Venetian glass begins at the eastern tip of the city, five centuries before the furnaces moved to Murano. Excavations at San Pietro di Castello, an island called Olivolo in the early Middle Ages, produced 957 glass fragments from early medieval levels. Of the 45 samples later analysed, two are plant-ash glass, and the earliest of those is eighth century (Gliozzo et al., 2025, pp. 4, 19).
That date carries weight. Plant ash replaced Egyptian natron as the flux of Mediterranean glassmaking during the early Middle Ages, and the San Pietro di Castello material places Venice among the earliest sites in Italy to take up the new recipe, as early as the eighth century (Gliozzo et al., 2025, p. 19). It is also the century in which Venice itself firmly emerges as a geopolitical entity, with its physical foundations taking shape between Rialto, San Marco and San Pietro di Castello (Gliozzo et al., 2025, p. 20).
The assemblage was excavated in the late 1980s and early 1990s, reassessed between 2021 and 2024, and published in full in 2025, so the account below reflects the current reading of the evidence (Gliozzo et al., 2025, pp. 1, 3).
Where is San Pietro di Castello, and what did the excavations find?
San Pietro di Castello sits at the eastern end of Venice, beyond the Arsenale, on the former island of Olivolo. The island holds a pivotal position in the formative stages of the city: near the port entrance, controlling access to the lagoon, sheltered by the ancient barrier islands of Sant'Erasmo and Sant'Elena. Historical records attest an episcopal centre there as early as the late eighth century (Gliozzo et al., 2025, p. 1). That fits the wider lagoon pattern: settlements grow in number and size from the fourth century AD, and the see of Altinum moved to Torcello in 639, dated by an inscription found in its crypt in 1895 (Paiano & Delpozzo, 2024, pp. 30 to 31).
Investigations between the late 1980s and early 1990s revealed a layered sequence from the fourth century onwards: waterfront structures first (Period 1, fourth to mid sixth century), then road networks connecting a building to the lagoon (Period 2, end of the sixth to end of the seventh century), an early attempt to organise the lagoon environment (Gliozzo et al., 2025, p. 1).
Most of the glass comes from Period 3, the beginning of the eighth to the beginning of the ninth century, and from a building that yielded a gold tremissis of the Byzantine emperor Heraclius, who reigned from 610 to 641, with three seventh-century Byzantine seals. The excavators suggest it housed the magister militum, a high-ranking Byzantine officer, or served as a repository for public documents (Gliozzo et al., 2025, p. 3).
How old is Venetian glass? The eighth-century evidence
The early medieval levels yielded 957 glass fragments, from which 80 distinct vessels were recognised, along with six further items of production waste and glass blocks. Forty-three samples covering 38 objects were selected for analysis; adding the crucible and the glass layer stuck to it brings the total to 45 samples (Gliozzo et al., 2025, p. 4).
Two of those samples are plant-ash glass rather than natron glass, and the earliest is eighth century. On that basis the study concludes that Venice was among the earliest sites to adopt plant ash-based glass, as early as the eighth century, and reads the difference as one between coastal and inland Italy: coastal settlements had steady access to imported glass through maritime trade with the eastern Mediterranean and North Africa, while inland sites were more isolated from those routes (Gliozzo et al., 2025, p. 19).
This is what makes the site more than a local curiosity. Natron, a mineral soda from Egypt, had been the standard flux of Roman and late antique glass. Its replacement by ash from salt-tolerant plants was the largest single change in Mediterranean glassmaking between antiquity and the Renaissance, and the Venetian lagoon was in on it early.
Where the glass in early Venice came from
Chemistry gives the answer directly. Provenance analysis identified 15 samples attributable to the Levantine region, 19 to the Egyptian area, two of uncertain origin, and two plant-ash samples most likely of Levantine and Egyptian production (Gliozzo et al., 2025, p. 19).
The chronology inside those numbers is the interesting part. Egyptian products appear earlier, in the sixth century, whereas all the Levantine glasses, along with the plant-ash samples, are dated between the seventh and eighth centuries (Gliozzo et al., 2025, p. 19). Early Venice was therefore buying from one end of the Mediterranean, then from another, as trade routes shifted.
The pattern held for centuries. By the mid-thirteenth century Venice was importing raw materials such as sand and ash from the eastern Mediterranean, and documentary evidence in the Venetian archives suggests a more or less uninterrupted tradition of glassworking from at least the tenth century (Occari et al., 2021, p. 1). Later the Republic turned that supply line into policy, creating a monopoly on the importation, sale and use of Levantine ash and other raw materials in order to secure its supremacy in glassmaking (Occari et al., 2021, pp. 24, 25).
What the objects actually were
The assemblage is domestic and liturgical rather than luxury. The 38 objects analysed break down into working wastes (6), tableware and lighting vessels (8 goblets, 6 lamps and 9 beakers or lamps), architectural glass (4 mosaic tesserae and 4 fragments of window glass) and 1 bead (Gliozzo et al., 2025, p. 4).
The commonest form is the Isings 111 stemmed goblet, blown from a single paraison. These vessels likely served a dual function as both tableware and lighting implements, and they have disc bases with an average diameter of 4 cm, predominantly flattened (Gliozzo et al., 2025, p. 4). A reader can picture the object easily enough: a small footed cup, roughly the width of a modern espresso saucer at the base, that could hold wine or float a wick in oil.
The contrast with what came later is sharp. When medieval glass from under the basilica of Santi Maria e Donato on Murano was analysed, the commonest form was the inghistera, a long-necked pot-bellied bottle used mostly for water and wine, alongside flat-based beakers called moioli and hanging lamps called cesendelli, a term that appears in the Venetian archives for the first time in 1313 (Occari et al., 2021, p. 3).
How eighth-century glassworkers coloured and decorated glass
The technical results are where the study is most surprising. For blue glass, there is clear evidence that a cobalt-bearing slag, rather than a refined cobalt pigment, was introduced as the colouring agent (Gliozzo et al., 2025, p. 19). The colourant was a by-product of metalworking, put to a second use in the glasshouse.
Recycling shows up again in the mosaic tesserae. One object dated to the seventh or eighth century contains both calcium antimonates and lead stannates. Since calcium antimonates were largely replaced by tin-based opacifiers from the fourth century onwards, that combination more likely reflects the recycling of an older tessera than the deliberate use of two opacifiers (Gliozzo et al., 2025, p. 14). Older Roman glass was being melted down and used again.
The decoration points forward. Several vessels carry white trails marvered into the rim, and compared with the underlying glass these trails have markedly elevated concentrations of tin (5198 to 32460 ppm) and lead (5167 to 30142 ppm) (Gliozzo et al., 2025, p. 15). An opaque tin and lead white, applied as a thread on a clear body, is recognisably an ancestor of the lattimo and filigree work that Murano became famous for centuries later.
Was glass made here, or only bought in?
Both, on the evidence. The high-status building also yielded glassworking debris, including small blocks, lumps, chips and a steatite crucible with a layer of green glass adhering to its base (Gliozzo et al., 2025, p. 4). A comparable situation is documented on Torcello in the same period, where a glass workshop stood approximately 50 m from the cathedral of Santa Maria Assunta and its baptistery (Gliozzo et al., 2025, p. 4).
The goblets sharpen the case. Plant ash-based and natron-based goblets from the site share the same morphology, a base and bowl formed from a single gather, blown and folded back to create a foot with a tubular rim, flattened profile and a diameter of 4 cm. The study reads that similarity as a sign that plant ash arrived as a raw flux or semi-finished material and was reworked locally, rather than arriving as finished vessels (Gliozzo et al., 2025, p. 20).
One caution belongs here. The Torcello workshop has long been claimed as the start of a Venetian glass industry, but the mosaic tesserae found in it place it with workshops excavated elsewhere in Italy whose activity was based on recycling old Roman natron glass (Occari et al., 2021, p. 1).
From Olivolo to Murano
Glassmaking stayed in Venice proper for several hundred years after the Olivolo material was buried. It moved in 1291, when a decree of the state banned the presence of furnaces in Venice and relocated production to the island of Murano, mainly for safety reasons but also to secure better ventilation and more space for the workshops (Occari et al., 2021, pp. 2 to 3).
The usual retelling gives fire risk as the only motive. The peer-reviewed version gives three, and the commercial logic behind the move becomes clearer in what followed: glass cullet imported from the Levant and from Tripoli in the thirteenth century, sand from Crete between 1293 and 1302 and from Sicily in 1340, and silica pebbles from the Ticino river and from Verona in 1394 (Occari et al., 2021, p. 25).
Seen from San Pietro di Castello, 1291 looks less like a beginning than an administrative tidying of something already five hundred years old. The lagoon had been importing, reworking and recycling glass since the sixth century, colouring it with metallurgical slag by the eighth, and decorating it with opaque white threads long before anyone wrote the word Murano into a statute. The quiet green campo at the far eastern edge of the city, which most itineraries skip, is one of the three places where Venice physically began, alongside Rialto and San Marco (Gliozzo et al., 2025, p. 20).
