Plaster Transennae and the Shaping of Light in Byzantium

Glass had a very important role in characterizing Byzantine interior design and architecture. Mosaics and opus sectile revetments have attracted much scholarly attention, but the study of the glass that was used to screen the window openings and so to filter the light penetrating the interior has long remained overlooked. Archaeological excavations have brought to light some evidence for glass working (Corinth), trade in window glass (Sardis), and display (Lesnovo, Studenica), but it is still very scarce. The excavations at Amorium (Afyon, Turkey), capital of the Byzantine thema of Anatolikon, has produced some remarkable evidence about glass making and the use of window glass in secular as well as in religious buildings. This evidence spans about half a millennium, between the fifth and the eleventh century. Glass-working waste, found in the so-called ‘Lower City’, proves that a variety of glass objects and windowpanes were produced in Amorium, although the site of a glass workshop has still to be located. The discovery of several chunks of frit leads to the conclusion that glass itself was also produced in the city from the raw ingredients. These were sand and an alkaline flux that could be either soda or potash. Soda gives a very durable glass, but potash makes the fabric of the glass rather fragile. The differing states of preservation found among the Amorium window glass fragments show that both soda and potash were employed as flux in the glassmaking, perhaps at different stages in the long history of glass production in the city. A large quantity of cylinder-blown window glass has been recovered in association with the bathhouse complex, built in the early Byzantine period and remodelled during the Dark Ages. This is not surprising, considering that glass windowpanes had been widely employed in Roman baths since the second half of the 1st century. The glass panes from the bathhouse are generally thick, naturally coloured, quite well preserved, and cut into rectangular shapes. They must have been set into plaster transennae, as shown by a fragment of glass pane still encased within the double surface of a plaster jamb. However, from the same area also came one large fragment of a marble transenna, in which the holes were left open, without any screening medium. This kind of transenna must have been used in rooms that needed good air circulation and were not sealed with glass windowpanes. Although the production of cylinder-blown window glass must have continued at Amorium throughout the Dark Ages, the ninth century saw the introduction of double-layered plaster transennae with circular holes encasing crown-blown discs. Examples have been found in the main church of the Lower City, whose construction dates back to the late antiquity but which was lavishly remodelled sometime after the Arab attack of 838. Although crown-discs appeared in the Roman Near East in the late 2nd century and are also recorded later on elsewhere in the Late Roman and Byzantine world, Amorium offers the first example of their employment in plaster transennae with circular holes, a form that was to become the standard window screening device in Byzantine architecture