resulted from the combination of torque from rotating the object and friction from pressing a flat tool against the tip of the soft glass.

Incredibly, the jug for Thutmose III appears to be the only surviving example of an enameled glass vessel that predates the Common Era. Before it was discovered, and then acquired by The British Museum in the 19th century, it had been entombed in the Valley of the Kings for considerably longer than two millenniums. Therefore, during that vast span of time, all subsequent reinventions and losses of the process that the jug so vividly reveals took place without its influence as creative inspiration or technical example. While we can debate, for example, whether the Renaissance Venetian enameling process was based more on Islamic or Byzantine practice, it is certain that in our current view of the history of enameled glass vessels, the remarkable Thutmose III jug represents both a beginning and an end.

The Roman Period

By the time the technique of enameling glass vessels was finally reinvented after 18th-Dynasty Egypt, the revolutionary process of glassblowing had been discovered. The lacuna in the history of enameling—some 14 centuries—is suspiciously immense. Because its first appearance coincided so closely with the start of glass vessel making, it is difficult to believe that this process did not reappear at least once before the Romans arrived on the scene. However, without any recognized surviving artifacts or documentary descriptions, we must—for the moment, at least—assume that both the concept of this type of decoration and the means to realize it were simply lost and long awaited rediscovery.

The absence of enameling on Greek and Hellenistic glass is very surprising, particularly during the last centuries B.C.E., when decorative enameling was used on gold. Because these glassworkers commonly employed hot-forming processes such as slumping (kiln forming), in which nascent objects sat stationary on molds while soft, it would have been easy for them to fire enamels on the surface of bowls, for example.¹¹ Their only technical challenge would have been to produce (or acquire from elsewhere) enamels compatible with the glass from which the objects were made. Apparently, this was a missed opportunity.

Roman vessels with enameled decoration were made over a period of about 300 years, beginning in the early decades of the empire. In this survey, each of those centuries will be represented.

First Century

The extensively decorated transparent green cup that is now in the Museo Civico e Archeologico in Locarno, Switzerland (Fig. 3), was found in the Muralto (Locarno) cemetery in 1936. It is the best-preserved example of a large group of similarly made and decorated objects whose find-places are widely distributed throughout Europe and the Middle East.¹² Datable to about 20–70 C.E., these cups fall into various categories according to the decoration of their bases. They may have originated in a Syrian or northern Italian workshop. Most of the examples are fragmentary.

The blank for the Locarno cup was made by a straightforward glassblowing process that left a tapered section of glass above the present rim, ending with a hole (see Figure 4.1, which shows a replica of the blank in my hand, as well as a drawing of the blank on the table). The various colors of enamel were painted on the cooled blank and allowed to dry thoroughly (see Figure 4.2; for a full description and illustration of the processes of making and applying enamels to a piece of glass, see Figures 2.1–2.3).

The decorated blank was very gradually reheated in a kiln to a point at which the glass would soon become distorted under its own weight if left unattended. The hot end of a blowpipe was coated with soft glass, but so as to leave open the hole in its end. The blank was then reattached to the blowpipe; the soft glass easily adhered to the edge of the hole.

Next, the blank was returned to the glassworking furnace (Fig. 4.2). Only a few seconds later, the “softer” (usually darker) colors undoubtedly began to shine, indicating that they had started to fuse. The “harder” (usually lighter) colors, such as yellow but often red too, were initially unaffected by the heat because of their more refractory nature. To prevent a sudden collapse, the partly fired blank was withdrawn from the furnace momentarily to allow the hottest enamels—and the glass on which they had been painted—to cool somewhat.

This procedure of alternately reheating the decorated blank in the furnace and cooling it outside the furnace was repeated several times. Eventually the hardest colors succumbed to the heat and fused. However, the softest colors unavoidably accumulated excess heat, and this caused the vessel wall to soften significantly. As these hottest areas began to collapse, the worker periodically blew gently into the blowpipe to slightly pressurize and thereby stabilize the vessel, thus retaining the desired form. When the firing process was judged to be complete, the decorated blank was broken free of the blowpipe and placed in the annealing oven to cool slowly.

Close inspection of numerous fragments from this group, together with extensive experimentation, supports this reconstructed process. The most refractory colors of historic enameled glass vessels often have a markedly granular quality, indicating minimal firing. Conversely, the softest colors frequently seem to be very thin, with an almost runny appearance. This indicates a great lack of uniformity in the intensity of the heat required to fully fire the different colors. Beneath the softest enamels, the inner surface of these vessels often feels slightly sunken in or slightly bulging. Together, these “symptoms” indicate that the firing process was executed by a highly skilled worker who successfully overcame formidable technical challenges.

Finally, the unwanted upper, tapered portion of the blank had to be removed in order to create the rim. This was accomplished by cracking it off, an extremely quick and easy technique that was invented shortly before the cup’s manufacture.¹³ In this process, a well-annealed blank is scratched with a sharp tool (a diamond or quartz chip works well) partway around its circumference at the desired location of the rim. The glass is then rapidly heated all around in a very narrow band. Today, a tiny flame is held in place as the object is rotated rapidly. In antiquity, workers may have used a tiny gather of soft glass that was stretched out and then quickly wrapped around the outside of the blank.

Cracking off is a well-understood and widely used technique in modern glass factories. Because glass has poor thermal conductivity, the exterior surface heats only superficially, placing great tension on the circumference in that narrow region. The scratch propagates a crack whose path is strictly limited to the zone of tension. The excess glass above the rim pops off with a pinging sound. The edges left on the vessel are razor-sharp, and they must be smoothed by rubbing them lightly with a soft stone.

Sometimes, the process of cracking off does not proceed entirely smoothly. When the wall thickness of a vessel is reasonably even all around, it is relatively free of such defects as cords or stones, and it has been adequately annealed to remove stresses in the glass, the worker has a good chance of creating a perfect rim. Frequently, however, the crack travels around the circumference in such a way that it misses its starting point and leaves an obvious flaw in the line. At worst, the crack can travel south, so to speak, and invade the vessel itself.

In the case of finishing an undecorated tumbler, this would waste perhaps three minutes of a glassblower’s time. But with an object such as the Locarno cup, an errant crack would ruin the piece at the very last moment of the entire manufacturing and decorating process. Perhaps as much to avoid such a mishap as to create a decorative feature, a cold-worked groove was cut just below the rim of the Locarno cup and many similar contemporaneous undecorated cups before the cracking-off process commenced. Such a groove can interrupt and halt a wayward crack and thus prevent the destruction of an otherwise excellent, nearly complete object. Whether or not the groove had a practical as well as aesthetic purpose, we can be certain that it was cut after the enamel was fired because numerous examples show it passing through fired decoration that, at its edge, is sometimes very slightly chipped (see Figure 3, far left side of the groove).

A second example of first-century Roman enameling is the blown glass amphoriskos from Kerch (Crimea, Ukraine) that is now in The State Hermitage Museum in St. Petersburg (Fig. 5). Its only close parallel is a considerably smaller jar in the form of an amphora that was found in Cyprus and is now in The Corning Museum of Glass. Since I have not handled the Hermitage object, it is impossible for me to reconstruct its manufacturing and decorating methods with great confidence. However, from the photograph, it is evident that the lower attachment points of the two handles sit atop the enameled decoration. Therefore, we can be confident that the glassworker made a blank without the handles, painted on the decoration, reheated the decorated blank, and added the handles after the enamels had been thoroughly fired. In general, this approach is identical to that employed in the creation of the Thutmose III jug (Figs. 2.1–2.8).

A close examination of the jar in The Corning Museum of Glass allows us to suggest a more detailed series of procedures. Unfortunately, this object is so badly weathered that it is difficult to gather much information about its manufacture. It appears—and feels—to have a pontil mark at its tip, and the handles seem to sit atop decoration at their attachment points (although this is very hard to confirm).

Because the rim of the Hermitage vessel clearly has a folded edge, it must have been held at the base at the conclusion of the glassblowing process. This would have been accomplished with a pontil or a clamplike holder that is sometimes called a snap.¹⁴ The vessel would have been held again during the firing of the enamel. This time, however, only a pontil could have been used, since gripping the lowest part of the vessel with a clamplike device would have badly damaged the decoration. Thus, it is likely that the Hermitage amphoriskos has two pontil marks that sit either near or atop each other. The first was produced by the manufacture of the blank, while the second was left from the firing of the enamel and the attachment of the handles. As we will see, this use of the pontil a second time for the firing of enamels on glass vessels became standard procedure beginning in the Middle Ages, and it continued through the 18th century (see Figures 11.1, 11.3, 15.1, and 15.3).

It seems likely that the Hermitage amphoriskos has at least one pontil mark on its base. However, if this is not the case, the manufacture and decoration of this object were probably similar to the process employed for the Locarno cup. A blank was made, leaving a small hole beyond the tapered shoulder that was similar to the formation seen in Figure 4.1. After the enamels were applied, the blank was reheated and reattached to a blowpipe, as shown in Figure 4.2. When the firing had been completed, the very hot but hardened vessel was broken free of the blowpipe, it was transferred to a clamplike holder, and the folded rim was created. At this point, with the enamels fired and the vessel cooled until it was stiff, a clamp could have been used without damaging either the vessel or its decoration. Finally, with the vessel still in the holder, the handles were added, completing the process.

Second and Third Centuries

The Daphne Ewer (Fig. 6.1 [55.1.86]), which was also found at Kerch, is now in the collection of The Corning Museum of Glass. This opaque white vessel, which is said to date from the late second to early third centuries, was blown. The surface is decorated with red and gray enamels, and it is gilded. In order to recognize that the decoration is not cold-painted, as it has often been described, but instead enameled, it is essential to confirm that the handle sits atop the decoration.

Upon close inspection with magnification and bright light, it is evident that the decoration disappears uninterrupted beneath the outer edges of the handle. There is also the unintended appearance of the end of one of the red bands on the vessel shoulder in a spot that was clearly meant to be undecorated (Fig. 6.2). In addition, it seems unlikely that the rosette on the vessel shoulder, which is now blocked by the handle, could have been so finely painted if the handle had been in place at the time of the painting (Fig. 6.3). These three features indicate that a blank was made without the handle. The decorator responsible for painting and gilding the blank carefully left bare a small area of the shoulder, anticipating the placement of the handle at the conclusion of the firing process. As is usual with handled vessels of this period, the lower attachment point was the first point of contact with the flattened, tooled gather of glass. Directing the molten glass toward the narrow gap in the decoration with the required speed would not have been easy, and we can now see that the glassworker, only slightly but tellingly, missed the mark. The handle was then attached to the rim and shaped, and the vessel was annealed.

 

The Daphne Ewer has enameled decoration on its somewhat concave base. There is no pontil mark. During the firing and the addition of the handle, the vessel would have been held by a clamplike device. From the start, it was probably intended that this object be held with a clamp, not a pontil. Where the sides of the object meet the base, it can be seen that the worker who made the blank carefully created a groove that could be tightly gripped with the holder. The decorator carefully placed the lowest bands high enough to allow them to avoid being marred by the clamp. Thus, the Daphne Ewer evinces careful planning and cooperation (assuming that the glass was worked and decorated by different craftsmen) in all stages of its creation, as well as meticulous execution.

A tall cobalt blue glass flute that is heavily decorated with painting and gold foil (Fig. 7) was found in Sedeinga (Sudan), and it is believed to have been made in the third century. This flute is one of two similar objects that were excavated at Sedeinga, and it is now in the Collezioni Egittologiche in Pisa. Its place of manufacture is a matter of conjecture. It may have been produced in Sudan or Alexandria. I have not examined this object, but careful observation by Dr. Robert H. Brill, research scientist at The Corning Museum of Glass, indicates that the gold was added as foil, and that some of the painted decoration is fired enamel while the rest is cold-painted.¹⁵

Photographs of the two flutes reveal a difference between the gold-foil decoration nearest the rim and that farther down the vessel body. They clearly show how both the gold and the enamels were fired. Near the rim, some of the gold appears to have been slightly broken up by vertically aligned “fissures.” Well below that area, most of the gold appears to be dense and unbroken. As we will see later, in the section of this article that discusses Renaissance Venetian glass, when a piece of glass with adhering gold is stretched, the decoration does not stretch, but instead it breaks up (see Figures 17.1–17.4). For this reason, we can be sure that the firing of these flutes involved returning the vessels to the furnace, after which the opening at the end of each vessel was slightly expanded or widened. This resulted in a slight breakup of the gold foil. Neither of these objects bears a pontil mark. The use of a clamplike device during the firing process is confirmed by the presence of a furnace-opened rim: the blanks could not have been made without one, and therefore the firing was undoubtedly accomplished using this tool.

Substantial heat is required both to soften the vessel for the thorough firing of the enamels and to ensure the permanence of the gold. My extensive experiments in replicating historic glassworking processes have demonstrated that gold fired onto glass at a temperature below the vessel’s softening point can, without much effort, be scraped off after the object has cooled. Thus, the high-temperature firing in the furnace accommodated both the gilded and enameled decoration on the Sedeinga flutes.¹⁶

Third and Fourth Centuries

The bottle with Apollo and Marsyas (Fig. 8.1 [78.1.1]), which is now in The Corning Museum of Glass, can be dated to the third or fourth century. It is painted in polychrome enamels, and it may have been made in Syria. This object exhibits two flaws that demonstrate technical challenges confronting any glassworker—then or now—engaged in enameling.

The first flaw, the shrinking and cracking of the decoration, also reveals one of the most easily detectable “symptoms” indicative of fired enamels in comparison with cold-painted ornament (Fig. 8.2). As we have noted, in creating a paintable substance, the particles of colored glass must be mixed with a liquid medium (Fig. 2.1–2.3). When gum arabic mixed with water is employed as that medium and the mixture contains an excess of gum arabic, the decoration (especially when it is thickly painted) has a tendency to shrink when it dries, leaving cracks or fissures in the enamel. Moreover, at the start of the firing process, as the gum arabic begins to burn away and the particles of enamel start to fuse, the decoration begins to shrink even further, resulting in additional widening of the fissures. Like the Locarno cup, this bottle was probably fired after it had been reattached to the blowpipe (Fig. 4.2). It has no pontil mark, and its rim was cold-worked. During the firing, it was probably somewhat inflated, widening the fissures still more. Because the edges of the fissures were more exposed to the intense heat than were the flat planes of the decoration, they softened sooner. These edges were quickly firepolished, and in the continuing heat that was required to complete the firing, they began to thicken and to form a bead. (These observations are supported by my experiments, in which trials that went wrong were as informative as those that did not.)

This thickening of the edges of cracks is symptomatic of fired decoration. Cracks that appear in cold-painted decoration will, of course, show no sign of smoothing by fire polishing or rounding at their edges. Similarly, any portion of the decoration that is raised well above the plane of the background will receive excess heat during firing, and points will tend to round and become bumps. On historic objects, it is especially helpful to examine these areas in order to determine whether the decoration was enameled or cold-painted.

The other flaw in the Apollo and Marsyas bottle is a pair of scratches that run horizontally between the two figures on the side of the bottle that is shown in Figure 8.1. The figure on the right also has a bad patch of scratches in the chest area. These were caused accidentally by rubbing the bottle on a rough surface after the enamels had been applied but before they had been fired. Unfired enamels are easily damaged, particularly when they are still wet, but also after they have dried. The damage seen here probably occurred after the bottle had been reheated and while it was lying on its side in the kiln as it was being reattached to the blowpipe. Had the damage occurred before the object was reheated, we might expect to see some evidence of retouching.

Flaws of this nature are common in enameled glass vessels. The damage often appears as a smudge. A typical example can be seen in the forward wing area of the bird at the top of the beaker shown in Figure 11.2. Because these objects had to be reattached to a pontil or blowpipe for firing (the Thutmose III jug still had its metal-rod handle attached at this point), they were probably placed on their side in the kiln to provide the glassworker with easy access to the base. However, this would have placed the decoration in a particularly vulnerable position. Today, kilns can be made with a door on the top, allowing the glassworker to access the base while the vessel sits on its rim. This arrangement would probably have been too difficult to construct before modern times.¹⁷

Byzantine Glass (10th and 11th Centuries)

The practice of painting enameled decoration onto glass vessels apparently lapsed again after the later Roman examples were made, this time for about 600 years, or until about the 10th to 11th centuries. During this time, however, the principle of fusing crushed colored glass to create decoration survived in at least one form. In early medieval Europe, enameling was used in the making of Celtic jewelry.¹⁸ Seventh-century examples found in Ireland have bronze cells atop copper-alloy structures that were filled with colored glass powder and then fired, creating a cloisonné effect. Transferring this technology to glass would not have been inordinately difficult.

The much-published bowl housed in the Treasury of San Marco, Venice, is the finest surviving example of Byzantine luxury glass (Fig. 9).¹⁹ It heralds the re-emergence of glass vessels with painted and fired decoration. This bowl was probably made in the 10th or 11th century and brought back to Venice by Crusaders following the sack of Constantinople in 1204. In addition to enameling, there is extensive gold-painted decoration and, as has only recently been noticed, silver staining. I have not been able to examine this object, but I have inspected a closely related vessel fragment and bracelet in The Metropolitan Museum of Art in New York City.²⁰ Both of these recently republished and reattributed pieces have silver stain. Thanks to their close association with the San Marco bowl, they shed considerable light on its probable method of decoration.

The fragment in the Metropolitan Museum consists of the lower wall and floor of a bowl or dish. Its style of decoration identifies it as a Byzantine object of about the 10th century. Like the San Marco bowl, this fragment has a very neatly made foot-ring and a pontil mark. Significantly, the mark on the fragment clearly appears to be a double pontil mark. Atop the first pontil mark, in the form of a small divot removed from the bottom of the vessel in the process of making the blank, there is a small amount of glass that was left behind after the pontil was attached a second time when the object was fired.

The bracelet, which is also thought to be Byzantine, is of a type that often includes gilding and enameling in addition to the silver stain that is seen on this example. From the visible presence of decoration that is trapped under an area of overlap where a join was made, as well as the presence of a small pontil mark, the sequence of events in the manufacture and decoration of the bracelet seems clear. A strip of blue glass was made and cooled. Next, silver stain was painted on. After the object had been gradually reheated, the decorated strip was picked up, at the middle of its decorated side, with a pontil whose tip had been coated with colorless glass. This is confirmed by a small pontil mark opposite the seam, which left traces of colorless glass and chipped away a little of the decoration. The strip was then reheated in a glassworking furnace and thoroughly softened. It was given its circular shape, perhaps by wrapping it around a mandrel, and its two ends were joined by overlapping them. Distortion in the decoration trapped between the two layers of blue glass indicates that manipulation (probably by squeezing with a tool) was employed to make a smooth join. During the intense heating that was required to soften the decorated strip, the silver stain fired.

From the double pontil mark on the vessel fragment and the trapped decoration in the overlap seam of the bracelet, it is clear that, in order to fire the silver stain, it was necessary to heat the decorated blanks until they softened. If a lower temperature had been adequate, the workers would surely have fired these pieces while they were in a kiln—and, in this case, the bracelet surely would already have been in its finished form. Compared with kiln firing, these firing procedures were more perilous for the objects and more difficult for the workers. They required great skill, and they consumed more fuel.

Because we now know that the decoration of the San Marco bowl includes silver stain as well as enamel, and because it is so closely related in origin to the Metropolitan Museum objects, we may conclude that this vessel, too, was probably fired in the glassworking furnace while it was reattached to a pontil. Although a published photograph of the bottom of the bowl appears to show a simple and tidy pontil mark, careful inspection with the eyes—and the fingers—would most likely reveal a double pontil mark.

Thus we can safely assume that the firing process for the San Marco bowl is consistent with that which we have seen for all earlier enameled glass vessels in this study namely, a high-fire process.

Islamic Glass

What might be termed “painterly” enameling on glass vessels (almost always combined with gilding) began in the Islamic world in the late 12th century, perhaps in Raqqa (Syria). The origin of enameling in Islamic workshops is unknown.²¹ There may have been a transfer of information from Byzantine workshops, or perhaps the technique was reinvented once again.

Before this time, in the 10th and 11th centuries, there had been a limited production of gold sandwich glass. Some surviving examples include a scattering of blue dots that may refer to some Koran manuscripts in which colored dots represent vowels or diacritical marks. The style of enameling on these objects is markedly more “primitive” than that which would flourish, with great virtuosity, a century or so later.

The 13th and 14th centuries are often referred to as the golden age of Islamic glass, when the making of enameled and gilded glass reached its peak. While mosque lamps are by no means the only form of this glass, they are certainly the best known. A wide variety of objects—including beakers, bowls, flasks, bottles, vases, and candlesticks—was made. Both enameling and gilding continued to reappear frequently in Islamic glass during the succeeding centuries, but there were no discernible technical advances beyond the brilliant medieval achievements until after the 18th century.

Ninth and 10th Centuries

The British Museum’s bottle of gold sandwich glass with blue enamel dots (Fig. 10) is important in the history of enameling primarily because it is the precursor of something that would be considerably more developed and grand. On close inspection, the dots appear to show vestiges of a somewhat granular stage, which strongly indicates enameling as opposed to, say, the addition of small pieces of colored canes or tiny bits of colored glass.

The glassworker began with a blown glass blank made in the form of a bottle.²² After the blank was cooled, it was decorated with gold leaf, and then the enamel dots were added (see Figures 2.1–2.3). The decorated blank was reheated, reattached to a blowpipe, and resoftened in the glassworking furnace. By this time, the enamel dots had been fired. Next, the glassworker carefully lowered the bulbous part of the soft blank into a previously made but still very hot glass cup. The lower part of the blank was further inflated, fusing the two parts together. Judging from the slight degree to which the gold decoration broke up, with radial (vertical) fissures in this and similar objects, we can see that the decorated inner part must have fit rather closely inside the cup. After the vessel was reheated, it was marvered and tooled into its final shape. Because the rim was to be finished by cold working, there was no need to transfer the vessel to a pontil. The completed vessel was then broken free from the blowpipe and annealed.

13th and 14th Centuries

The three objects that will be discussed in this section provide a clear body of evidence for the process by which beakers were decorated in certain Islamic glass workshops in the second half of the 13th century. They are a monumental beaker in the Calouste Gulbenkian Museum in Lisbon (Fig. 11.1), a beaker in the Hessisches Landesmuseum in Kassel (Fig. 12), and fragments of a beaker in the al-Sabah Collection, Dār al-Āthār al-Islāmiyyah, Kuwait National Museum (Fig. 13).

The upper portions of all three beakers were slightly flared during the firing process, thereby slightly stretching the decoration horizontally. On the Gulbenkian beaker (Fig. 11.2), the small red enamel dots that are seen on the birds near the middle of the vessel are round. Closer to the rim, however, the dots are slightly elongated and oval because the diameter of this part of the object was increased. This is also true of the Kassel beaker. Here, however, the “distortion” during firing contributed positively to the decoration: the wings and bodies of the birds were thus given a heightened and dynamic wind-swept quality.

The degree to which the diameter of one of these beakers was increased can be estimated fairly closely. It can be argued, of course, that the dots on the Gulbenkian beaker were painted rather casually, and that not all of them were precisely round before the object was fired. This would call into question their usefulness as a reliable measurement standard. However, the eyes of the birds on the Kuwait fragments (Fig. 13) are impressive in the precision of their initial round shape. As we would now expect, eyes near the rim are markedly oval compared with those farther down the side of the beaker. What is particularly helpful about the Kuwait fragments is the evidence they offer of the way in which the eyes were created.

The lower center fragment shows that the eyes were made, not by leaving a small area bare during the painting, but by clearing away wet enamel immediately after the painting, using a round and perhaps somewhat pointed tool. This tool was probably turned as it rested firmly against the wall of the vessel, pressing into the enamel. This resulted in the displacement of a small amount of enamel around the bare spot. It also helped to produce a perfectly round eye. This perhaps inadvertent precision leaves us with a fine standard by which to measure the distortion created during the firing process.

A double pontil mark was required on the base of objects that exhibit these features. The occasional Roman practice of employing a Clamplike device (snap) to hold an object during reheating was apparently unknown. One scar was left in the making of the blank, while the second one, positioned atop the other, was produced when the vessel was reattached during the firing process. The double pontil mark of the Gulbenkian beaker is fairly easy to distinguish in a photograph (Fig. 11.3).

From a close inspection of several medieval Islamic beakers, it seems safe to generalize that a slight further flaring of the vessels was an integral part of their firing. Because the glass softened and began to flex under its own weight while it was rotated on the pontil inside the glass furnace, the glassworker found it much easier to maintain the form of the object by gently expanding the upper portion with a tool when the glass as periodically removed from the furnace.

The mosque lamp in the Calouste Gulbenkian Museum (Fig. 14.1) is representative of many such 13th- and 14th-century lamps that have been closely examined. The evidence of this lamp’s decorating process, including its double pontil mark (Fig. 14.2), is identical to that seen on the beakers, with the exception of one important feature: there is no sign of distortion in the decoration near the rim. We can therefore be sure that when the glass was at its softest during its firing, the glassworker did not widen the opening at the upper end of the vessel. Although the Gulbenkian lamp is fairly straight, many mosque lamps are noticeably crooked.

Broadly speaking, this crookedness and asymmetry can also be observed in other non-beaker forms of medieval enameled and gilded Islamic glass. At the “Glass of the Sultans” exhibition, which was held at The Corning Museum of Glass, The Metropolitan Museum of Art, and the Benaki Museum in Athens in 2001 and 2002, it was possible to see, in one room, arguably the most comprehensive display of such objects ever assembled. Well after the shock of beholding their beauty had subsided, viewers would have been able to note the marked crookedness of nearly every one of these objects. Had beakers been represented in this show (only the exceptional and immense Gulbenkian beaker was included), they would have stood in stark contrast because of their relative straightness and symmetry. It is difficult to imagine that a glassworker who had fired decoration on beakers would have failed to employ the same strategy for keeping mosque lamps, vases, and other items straight when they were fired if he had been aware of that decidedly superior procedure.

This subtle yet, practically speaking, enormous difference invites speculation. The evidence of two similar yet distinctly different methods of firing gilded and enameled objects of two separate types suggests two different working traditions by two distinct groups of workers. At some point, one of these parallel activities ceased in favor of the other, or perhaps they were merged. As we will see, in the 13th century, the Venetian firing practice was identical to that used by the makers of Islamic beakers. Later, in the 15th century, Venetian workers would pursue that same strategy to a rather extreme degree (see Figures 19–24). By contrast, the process by which Islamic workers fired objects other than beakers seems, justifiably, to have been eventually abandoned because it resulted in a markedly inferior degree of control and invited asymmetry.

European Glass (13th–19th Centuries)

Finally, after centuries of isolated beginnings and endings, rediscoveries and losses, Venetian workers in the mid-15th century commenced a tradition of enameling on glass vessels that would become widely disseminated in other European glassworking cultures—and continue to be practiced, without interruption, to the present day.

The precise manner in which enameling arrived in Venice is open to speculation.²³ Sophisticated enameling had been rendered on perilously thin vessels there during the late 13th and early 14th centuries, but without surviving examples datable between that period and the 15th-century production, we must assume that enameling in Venice somehow ceased. While it has long been assumed that the Venetian enameling practice of the 15th century was based on late medieval Islamic techniques, it is perhaps equally plausible that the technical foundation was Byzantine in origin. Indeed, as was mentioned earlier, an argument can be made that the forerunner of the Islamic enameling process was the Byzantine practice of the 10th and 11th centuries.

In any case, probably through the pan-European dispersal of Venetian glass objects and workers, the processes of enameling on glass vessels spread to the Low Countries, Spain, France, Germany (the Holy Roman Empire), England, and beyond during the 16th century. For about three centuries, as we will see, this technology was unchanged in that enameling required that objects be reattached to the pontil for firing. However, this would change in the 19th century, when a revolutionary development in the making of the enamels themselves would allow methods to be established that continue to be employed today namely, low-fire enamels capable of being fired in a kiln.

Venice (13th and 14th Centuries)

The Aldrevandin Beaker (Fig. 15.1), which has been in The British Museum since 1876, has been the iconic head of a group of more or less similar objects since the early 20th century.²⁴ Characterized in part by enamel painting on both the outer and inner surfaces, this beaker gets its name from the inscription in its frieze, which reads, “magister aldrevandin me feci(t).” The name is thought to be that of the decorator. This beaker may be the most widely known and published medieval European glass vessel. Art-historical considerations aside, the object’s importance is entirely justifiable strictly on technological grounds for the clarity with which it demonstrates the complete constellation of “symptoms” of the enameling process as practiced in Venice during the late 13th and early 14th centuries.

The Aldrevandin Beaker (H. 13 cm) is imposing. It is not a small object, and because of the sheer density of its decoration, it has considerable visual “gravity.” When it is held, however, it is shockingly lightweight. The accomplishment of the glassblower who made the blank is, by itself, noteworthy.²⁵ In most areas, the wall is scarcely more than a millimeter thick. The process of glassblowing becomes exponentially more difficult as the wall thickness of a vessel diminishes.

During the firing process, as the decorated blank reattached to the pontil was held in the glass furnace, the vessel became soft. This is evident from the easily felt localized distortion of the glass on the side of the wall behind the enameled decoration. For example, the letters of the frieze can easily be felt in reverse by gently rubbing a finger on the inside of the beaker. This is because, as the blank was heated, the white enamel protected the glass from direct infrared exposure, causing the glass around the letters to heat more rapidly and thicken slightly in comparison with the glass that was directly beneath the white enamel. The result is a slight but detectable variation in the wall thickness of the beaker. Conversely, the black enamel used in one of the armorial shields and the glass on which it sat thickened more than adjacent areas because they heated more quickly and remained soft longer.

A slip of the hand by the decorator, although it did not in any way compromise the excellent result achieved, fortuitously provides a means of measuring the degree to which the shape of the vessel was changed during the firing. A small amount of yellow paint, accidentally spattered in a patch on the blank, is now visible with only slight magnification (Fig. 15.2). Judging from the considerable number of dots, as well as their thinness and the fairly large area they cover, we can see that the paint was fairly “runny” (or low in viscosity), and it seems safe to assume that, at the moment the spatter occurred, the dots were perfectly round. Therefore, we can interpret their distortion as an accurate indicator of where and how much the glass beneath them moved as the object was being fired (see Figure 17).

Dots well below the frieze are still round, while dots nearer the rim appear to be progressively more oval. Thus, during the firing, the beaker was flared a bit and the vessel’s diameter in the vicinity of the frieze was slightly increased. This expanding of the soft vessel allowed the glassworker to keep the object completely under his control, thereby avoiding a crooked appearance and even a collapse. That the reheating and reshaping took place with the beaker reattached to the pontil is confirmed by the vessel’s exceptionally clear double pontil mark (Fig. 15.3).

Beakers of the Aldrevandin group are often divided into two categories: relatively low and wide ones, such as the Aldrevandin Beaker, or narrower and more flared ones, such as the example in the Museum für Kunsthandwerk in Frankfurt am Main (Fig. 16). On the Frankfurt beaker, the two rows of dots (conveniently placed at the very top and bottom of the object) have been painted far too casually to make them useful as a precise reference, unlike the accidental yellow dots on the Aldrevandin Beaker. However, photographs show that some dots in the bottom row are round and all of the dots in the top row are oval, so it seems safe to assume that this beaker, like the Aldrevandin itself, was somewhat expanded during the firing process.

From close and extensive hands-on inspections of the Aldrevandin Beaker and its close parallels, the numerous Foster Lane fragments found in London in 1982, and close observation of the Frankfurt and Chur (Switzerland) beakers in museum displays, a consistent picture emerges. In the late 13th and early 14th centuries, enamels in Venice were fired in a furnace while the decorated blank was reattached to a pontil. During this process, the diameter of the upper third or so of the vessel was slightly increased. Based on my experiments aimed at re-creating the processes used to make the blank and to decorate the Aldrevandin Beaker, it seems safe to say that the same level of skill was required for both of these activities. Therefore, theories about the origin of these beakers that are based on the assumption that the blanks could have been decorated and fired “on location” for specific customers by workers who were skilled only in the painting process seem to be doubtful, given the technical evidence to the contrary.

Venice (15th–17th Centuries)

With the reappearance of enameling on glass vessels in Venice during the 15th century, a system of procedures was developed and standardized by which the makers of these objects apparently sought to control their shape during the firing process to an unprecedented degree. This was accomplished by leaving the blanks in an unfinished form (they often looked markedly different than when they were completed), decorating them, and then resuming and completing their shaping when they were fired. This approach must have required exceptionally close cooperation between the glassblowers and the decorators. Its effectiveness is attested by the fact that we fairly rarely encounter significant crookedness in 15th- and 16th-century Venetian enameled glass objects.

The shape of the unfired blank can be determined rather accurately by closely observing both where and how much the decoration appears to have stretched on an object when it was fired. This can be seen in the distortion of enameled decoration and alterations in the appearance of gold-leaf decoration (Fig. 17). The shapes of enamel dots range from approximately round when they were painted to oval or elongated after firing, depending on how much they were stretched. Stretching can sometimes be observed in repeated enameled motifs such as foliate patterns, and in representational painting. The faces of putti painted near the rim of a goblet, for example, may be broadly distorted because the vessel expanded during the firing process.

As was noted with the Sedeinga flute (Fig. 7), gold visibly breaks up when the glass on which it sits is stretched. The orientation of the fissures, which is always perpendicular to the stretching, indicates the exact direction of that movement. Other useful indicators are variations in the width of scratched decorative lines such as fishscale motifs and lettering in gold leaf, which were common during this period. These are particularly helpful in estimating the magnitude of the stretching. Scratches aligned with the direction of the stretching (such as the circumferentially oriented scratches near the rim of a vessel) become narrower, while those aligned perpendicularly to the stretching become wider.

Fifteenth- and 16th-century Venetian and Venetian-style enameled and gilded glass vessels can exhibit a considerable contrast between the shapes of the blanks and those of the finished, decorated objects. Some display a slight degree of transformation (Fig. 19), but at the other extreme, some blanks bear almost no resemblance to their completed forms (Fig. 23).

Experiments that attempt to reconstruct the historical processes help to explain the evidence provided by the original objects and help to direct future investigation. Figures 19–24 include three well-known 15th- and 16th-century Venetian glass types, followed by illustrations of my re-creation of their firing processes. Three degrees of transformation are shown: slight (Fig. 19), substantial (Fig. 21), and extreme (Fig. 23). The table of drawings contrasting shapes of blanks and of finished objects (Fig. 18) illustrates the various degrees of transformation, ranging from none (nos. 1 and 2) to extreme (nos. 6–8).

Some goblets required two attachments of a pontil at the time of firing (Fig. 18, no.7). The first pontil was attached to the floor of the inside of the bowl, which allowed the foot to be fired. When that step had been completed, a different pontil was attached to the object (atop the pontil mark that had been left from the making of the blank, thus creating a double pontil mark). Immediately thereafter, the pontil attached to the inside of the bowl was very carefully removed, thus allowing the upper portion of the goblet to be fired. This perilous procedure seems to have been employed only for particularly tall objects and/or those with exceptionally elaborate decoration on the foot.

Some types of objects, including pilgrim flasks (Fig. 18, no.1) and conical footed beakers (Figs. 18, no. 2, and 28), have decoration that shows no sign of stretching. However, as we will see just below, there are other pieces of evidence showing that these vessels were undoubtedly fired in the furnace while reattached to a pontil.

The best and most easily observed evidence of furnace firing is often the pontil mark itself. When the process of making the blank and the firing process were completed, the pontil was tapped firmly to free the object (Fig. 25), and each time, this left a scar. Many 15th- and 16thcentury Venetian and Venetian-style objects have an easily identifiable second pontil mark that sits atop the first. The first separation often leaves a divot in the bottom of the vessel, while the second produces an unshapely mass of glass that rests on top of the divot, visibly off-center. A nuptial goblet in The Corning Museum of Glass (Fig. 26.1 [79.3.170]) has a double pontil mark, revealing the positions in which the object was held during both “breakoffs.” The straight and parallel set of fracture lines from the first pontil mark, which occupies the upper quarter of the figure, is aligned between about the 11-o’clock and one-o’clock positions. The second mark (downward in the photograph) sits atop the first, slightly off-center. Its fracture lines occupy the left quarter or so of the illustration, and they are aligned between about the eight-o’clock and 10-o’clock positions. This shows that when the firing was finished, the worker was holding the vessel 90 degrees in rotation compared with its orientation when the completed blank was broken free of its pontil. Seldom do pontil marks reveal so much!

Two additional symptoms of a furnace-firing process can be observed in some glasses. The first is the previously described phenomenon of localized thickening of the glass under or around the enameled decoration. As was already mentioned, if this thickening is significant, it can easily be felt on the inner surface of a vessel that has enameled decoration on its outer surface. Sometimes, in carefully positioned raking light, the undulation of the surface can also be seen (Fig. 27 [54.3.244]).

The second and more rarely seen symptom also results from localized thickening of the vessel wall, but it manifests itself differently. Where large areas of soft, dark enamel (such as blue) are painted fairly close to the top of a vessel, the substantial thickening of the glass pulls the rim downward. A visibly uneven rim results (Fig. 28 [79.3.191]).

If there is still any doubt that 15th- and 16th-century Venetian glassworkers fired their enameled vessels in the furnace while they were reattached to a pontil, and that they sometimes made their blanks in a highly unfinished state, a sizable group of ewers whose handles and spouts are obviously attached atop the decoration should put it to rest. A lavishly enameled and gilded ewer in The J. Paul Getty Museum (Fig. 29) conveniently has a colorless glass handle that can clearly be seen to have been placed, at both the upper and lower attachment points, atop the fired decoration. This proves that the handle was not attached when the blank was decorated.

We can see that the spout, too, was a late addition. It was made by attaching a large gather of colorless glass to the side of the vessel. Next, the gathering iron, the metal rod with which the added glass had been collected from the furnace, was pushed through the now softened wall of the vessel. The rod was then immediately pulled outward and upward, giving the spout its present shape. Finally, the tip of the spout was broken free from the gathering iron, leaving a hole at its end. From the long colored striations in the spout, it can be seen that when the spout was formed, the enameled and gold decoration was also resoftened and drawn outward with the colorless mass of molten glass. No enamel dots in Venetian Renaissance glass were ever stretched longer than these!

This surprisingly complicated procedure was necessary because, had the handle and spout been in place during the firing of the enamels, the localized additional pull they would have exerted with centrifugal force would have badly distorted the form of the vessel’s body.

The Low Countries (16th Century)

A tazza in The British Museum (Fig. 30) and the closely related Margaret of Austria roundel in The Corning Museum of Glass (Fig. 31 [81.3.43]) are exceptional examples of Renaissance gold sandwich glass. Both of these objects were probably made in the Low Countries during the 16th century. They are so closely similar, in both technique and size, to fourth-century Roman roundels as to suggest that the former were made in imitation of the latter. The tazza is notable for the complexity of its construction and for the evidence it reveals of an accident that nearly ruined it during its manufacture. In addition, by fully understanding how the tazza was made, we can be more certain of the origin of the Margaret of Austria roundel.

The manufacture of the tazza began with the blowing of a disk that was about nine centimeters in diameter and a little more than two millimeters thick. The disk was decorated on the side opposite the pontil mark. Then, with the decorated side down, it was gradually reheated in a kiln to a temperature somewhat above the annealing point of the glass (about 950°F or 510°C for a modern soda-lime glass). By the time the decorated disk was hot enough to be reattached to the pontil at the location of the first pontil mark, another glassblower had made, on a blowpipe, an ovoid bubble of glass a little larger in diameter than the disk itself. Unfortunately, in the reheating of the disk, its reattachment to a pontil, or its return to the furnace in order to fire the enamels, it cracked near its center. Fortunately, the crack extended only a few centimeters. It could easily have split the disk into two pieces, rendering it unusable. The crack avoided the thicker—and stronger—center area, thus forming a line that curved slightly. When the disk was reheated to the point of softening in order to fire the decoration, the decision was apparently made to continue the work despite the flaw.

Next, while the disk and the end of the ovoid bubble were soft, they were carefully pressed together, thereby trapping the decoration between the two layers of glass. The pontil was then broken free, leaving the disk attached to and nearly covering the entire end of the bubble. Where the decoration stood highest above the disk, complete contact between the two surfaces was prevented and some pockets of air were caught between them (Fig. 31.2).

During the reheating, tooling, and slight inflation to fully smooth and shape the glasses, the crack opened slightly. Its edges, especially at the ends, rounded. When a bit of molten glass was added to the tip of the bubble and tooled to create a merese, air was trapped along a tiny crevice that was itself a vestige of the crack (Fig. 30.2).

When the highly unusual process for creating the trapped decoration in the upper part of the goblet had been completed, work proceeded in the customary manner to make the goblet.²⁶ A merese was added to the center of the disk, and the stem and foot were added atop the merese. A pontil was attached to the foot, the vessel was broken off the blowpipe, and the top of the goblet was given its final shape. The finished object was then removed from the pontil (leaving a single pontil mark, of course) and annealed.

From a close inspection of the British Museum tazza and our knowledge of how it was probably made, it seems clear that the Margaret of Austria roundel was once the featured decorative element in a similar tazza that was subsequently damaged. The outer edge (visible in Figure 31.1), which is usually mounted in a turned-wood frame that hides it, is rough. This resulted when the outer, undecorated part of the bowl was removed by clipping it away with a pliers-like tool, a commonly used flat-glass technique known as grozing.

The back surface of the roundel has an “excavation” about two millimeters deep and the approximate diameter of the upper merese on the British Museum tazza, which was carefully created by cold working (Fig. 31.2). This cavity was left when the merese was removed.

It is curious that the sophisticated cold-working process used to remove the merese was not also employed to more safely and attractively cut away the unwanted glass from the edge of the roundel. Perhaps the grozed edge dates from the time of the accident and the cold-worked excavation (which was probably accomplished by a skilled copper-wheel engraver) took place at a later time, possibly in a different location. The merese could have been more carefully removed because, by this time, the disk had likely achieved the status of an heirloom and was being prepared for a suitable mount. In any case, the Margaret of Austria roundel is among the finest examples of enameled decoration on a glass vessel of any period.

Spain and France (16th Century)

The enameling procedures employed in Spanish and French workshops of the 16th century seem to have been essentially the same as those used in Venice. If any difference can be distinguished, it may be safe to say that objects made later than about 1525 seldom reveal evidence of substantial expansion during the firing process. Thus, during the 16th century, there seems to have been a general trend away from making blanks in shapes that were expected to be altered substantially when they were fired.

A late 15th-century goblet in The Corning Museum of Glass (Fig. 32 [63.3.37]), which was probably made in Catalonia, has a double pontil mark on the inside of the foot and bears evidence of limited stretching of the dots near the vessel’s rim. Although the foot is extensively decorated, there is no pontil mark in the floor, as we might perhaps expect (see Figure 18, no. 7). Therefore, no stretching of the decoration is present on the foot.

Two French drinking glasses in the Corning Museum (Fig. 33), dating from the mid-16^th century, indicate that their decoration was very slightly stretched during firing, and both have double pontil marks. The later example (Fig. 33, right) is most interesting because of the way in which the blank was made.

Having blown a fairly thin bubble of glass, the worker made a narrow hourglass-like constriction at its center. The tip of the outer bulb was reheated. When its outer half was thoroughly resoftened, the glassblower sucked air out of the blowpipe, causing the soft glass to move upward to the point where it was halted by the constriction. This resulted in a double layer of glass beyond the constriction. After reheating, the doubly thick glass was tooled to form the knop and the tall, spreading foot. The vessel was broken free of the blowpipe and transferred to the pontil, and the bowl of the goblet was given its final shape.

It can be seen in the photograph that attaching the pontil to such a tall, hollow foot cannot have been easy. The second pontil, which was used for firing the enamels, was probably allowed to inadvertently touch—and adhere to— the foot in a place much lower than was intended, producing a bit of unwanted glass.

Germany (16th and 17th Centuries)

The longstanding and lucrative Venetian practice of making and selling glass in styles that were carefully tailored to suit foreign tastes is well known. It is not surprising, therefore, that enameled and gilded glass was ordered, custom-made, and exported to Germany and other parts of Europe during the late 15th and early 16th centuries. Some of these glasses were decorated with the arms of patrician families from Nuremberg, Augsburg, Liechtenstein, and Alsace. Enameled and gilded glass seems to have become so popular in Germany and Bohemia during the late 16th century that, even though it was no longer fashionable in Venice, Venetian glassmakers continued to make it for wealthy northern European customers. By the last quarter of that century, enameled glass was being produced in local workshops where the craft had been learned and mastered.

Undoubtedly, the best-known form of German enameled drinking glasses is the Humpen (Fig. 34.1 [57.3.81]). The earliest example, an object in The British Museum, is dated 1571. The inherently strong shape of a cylinder resting on a small foot might be expected to withstand a brief period of heating sufficient to soften it without deformation, but this does not seem to have been the way these objects were fired. Of some 25 closely studied Humpen, most offer clear evidence of having been fired on a pontil.²⁷ Many of them have double pontil marks (Fig. 34.2), as well as distortion of the inside wall under or around the enameled decoration that can be felt or seen in raking light. On the illustrated example, from the Corning Museum collection, and on many other Humpen, the gold band near the top looks—and feels—constricted in comparison with the vessel wall just above and below it. This was caused because the gold insulated the glass during the firing process, allowing excessive heat to accumulate beneath it. Since the glass below the gold became hotter and remained hot for a longer time than the glass around it, the force of surface tension during this period of lowered viscosity caused the glass to thicken and shrink, thereby forming the constriction. With this amount of softening, the vessel would certainly have collapsed under its own weight well before this gold-band phenomenon occurred if the firing had taken place while the object was sitting in a kiln.

Similarly, on a much later (1662) beaker with filigrana (filigree) decoration in The Corning Museum of Glass (Fig. 35), the vessel wall sank inward beneath the thickest enamel and the gold leaf. This demonstrates that, during the firing, the entire object was significantly softened and thus the enamels and the gold must have been fired in the furnace after the object had been reattached to a pontil. The Corning Museum also has two dated beakers of the late 17th century that bear double pontil marks and a slightly detectable distortion of their inner surface, revealing that they were fired using the same process (Fig. 36 [Group image 79.3.311]).

Because of its thinness, it might be expected that Schwarzlot decoration (Fig. 37.1 [79.3.253])—lightly painted enamel that is usually black, gray, or rust red—could be fired while the object sat in a kiln. However, the late 17th- and early 18th-century objects ornamented in this manner were evidently furnace-fired. While the enameling is much too thin to have caused any detectable change in the wall thickness of the vessels, many examples have double pontil marks (Fig. 37.2). This evidence of high-temperature firing is consistent with what we see in contemporaneous glass panels decorated with grisaille, which is essentially identical to Schwarzlot. Their undecorated side is sometimes noticeably roughened by contact with the uneven surface on which it lay softened during firing.

A large group of wineglasses, usually called “peasant” glasses, often bear 18th-century dates and are decorated with brightly colored enamels. A representative selection from The Corning Museum of Glass (Fig. 38 [79.3.734]) is thought to have been made in Switzerland, Germany, or Bohemia. These objects show clear evidence, such as double pontil marks and distortion of the vessel wall, that even at this late date enamels were still routinely fired in the furnace while the vessel was reattached to a pontil.²⁸

Venice and England (18th Century)

As noted in the discussion of 16th-century German enameled glass, the Venetian glass industry was ever ready to supply a niche market with custom-made wares. In the 17^th and 18th centuries, a growing European demand for fine porcelain imported from China (which was both expensive and in short supply) was partly satisfied by credible imitations made of glass (Fig. 39 [51.3.316]). Although porcelain was made in the area of Venice as early as the 1720s, we know that the Miotti glass factory was making various imitations of it during the ensuing two decades. While it is not surprising to find clear evidence that 15th- and 16th-century Venetian opaque white enameled glass objects (which were also inspired by imported Chinese porcelain) were fired in the furnace while reattached to a pontil, we might expect that by the relatively late date of the Miotti examples, the technology of enameling would have progressed to the point that kiln firing was possible. But this was not the case; both tumblers shown in Figure 39 have double pontil marks that are easily identified.

Imitations of porcelain became an important part of production at many European glasshouses during the 18th century. In addition to Venice, factories in Bohemia, Germany, Spain, and France worked to satisfy what must have been a considerable demand. English factories in London and southern Staffordshire made highly credible imitations of porcelain in opaque white glass that was impressive in its chalklike opacity (Fig. 40.1 [86.2.16]). The decoration on these objects, like those made in Venice, shows no sign of having been stretched, but most of the glasses have an easily seen double pontil mark (Fig. 40.2), confirming a furnace-firing procedure.

Mid- to late 18th-century English stemware decorated with enamels (Fig. 41.1 [51.2.192, 50.2.24, 79.2.32, 51.2.205, 79.2.67, 50.2.9, and 50.2.69]) and, in some cases, gold (Fig. 41.1, third from left) usually displays all of the symptoms of furnace firing, with the exception of stretching. Perhaps because this glass has a significant lead content—unlike all of the other glass in this study—the sites of the pontil attachments often have a badly splintered appearance, shrouding any clear evidence of a double pontil mark. Still, we can be sure that the firing took place in a furnace with the goblet reattached to a pontil because distortions on the inner surface of the glass can be felt under some of the enameled and gold decoration, indicating that the vessel wall softened significantly during the firing. Despite the splintering phenomenon, some 18th-century English enameled wineglasses have a second pontil mark that is fairly easy to discern (Fig. 41.2).

Europe (19th Century)

During the 19th century, thanks to a growing sophistication in the formulation and manufacture of enamels, a great advance took place that made firing them on glass vessels an infinitely easier process than ever before. This advance was the development of low-fire enamels. They were carefully made to soften at a temperature below that at which the vessel would have started to become distorted from the force of gravity during the firing process. This meant that, for the first time in the then 3,400-year history of enameling glass vessels, firing could occur while the objects sat in a kiln.

In terms of their chemistry, low-fire enamels are a tall order for a glass technologist, as we would term such a specialist today. Three essential criteria must be perfectly met in order to ensure success. First, the color must be correct and of great intensity because enamels are often painted thinly. Second, the enamel must be completely compatible with the glass from which the vessel was made. That is, the two materials must gain and lose volume at about the same rate and to about the same degree as they cool from a temperature at which they are slightly soft down to room temperature. If they are incompatible, they will crack apart. Third, the enamel must be capable of being fired at a temperature well below the point at which the vessel might become distorted under its own weight during the firing process.

Practically speaking, the development of lowfire enamels meant that essentially no manual skill was needed for firing, only knowledge of how to operate the kiln. This was a vast change from the days of the Thutmose III jug, the Daphne Ewer, the Aldrevandin Beaker, and especially the 15th- and 16th-century Venetian objects, for which the skill required to fire the decoration was equal to that needed to produce the blank. In addition to making it easier to create complex one-of-a-kind objects, the new firing process facilitated the decoration of large numbers of identical objects. This was essential for the creation of matching sets and for the attainment of a mass-production capability.

Figure 42 shows a selection of 19th-century enameled objects made in various countries. Each of these vessels imitates, more or less closely, a well-known historical object or style. While the overall visual effect of the enameled (and gold) decoration is very similar to that seen in its precursors, the makers of these vessels used the new low-fire enamels and fired the objects while they sat in a kiln. Thus, upon close inspection, numerous details are different. There are no double pontil marks, there is no evidence of decoration altered by stretching during the firing process, and there is no sign—either visual or tactile—that the wall became soft when the enamels were fired. Where and when low-fire enamels began to be made, and how the new technology spread, will (as much as can be determined) be the subject of a separate study.

Conclusions

In applying the information in this article, it is essential to accurately collect the evidence that an object reveals about the processes that were used to decorate it. For ease of reference, two checklists follow.

Furnace firing of the enamels is indicated if one or more of the following symptoms is present:

1. A double pontil mark (Figs. 1.3, 14.2, 15.3, and 26.2).
2. Evidence that the decoration has been stretched (Figs. 8.2, 11.2, 12, 13, 15.2, and 23).
3. Distortion of the surface of the object below the decoration (Fig. 27).
4. An unevenness of the rim that corresponds to the presence of large areas of decoration nearby (Fig. 28).
5. Additions such as decorative wraps, handles, spouts, and prunts that sit atop the decoration (Figs. 5, 5, and 29).

Kiln firing of the enamels is indicated by the complete absence of the symptoms of furnace firing, as described above. That is:

1. A single pontil mark (or no pontil mark).
2. No evidence that the decoration has been stretched.
3. No distortion of the surface of the object below the decoration.
4. An even rim despite heavy areas of enamel (or gold) decoration nearby.
5. Decoration appearing on additions such as handles and spouts, indicating that they were in place at the time the decoration was applied.

The conclusions that can be drawn from a careful assessment of these symptoms are limited but useful. If the method of firing can be determined with certainty to have been the kilnfiring process, the object can safely be said to have been made in the 19th century or later. On the other hand, it is important to understand that a complete display of the symptoms of the furnace-firing process does not necessarily mean that the object was manufactured before the 19th century. This is because undeniable symptoms of furnace firing can clearly be seen on objects that are thought to have been made during and after the 19th century.²⁹ This process was probably employed very rarely in that later period, and it may, on occasion, have been intended to mislead.

The ability to detect the distinct differences between these two categories of enamels (highfire and low-fire) and their respective methods of firing (in the furnace and in the kiln) is essential for the historian.³⁰ When this information is used—together with that compiled from the traditional scientific, archeological, art-historical, typological, and archival disciplines––it may afford researchers a significantly deeper understanding of these objects and further illuminate their place in the history of glass.

---

This article was published in the Journal of Glass Studies, Vol. 48 (2006), 23–70. Due to copyright restrictions some Figure images have been removed. Please refer to the original printed article to see all Figure images.

---

*This article is respectfully dedicated to the late Hugh Tait, former deputy keeper of the Department of Medieval and Later Antiquities in The British Museum, an eminent colleague and close friend. Between 1982 and 2003, he kindly helped with my studies of many of the objects mentioned herein.

Acknowledgments. I thank the following colleagues at The Corning Museum of Glass for their assistance in the preparation of this article: Dr. David Whitehouse, executive director; Amy Schwartz, deputy director for education programs and The Studio; Dr. Dedo von Kerssenbrock-Krosigk, curator of European glass; Richard Price, head of the Publications Department; Jill Thomas-Clark, rights and reproductions manager; Nicholas Williams and Andrew Fortune of the Photographic Department; and Jacolyn Saunders, publications specialist. I am also grateful to Prof. Ian Freestone of Cardiff University, formerly research scientist at The British Museum. Steven Barall, an independent photographer in New York City, patiently and expertly photographed all of the decorating and furnace-working procedures.

1. For a useful historical overview, with numerous quotations from various periods discussed, see R. J. Charleston, “Enameling and Gilding on Glass,” The Glass Circle, v. 1, 1972, pp. 18–32. Many technical details will be found to disagree with this article, however.

2. Generally, in publications that include technical aspects of glass history, pre–19th-century enamels are uniformly—and inaccurately—described as having been made to fire at temperatures low enough so as not to soften and thereby distort the object during the firing process. A partial list of publications containing this error will be found in note 30.

3. I n this article, the term “kiln” will be used to describe an oven whose heating chamber can attain a maximum temperature of about 1,200°F or 649°C. In practice, there is no difference between a kiln and an annealing oven. The term “glass furnace” denotes a specialized piece of equipment intended for making glass or for reheating an object during its manufacture. Typically, such a furnace operates at temperatures of about 2,100°F (1,149°C) for glassworking and about 2,250°F (1,232°C) for glassmaking.

4. Although it has been an active field for decades, the scientific investigation of historical glass has recently increased noticeably. For example, chemical analysis of Islamic objects has focused on the glass from which vessels were made as well as on the enamels used to decorate them. Robert H. Brill, Ian C. Freestone, Julian Henderson, and Mark T. Wypyski are among the scientists who are researching and publishing in this area.

5. I an C. Freestone and Mavis Bimson, “Early Venetian Enamelling on Glass: Technology and Origins,” Materials Issues in Art and Archaeology IV, Pittsburgh: Materials Research Society, Proceedings, v. 352, 1995, pp. 422–425 (here, the two types are called “pre-melted” and “cold mixed”); Ian C. Freestone and Colleen P. Stapleton, “Composition and Technology of Islamic Enamelled Glass of the Thirteenth and Fourteenth Centuries,” in Gilded and Enamelled Glass from the Middle East, ed. Rachel Ward, London: British Museum, 1998, pp. 126–127 (here, the two types are called “pre-fritted” and “unfritted”). Although these articles discuss Islamic and early Venetian enamels (and vessel glasses), the distinction between the two types may be useful for studying and describing historical enameled objects in general.

6. Freestone and Bimson [note 5], figs. 4 and 5.

7. According to hyperphysics.phy, the thermal conductivity of some common materials [unit: (cal/sec)/(cm^2C/cm)] is as follows: silver, 1.01; aluminum, 0.50; lead, 0.083; “ordinary” glass, 0.0025; and asbestos, 0.0004.

8. Hugh Tait, The Golden Age of Venetian Glass, London: British Museum Publications, 1979, p. 29. This quotation, which was widely published during the 20th century, first appeared as early as 1849 in Mary Philadelphia Merrifield, Original Treatises on the Arts of Painting (New York: Dover Publications, 1967).

9. For an illustrated description of the core-forming process, see William Gudenrath, “Techniques of Glassmaking and Decoration,” in Five Thousand Years of Glass, ed. Hugh Tait, London: British Museum Press, 1999, pp. 214–215.

10. Dudley Giberson, A Glassblower’s Companion, Warner, New Hampshire: Joppa Press, 1998, pp. 17–18. See also Dudley Giberson’s Core Vessel Video 2004, produced by the Joppa Press.

11. Gudenrath [note 9], p. 221.

12. Beat Rütti, “Early Enamelled Glass,” in Roman Glass: Two Centuries of Art and Invention, ed. Martine Newby and Kenneth Painter, London: The Society of Antiquaries of London, 1991, pp. 122–136.

13. Gudenrath [note 9], pp. 226–227.

14. Ibid., p. 226.

15. R obert H. Brill, “Scientific Investigations of Some Glasses from Sedeinga,” Journal of Glass Studies, v. 33, 1991, pp. 11–28.

16. For enameled glass beakers from Begram, Afghanistan, dating from 50 to 125 C.E., see Joseph Hackin, Nouvelles recherches archéologiques à Begram, Paris: Imprimerie Nationale, 1954, figs. 257–269; and David Whitehouse, “Begram Reconsidered,” Kölner Jahrbuch für Vor- und Frühgeschichte, v. 22, 1989, pp. 151–157. For enamel-decorated cups that may have originated in the Rhineland from 150 to 270 C.E., now in Copenhagen’s National Museum, see Hugh Tait, Five Thousand Years of Glass [note 9], p. 87, fig. 108.

17. Omitted from this study because of its minimal use of enamels is a group of much-published base-disks, formerly part of bowls, dating from the fourth century. See Tait [note 9], p. 97, fig. 123. These objects more properly belong in a study of gold decoration on glass.

18. Hugh Tait, ed., Seven Thousand Years of Jewelry, London: British Museum Press, 1986, pp. 101–114.

19. See also D. Whitehouse, “Two Medieval Drinking Glasses with Gilded and Enameled Ornament,” in Hyalos, Vitrum, Glass. First International Conference: History, Technology and Conservation of Glass and Vitreous Materials in the Hellenic World, Rhodes, 2001, ed. G. Kordas, Athens: Glasnet Publications, 2002, pp. 199–203.

20. David Whitehouse, Lisa Pilosi, and Mark T. Wypyski, “Byzantine Silver Stain,” Journal of Glass Studies, v. 42, 2000, pp. 85–96.

21. Hugh Tait, “Venice: Heir to the Glassmakers of Islam or of Byzantium?,” in Charles Burnett and Anna Contadini, Islam and the Italian Renaissance, Warburg Institute Colloqui 5, ed. Charles Burnett, Jill Kraye, and W. F. Ryan, London: The Warburg Institute, University of London, 1999, pp. 77–104.

22. For a full description with illustrations, see William Gudenrath, “A Survey of Islamic Glassworking and Glass-Decorating Techniques,” in Stefano Carboni and David Whitehouse, Glass of the Sultans, New York: The Metropolitan Museum of Art, 2001, pp. 66–67.

23. Tait [note 8], pp. 25–28; idem [note 21].

24. Ingeborg Krueger, “A Second Aldrevandin Beaker and an Update on a Group of Enameled Glasses,” Journal of Glass Studies, v. 44, 2002, pp. 111–132.

25. For an illustrated description of the glassblowing process used to make the blank, see Gudenrath [note 9], pp. 223–225.

26. For an illustrated description of the process used to complete a goblet, see ibid., pp. 230–231.

27. Based on an ongoing study of German enameled glass with Dedo von Kerssenbrock-Krosigk.

28. As late as 1780, this process was described in association with German glassmaking by Georg Ludewig Hochgesang. See Axel von Saldern, German Enameled Glass: The Edwin J. Beinecke Collection and Related Pieces, Corning: The Corning Museum of Glass, 1965, Introduction, n. 1.

29. For example, since 1968, the “Hope Beaker” in The British Museum has been widely judged as having been made in the 19th century. Nevertheless, its decoration indisputably shows many signs of having been fired in the furnace while the object was reattached to a pontil. Whether this argues in favor of the previously held view that the beaker was of medieval origin or suggests an even greater cleverness on the part of a “faker” must be the subject of another study. See Hugh Tait in Fake? The Art of Deception, ed. Mark Jones, London: British Museum Press, 1990, pp. 182–183; and Krueger [note 24].

30. Among the recent publications including the error that pre–19th-century objects were decorated with low-fire enamels that were fired in a kiln are Linda Komaroff, “Color, Precious Metal, and Fire: Islamic Ceramics and Glass,” in The Arts of Fire, ed. Catherine Hess, Los Angeles: Getty Publications, 2004, p. 36; Julian Henderson, “Technological Analysis of Mamluk Glass,” in Mamluk Enameled and Gilded Glass in the Museum of Islamic Art, Qatar, ed. Stefano Carboni, Qatar, 2003, pp. 28–31; Erwin Baumgartner, Venise et façon de Venise: Verres Renaissance du Musée des Arts Décoratifs, Paris: Union Centrale des Arts Décoratifs, 2003, p. 142; Stefano Carboni, Glass from Islamic Lands, New York: Thames & Hudson, 2001, p. 325; Robert H. Brill, “Some Thoughts on the Chemistry and Technology of Islamic Glass,” in Glass of the Sultans [note 22], p. 35; and Stefano Carboni, “Painted Glass,” in ibid., p. 203. Many other recent and earlier books and periodical articles also reflect this misunderstanding.

Published on July 9, 2013