In over two decades of archaeological research at the Roman villa of Aiano (Tuscany), discoveries have been numerous and diverse, and the intricate nature of their depositional contexts have fostered a consistent report activity on these findings and on their interpretation (Cavalieri, 2023). Once pieced together, these scattered pieces of evidence have contributed to redraw a long and complex history of the villa that spans from the end of the third century to the 7^th century A.D. A monograph gathering the results of 20 archaeological campaigns is currently underway and hopefully will be published shortly. In this article, I provide a synthetic and updated review of recent analytical data of archaeological material that all demonstrate the wealth of the villa’s inhabitants in Late Antiquity. Such demonstration of wealth in aristocratic domestic contexts is, in fact, rather common in the Tuscany, with notable examples at the villa dell’Oratorio of Capraia and Limite (Florence) (Cantini 2017), San Vincenzino of Cecina (Donati, 2013), Ossaia in Cortona (Gualtieri, Fracchia, Ferrari, 2014), and more recently the Vignale (Giorgi, Zanini, 2024).

This paper aims at a definition of the luxuria villae of Aiano that goes beyond the stylistic and morphological elements of the residential architecture, such as floors and wall paintings, which have been discussed elsewhere (Cavalieri 2023). It proposes to «quantify wealth» through the prism of glass paste sectilia and beads, by focusing on composition and production techniques of these artefacts, analysed archaeometrically. Indeed, recent and ongoing research at Aiano and at the other villas of northern Tuscany has been clarifying the characteristics of late-antique residential settlements in the region. While the topic has remained for several decades ill-documented in the academic scholarship, recent archaeological research has contributed to draw the picture of a cluster of prosperous residential settlements that stand in sharp contrast with the long-standing historical model of decadence and economic recession of the Italian peninsula at the dawn of Antiquity. Once the environmental and productive landscape of the villa has been presented (in particular wine-making in dolia), the methodology aims to combine and discuss depositional history, typology, forming technique, and chemical composition of the glass-pastes at Aiano. The focus is on archaeological disiecta membra that have been thus far published in various articles: namely, sectilia with marine scenes and ornamental beads.

The remains of the late-antique villa of Aiano (4^th-5^th century A.D.) lay in the middle-basin of river Elsa, below the confluence of the Fosci and Riguardi torrents, on a river terrace slightly decreasing towards the South and sheltered on the North by a low hilly ridge (circa 140 Mt above sea level). Most likely, this villa was part of the large administrative district of Volaterrae (Volterra) —also the closest city, located about 17 km to the west (Fig. 1). Although the settlement was inaccessible from the main communication routes, it was easy to reach. Its position nec in via nec a via procul mirrors Columella’s prescriptions on the siting of villas (De re rust. I, 3, 3-5; I, 5, 6-7). Referring to Cato’s opinion on the opportunity to build villas where they could be well-served by the road network, Columella recommended avoiding building in direct vicinity to the road, and particularly public roads that were frequented by the military.

As for water supply, the villa could rely on the phreatic resources of the nearby hill —where the local toponym Castellaccia might indicate a castellum aquae, i.e. a cistern. Moreover, a constant water supply was guaranteed by the Fosci torrent —a hydronym coming from the Latin word fauces, or «upper outlet», indicative of the narrowness of the valley. The torrent runs down westward near the villa, and is supposed to have had a substantial flow rate even during the driest months of the year. The field around the site shows a gentle slope towards the torrent; it was probably exploited for agricultural purposes. Hordeum (barley) has, in fact, been detected among the pollens retrieved in archaeological contexts around the villa, which demonstrates further the agricultural use of the surrounding land. The cultivation of vines is also attested following the discovery of pollens recently collected in the stratigraphy of villa’s pars rustica. The area also featured —still in the Middle Ages— a vast forest providing wood, pasture for livestock (pig bones with butchering marks have been found in the villa), wild fruit and game, which matches the Roman definition of a saltus. All in all, the villa’s location fulfils the recommendations of Roman agronomists: a structure facing south and sheltered from the northern winds, located at the foot of a hill but above the torrent, thereby close enough to both water and woods (Fig. 2). The reasons behind the choice for this locus amoenus remain uncertain. While the origin and identity of its wealthy founder are unknown, one can perceive their intention to build a sumptuous villa close to a major crossroad, far from the threats of roads and the form the noise of the cities, where one could enjoy otia, hunting, and agriculture (Cavalieri, 2021, p. 69).

Recently, archaeology has been challenging more and more the idea of a depopulation of the Val d’Elsa between the 1^st century B.C. and 3^rd century A.D. (Cavalieri, 2016, p. 115-116). The local toponymy has provided further support to this idea. Numerous praedial place-names are derived from Etruscan and Latin names: they dot the area between the Western sector of the municipality of San Gimignano and the left bank of the river Elsa. Such evidence indicates that the late-antique villa of Aiano (the predial toponym being still of uncertain derivation, perhaps VIano>Virius) was built in a hilly area already that was fairly inhabited during the High Empire. However, even prior to the Roman conquest of the region, the whole Fosci valley hosted a consistent and widespread network of settlements, as is confirmed by the discovery of a votive depositat Torraccia di Chiusi, on top of a small hill located a few hundred meters to the North-West of Aiano. The sacred area lays in the immediate vicinity of a spring that shows evidence of frequentation from the archaic period down to the Middle Ages, and it has produced numerous findings of exceptional historical value. The long-standing occupation history of the region and the density of settlement patterns are also indirectly confirmed by the archaeological research at Aiano and at surrounding sites. Not far from the villa, at the Monti settlement, a marble funerary urn, dated to the end of the 1^st century or the beginning of the 2^nd century A.D., displays protomes of Amon (Fig. 3).

It testifies to the existence of a funerary site in the vicinity. Finally, in archaeological contexts of the 6^th and 7^th century at Aiano, bronze artefacts and pottery of Etruscan time were found: their chronology spans from the early archaic to the hellenistic period (Cavalieri et al., 2010).

These findings are probably the result of burial looting, not much with the purpose of collecting precious items, but instead to recycle materials —metals in particular— during the early Middle Ages. Indeed, during that period, the villa transited from a residential settlement to a craft production area using building and decorative materials from earlier occupation phases of the villa, such as metal, glass, plaster, marble, etc.

Since the beginning of our archaeological research program at Aiano in 2005, excavations have uncovered a total of 5,000 square meters over the 10,000-15,000 estimated through geo-physical analysis. This confirms the archaeological value of the site, further enhanced by a good conservation state (Fig. 4).

A lavish villa was brought to light through stratigraphical research, for which we propose to identify six successive occupation phases:

1. Foundation of the villa (end of 3^rd - first half of 4^th century A.D.); 2. Re-organisation of the settlement and transformation of the residential area into what represents —in the late-antique Italy— an architectural unicum —in the second half of 4^th or second half of 5^th century A.D.; 3. Abandonment of the site in the end of 5^th or mid-6^th century A.D. —and transformation into a demolition site in which building materials and artefacts were purposely recycled; 4. Between the second half of the 6^th and the first half of the 7^th century A.D. the site ceases its activity and is definitely abandoned; 5. In the mid-7^th century, amid the ruins of villa, two people are buried —as shown by the osteological analysis they were possibly pilgrims, for signs of stress appeared only in the lower limbs. They might have died on their way on the Via Francigena, which passes nearby, and had been buried by the building’s collapse; 6. Disappearance of the site from the local landscape, and rediscovery through archaeological excavations —20^th century.

This data provides the picture of a highly prestigious site in Tyrrhenian Italy, probably fostered by the direct links of its inhabitants with Rome. The discovery, over the past few years, of more than 2,000 fragments of sectilia in glass paste shows the level of wealth at the site and the cultural and economic standards that were enjoyed by the Tuscan aristocracy in 4^th and 5^th century A.D.

One should stress that the entire decorative apparatus of the villa as we know it (marbles, sectilia, mosaics, frescoes, and stuccoes) has been found disconnected from its original decorative function, as materials and items had been recycled before the abandonment of the site in 7^th century A.D. Hence the impossibility to reconstruct with precision spaces and functions of the findings, apart from referring them to the so-called «triconch hall», a living area which bears some resemblance to a trichorium. The building has a central plan and the shape of an equilateral triangle with apses adjacent to each summit, the southern one being a double-threshold vestibule. The hall itself is placed within a curved wall system, a sort of ambulatio polylobata, where five wide circular lobes are closed on the South side by a quadrangular vestibule connecting the central hall and the surrounding corridor. While most of the marble and mosaic decoration of the villa is lost, the floor of the «triconch hal» has been preserved, although with some gaps, and it seems reasonable to date it to the beginning of the 5^th century A.D. It is an opus signinum with geometrically decorated mosaic inserts (Fig. 5). What is left of it allows a partial reconstruction which shows extremely refined materials: inlaid marble flooring, wall paintings imitating opus sectile, mosaics with stone and opaque glass paste tesserae, and gold-leaf ialina. Here are traces of stucco architectural decorations, crustae marmoreae, glass paste sectilia reproducing ichthyic subjects and connected with craftsmanship coming from Alexandria, the port city in Egypt (Fig. 6) (Cavalieri et al., 2018, p. 493-498).

The outline provided above demonstrates that the ancient settlement network in the Fosci basin area persists without major upheavals until the modern era. The process of systematic recycling and repurposing decorative materials at the beginning of the Middle Ages contributes to this continuity. When managing this large, lavish residential villa became no more economically viable and when the structure decayed (either following the dissolution of the property or due to the discrepancy between the income generated and the luxurious features of the building), craftsmen began dismantling the structure bit by bit and repurposing its materials, effectively re-inserting them into the production line. So, while life went on in the surrounding rural settlement, the villa lost its purpose and original function and became a workshop, undergoing a radical transformation where its materials, if recyclable, were systematically taken away. The systematicity of the process in itself testifies to the ongoing vitality of the settlement, despite its changed function.

Ever since 2019, research in the north-eastern sector of the villa has brought to light a large pillared room. This has been a crucial discovery, providing fundamental data and information for the reconstruction of the history and importance of the villa between the 4^th and 5^th century A.D. It shed light upon the villa’s productive aspect at the core of the Val d’Elsa. This sector of the site is divided into four main units (Fig. 7). In the centre, there is a large cella vinaria (β) of about 30m × 9m, marked by six axial pillars dividing it into two naves (Fig. 8). Within this room, which is until now one of the widest of the villa, 22 well-preserved dolia defossa have been discovered and partially studied. According to their position and to the large surface of this room, it is possible to infer the presence of about fifty dolia, located along four lines (Cavalieri et al., 2024, p. 303-310). Immediately to the south, unit α seems to have been a large water vat, as suggested by the buttresses leaning against the outer walls.

In 2024, two rectangular vats have been excavated (circa m 2.1 × 1.7): they were sealed with cocciopesto and each of them was equipped with a small staircase in the south-east corner. The floor shows a slight declination towards a central a ceramic cuvette to collect the scum. These vats, or lacus in Latin, were used in the first phase of the collection and settling of grape must (Fig. 9). The cellar opened towards the south, on a long porch (γ) which bordered a large courtyard, most probably the core of the villa’s pars fructuaria. Lastly, unit ε: here, excavations have produced a sub-circular pavement which, despite the fact that it was completely taken apart and stacked, has provided evidence of the profile of the torcular.

The Department of Chemistry at the University of Pisa has been carrying out archaeometric analysis since 2022, using soil samples collected from several dolia defossa and from the two lacus. Whereas qualitative results regarding the last samples are not yet available, the analysis of the soil from the dolia is instructive.

All samples have been analysed using gas-chromatography and mass spectrometry (GC/MS). They revealed traces of pine tar, which demonstrates that the dolia were liquid-proofed with tar, a technique commonly used to preserve wine, as witnessed by Columella (Colum. xii, 18, 3-6) (Fig. 10). However, until today no trace of grape residues have been retrieved, for example grape seed. This does not exclude the presence of residues —perhaps it is just difficult to find any because of the repurposing of the space (phase 3, detailed above) and of the almost systematic leveling of the vessels. The next 2025-2027 excavation campaigns will focus on finding further organic proofs of the wine-making process.

Lastly, several samples also produced also traces of fat acids, an element suggesting the presence of vegetable oil. We cannot therefore exclude that these dolia were used (or re-used) after the residential phase of the site, as olive oil containers.

Given the large dimensions of the cella vinaria and the quantity of wine that was possibly preserved in the dolia —an average quantity has been estimated around 1,000 litres per dolium— it is possible to infer that such production was not meant for self-consumption nor aimed at the local trade. Instead, it is more likely to have been serving the Ostia market, and hence Rome which, in the fourth and fifth centuries, was increasingly dependent on food supplies coming from the whole peninsula. Winemaking may have been a fundamental source of income —perhaps not the only one— for the villa’s owner. Although unfortunately, we do not know his name, thanks to the fine architecture of the site and its precious decorations provide a glimpse into a high social standing and a culture inspired from the most refined tradition of luxuria Urbis.

The most significant cluster of fragments related to the sectilia was retrieved in the area east of corridor R (see Fig. 4) —a place used as a sort of warehouse for waste material coming from the villa (Cavalieri et al., 2016, p. 286-291; Cavalieri et al., 2018, p. 495-498). Glass was probably transferred there before being repurposed, following a systematic reuse of the settlement, which can be dated to the late 6^th century according to numismatic data (Cavalieri et al., 2010).

Our research program at Aiano involved a reconstruction of the fragments by extrapolating them from the UUSS they belonged to. It seemed a suitable approach, especially because, in doing so, we could suggest the original position of the sectilia fragments with marine decoration. These appear to have been organized in several panels, each of which presumably measured 20-25 cm in length. Glass fish, made using multiple techniques similar to those in Aiano, are attested as early as the 1^st century A.D. For instance, a glass cup retrieved in the Athenian Agora displays a fish pattern in silver, gold and green colours at the bottom of the vessel. More complex arrangements, always featuring fish patterns as integral parts of big glass paste sectilia are a testimony of the wealth associated with this type of artefacts at least from the mid-Imperial age onwards. At this time only the elite commissioned glass panels, as witnessed by an excerpt from the Historia Augusta (xxix, Firmus 3.1) mentioning the case of the Alexandrine usurper Claudius Firmus who, at the end of the 3^rd century A.D., wished to enhance the decoration of his house with glass paste sectilia applied with tar or another similar material. Such sumptuous apparatus was most probably not much different from the retrieved thousands of fragments from the villa of Lucius Verus on the Via Cassia, in Rome. These fragments belonged to panels featuring phytomorphic decorations filled with griffins. What emerges from the Historia also seems to suggest the Egyptian origin of this decorative style (Brill and Withehouse, 1988, p. 37): a generally accepted theory also supported by analyses of the Thomas Panel of the Corning Museum, presumably coming from El-Fayum. There is no lack of evidence, with examples of sectilia from Egypt which can be added to numerous fragments of Egyptian glass fish appeared on the antiquarian market.

Two glass paste panels discovered in Rimini and Corinth and featuring marine subjects can be traced with certainty to their original context. In the first case, the Surgeon’s domus (Casa del Chirurgo) destroyed in 257-258 A.D. (Ortalli, 2000): here a circular panel has been brought to light inside a room for triclinia. In the panel, a series of multiple frames included the image of three fish against a light-blue/turquoise backdrop. In the second case, the Corinthian one, the sectile comes from a space East of Corinth’s theatre, which collapsed for an earthquake at the end of the 3^rd century A.D. In it, are four fish against a blue backdrop. The fish are made with multiple techniques: millefiori next to banded decoration.

In both cases, the composition is linked to decorative schemes well-known to the contemporaneous mosaic art. Images of fish are included in the landscape-like the glass paste sectilia of Kenchreai; or in fishing scenes, or also represented as individual subjects, in groups or three or four. As for the latter layout, which had been already seen in Pompeii, probably the peak of its diffusion was in the 3^rd century A.D. In the case of the Zliten mosaics in Libya (first half of the 3^rd century A.D.) the layout is very similar to the glass paste sectilia under investigation here (Parrish, 1985.).

A cluster of sectilia showing at least six fish of about 20-30 centimetres was found in the 1990s in Trier (Augusta Treverorum, Germany), in the presbyterial area of one of the two halls of the bishopric —this was dating back to the mid-4^th century A.D. It is presumed that they are what is left of a wall decoration attributed to Egyptian craftsmen arrived in Trier with Athanasius, bishop of Alexandria, who was exiled there in the mid-4^th century (Merten, 2012-2013).

The Kenchreai panels were found near Isis’ sanctuary, in Corinth, inside buildings which were lost to the 375 A.D. earthquake. The panels show how fish patterns were still used in glass-paste sectilia decorations at the end of the 4^th century. Over one hundred panels display a large range of fish species —sea bass, sea bream, red mullet and cuttlefish, all within a landscape which looks inspired by the Nile area.

The Thomas Panel exhibited at the Corning Museum in New York is more recent, as it is attributed to the first two decades of the 5^th century A.D., and is probably among the most similar to the Kenchreai examples, making it plausible that all these sectilia came from one Egyptian workshop operating between 4^th and 5^th century A.D. Interestingly, the glass paste panels from the villas of Italian potentiores (for example the Albinus Caecina’s one in San Vincenzino —Tuscany, Livorno (Donati, 2012, p. 444-446)— and the one belonged to the Scipiones Orfiti, in Faragola —Foggia, Apulia—) were also composed around that time (Volpe, Turchiano, 2012, p. 309 ss.). Moreover, their technical features and decoration suggest that all of them were the work of craftsmen from Alexandria. In particular, a fragment found in San Vincenzino provides a close comparison to the ones in Kenchreai, where the landscape displays Egyptian features. This being said, rather than a mere diffusion of stylistic features among glass craftsmen communities of the Late Roman Empire, it is more likely to suggest that these craftsmen came, in fact, from Alexandria.

As for the executive technique, we have analysed the murrine (sections of glass-past canes) present in the fragments. The sectilia from Aiano, in fact, are extremely refined pieces of work showing a high-quality manufacturing technique. Each fish pattern is made from several murrine, aiming at creating the illusion of scales, with their nuances and the various details in the animal bodies, resulting in extremely realistic pieces of work.

Significant differences define the bodies of fish in our examples: in the Riminese sectilia fish bodies are divided into bands, each of which is made with monochrome glass paste; some of the Corinthian fish are constructed with murrine arranged as a mosaic, without a real division in bands (Oliver, 2001, p. 349-363). The specimens from Aiano demonstrate a rather complex processing technique: three different chromatic bands, each of which made of several series of murrine were aligned in a very regular fashion, due to the parallel juxtaposition of the linearly cut glass canes.

For example, in the fish fragments displayed in Figure 11 (perhaps a sea bream), the body of the animal is divided in three separate chromatic bands: the upper band is dark turquoise, the middle one is light yellow and the lower band is grey. Each band in turn is made of a sequence of several adjacent series, composed of five murrine next to each other.

More specifically, the upper band (dorsal) murrine are taken from a circular glass paste cane made of five concentric stacked strata of the following colours (Fig. 12, from the outside to the inside):

Once assembled, the cane was pressed and cut into two parts, to obtain an oblong shape imitating real fish scales (roughly 3.5 mm section), the canes then were put next to each other in a series of five and cut to obtain the murrine which in turn were pulled together and joined through heat.

This process made one band. Two more similar bands were then added, together with the black upper edge: all the fish body was therefore modelled through heat, stretching it when necessary and pressing the ends, to obtain the desired shape.

Subsequently, other murrine were added for the eye, mouth and teeth; and then the red line corresponding to the gill and the pectoral fin, and lastly the outer fins (back, pelvic and tail) made with red, yellow and black glass paste stripes.

A close analysis of individual details therefore demonstrates the use of extremely sophisticated techniques, together with a very high level of craftsmanship. Similar quality is found in a glass paste fragment preserved at Kelsey Museum, University of Michigan. This fragment, as is the case with the one from Corinth, features the murrine arranged in a mosaic shape and not in sequences gathered.

Furthermore, the analysis of sectilia has allowed us to understand the way in which the fish patterns were inserted into the backdrop: we suppose that they were placed within the panels after carefully carving in the panels the necessary space through cold cutting, presumably starting with a first hole made with a curved steel blade and abrasive, until the shape was defined. Such inlays are made with great accuracy, as can be seen in the fish mouths and their backgrounds, thus suggesting not only precision tools, but also a in-depth understanding of materials and their characteristics.

Such complexity and sophisticated production skills justify the hypothesis of an Alexandrine origin for these artefacts. Alexandria was well-known for its exquisite glasswork and, since the Ptolemaic age, for its thorough study of flora and fauna, which in turn merged in the Library of Alexandria, the famous extraordinary repository of ancient knowledge. These precious and fragile objects were most probably imported to Aiano by high-level clients who enjoyed cultural connections with Rome and could afford expensive purchases from the Oriental market.

Twenty glass paste beads and one pendant were retrieved during the excavation campaigns conducted at Aiano from 2005 to 2019 (Boschetti et al., 2022). The beads were all deposited in backfills formed between the 6^th and the 7^th century, when the artisanal activities came to an end and the area was abandoned. These backfills are characterized by a high level of fragmentation. Finds dating to the second half of the 4^th century, the time of the monumental renovation of the villa, are accompanied by late 5^th- to 7^th-century material. Earlier material is less frequent. The beads were scattered across the vast area occupied by the villa, without any significant clustering, and no bead was deposited in fills associated with the glass-recycling furnace excavated in the ambulatio of the five-lobed hall (Room O). An important concentration of glass-working waste was retrieved from Area N, located immediately outside the central hall and interpreted as a zone used as dump, collecting the waste generated by the craft activities. The visual examination of all the glass finds from Aiano ascertained that there is no evidence that could be linked to bead working, such as tools, glass-past canes, and/or failed beads (Cavalieri and Giumlia-Mair, 2009). The glass-working waste from Area N includes fragments of vessels and tesserae deformed by heat, fragments of refractory material with adhering glass, and masses of tooled coloured glass (Fig. 13). These finds testify to the recycling of colourless and coloured glass, but it is impossible to ascertain whether this glass was shaped into new objects at Aiano or whether it supplied some secondary workshops in unprocessed glass. As documented during the 4^th century A.D. at Aquileia and during the 9^th century at San Vincenzo al Volturno in Molise (Schibille & Freestone, 2013), colourless glass was recycled and coloured by adding mosaic tesserae. A similar recycling process probably took place at Aiano, where the deposits associated with the furnace yielded an exceptional concentration of windowpanes and mosaic tesserae.

Although the bead typology in the Roman and early medieval Italy remains ill-documented in the scientific scholarship, the data published so far allows for a basic characterization of the Aiano assemblage. The beads are clearly different from Migration-period types (Boschetti et al., 2020) although they do display some continuity with the Roman and Late Antique traditions (Mandruzzato, Marcante 2008). Their manufacturing techniques are equally revealing: only wound and segmented glass beads have been identified at the site, which are only two of the four production methods reported thus far in the Ancient Mediterranean world (Fig. 14). Drawn, cold-cut, hot-polished and mosaic varieties of beads are notably absent from the Aiano assemblage.

Most specimens (that is, 17 beads and one pendant, TCC09) were formed by winding. Among these, three small monochrome beads (TCC01, 05, 015) clearly imitate semi-precious stones, a practice that was widely attested between the 3^rd and 6^th centuries in the Late Roman world (Swift 2003). Another specimen of a distinctive light blue colour (TCC04) belongs to a popular type in Egypt and Nubia during the 4^th-6^th centuries (Then-Obłuska 2018); it was probably inserted in a gold or bronze frame alongside real gemstones to create a complex and colourful pattern.

Another set gathers seven beads and the pendant TCC09 which are all made of black glass (TCC06, 07, 13, 17, 18) which was a highly sought-after for crafting jewellery between the 3^rd and the 5^th century (Cosyns 2011). The mottled black bead TCC12 exhibits red, white and pale-blue elements, thereby conveying a fashion observed in Late Antique Fayoum burials and in the jewellery workshop in Alexandria’s Diana quarter (Boschetti et al. 2022). Other similar pieces are known from Italian contexts spanning the 1^st to the 5^th century (Mandruzzato, Marcante, 2008). The spiral bead TCC18 is also very similar to examples found in the Alexandrian workshop.

Jug-shaped pendants, represented by a single example at Aiano (TCC09), were widely attested across the Levant and Egypt, and they also became popular in Italy in the late 4^th and early 5^th century (Mandruzzato, 2017). They are frequently interpreted as Christian amulets and are often found in women and children burials (Bolla, 2011). Another bead, TCC21, may also be of Egyptian origin, given that its shape and decoration closely match Late Antique Egyptian spindle whorls (Spaer et al., 2001).

Although Roman beads are predominantly seen as commodities exchanged on local and regional markets (Mandruzzato, Marcante, 2008), evidence from Aiano challenges this view when examined against a wider geographical backdrop. The large black spherical bead TCC07, with crossed white trails and green-yellow eyes, is identical to a specimen from Aquileia dated to the 4^th-5^th century. Other specimens have been found in Austria and suggest that this region hosted the main production centre of this type of bead. Comparable materials are also reported in Britain, among burials dating to the mid-5^th and early 6^th century; they are considered local productions. Already in the 1970s, similar British beads were compared to continental forms from the 4^th-5^th century and were thereby classified as «exotic» Roman imports from the European mainland (Guido, 1978). An even earlier example, discovered in a 3^rd century grave at Classe —the port of Ravenna— was associated with black beads decorated with coloured zigzag trails akin to Aiano’s TCC06 (Montevecchi, 2000).

Other wound beads from the site, such as the melon bead TCC19 and the cylindrical beads TCC02, 03 and 20, belong to very common types of uncertain date. The few drawn and segmented beads (TCC08, 14, 16) form a small group; one of them, the green bead TCC08, was finished up by heat polishing. These drawn beads originate from Egypt and circulated widely in Europe from Late Antiquity onward, with a major diffusion beginning in the 4^th century A.D.

Elemental and molecular analyses have proved to be fundamental for examining the chronology and provenance of the Aiano beads, especially where typology alone is insufficient. As for monochrome wound beads, this approach is particularly decisive, since typologies span several centuries (Cavalieri et al., 2022, p. 10-12). Four specimens are composed of Roman base glass: the colourless melon bead TCC19 and the drop-shaped beads TCC04, TCC05, and TCC15. Among these, TCC04 —an opaque green bead— combines Roman glass with calcium antimonate as an opacifier, a technological know-how that should be no later than the 4^th century. By contrast, the blue bead TCC05, opacified with calcium stannate, reflects a craft tradition that postdates the 4^th century.

Other monochrome beads display different trends. The spiral black bead TCC18 consists of Foy 2.1/HIMT glass and aligns closely with the composition of black Egyptian beads of the 4^th and 5^th centuries, while the colourless barrel-shaped bead TCC01, made of HIMT glass, likely belongs to the same chronological horizon.

The polychrome wound beads separate into two distinct technological groups. The first group comprises six specimens (TCC03, 06, 07, 10, 13, 20), all produced using recycled Roman glass combined with antimony-based opacifiers and Roman cobalt characterized by a high Co/Ni ratio (as seen in TCC13). These criteria are suggestive of a production no later than the 4^th century. Within this group, two beads (TCC06, 07), whose typological parallels can be tracked back to the 3^rd century, may represent residual material from the earliest occupation phase of the villa.

The second group consists of four beads (TCC02, 11, 12, 21) and it also includes the jug-shaped pendant TCC09. These specimens all share the assemblage of core material made from mixtures of Foy 2.1 and HIMT glass, and decorative threads using recycled Roman glass. TCC02, for example, combines a green Roman-glass thread opacified with calcium antimonate and a red core matching a HIMT/Foy 2.1 composition. In the remaining pieces, Roman-glass decorations contain various opacifying agents, such as calcium stannate (TCC11, 12), calcium antimonate (TCC21), and lead stannate (TCC09). Roman cobalt is also present in the translucent Foy 2.1 body of the mottled bead TCC11, while the blue thread on TCC21 —a bead which shape recalls the Egyptian spindle whorls— is coloured with a later cobalt source and opacified using a calcium antimonate.

One bead, however, is considered a loner sample: that is TCC17. It is exclusively composed of Foy 2.1 glass; its translucent blue core is coloured with a post-4^th-century cobalt raw material, and its white dots are obtained using a calcium stannate opacifier. This is the only example in which the base glass mixture, the colourant, and the opacifier are all attributable to mid-5^th century (or later) traditions.

In conclusion, the compositional characteristics defined archaeometrically at Aiano all strongly support an Egyptian origin for several bead specimens, and more particularly for the three mottled beads (TCC11, 12, 17), the bead/spindle whorl TCC21, and the jug-shaped pendant TCC09. Furthermore, the three drawn-segmented beads (TCC08, 14, 16) are all made of Foy 2.1 glass (Fig. 15), and their forming technique aligns with Egyptian traditions from the 5^th to 7^th centuries. The presence of Roman cobalt in the blue bead TCC14 further suggests that this type belongs to an early type of Foy 2.1 from around the mid-5^th century.

Finally, one should note that the findspots of the beads do not display any specific clustering, according to their chemical composition nor forming technique. The only exception concerns the three drawn-segmented beads from Area N, which come from stratigraphically related contexts and may thus have once been strung together, although archaeological evidence is so far insufficient to confirm this.

I have defined the Aiano villa using rather emphatically a Pindaric image, «the shadow of a dream»1: the architectural and decorative luxury of the villa is, in fact, perceivable only as a shadow, the skeletal remains of what once was a monumental complex of buildings where materials (marble, stuccoes, metal, glass etc) were systematically sacked from the 6^th century onwards. All that was left were scarce traces of the creative and imaginative dream of its creator. For an archaeologist, albeit used to work with the concept of scarcity, it is a highly frustrating experience, only partially overcome by the advantages of virtual archaeology and the use of archaeometry... stating the opposite, after having experienced archaeological excavations, would be a lie.

In any case, analytical methods borrowed from the applied sciences allowed us, on the one hand, to reconstruct spaces, volumes and colours of the villa (Cavalieri, 2023, p. 88-98); on the other hand, to identify at least some types of materials which lay scattered and fragmented in the archaeological warehouse. For the sake of consistency, this article focussed on glass paste sectilia and personal ornaments, which have been studied following criteria of typology and production through archaeometric analysis.

Starting from and combining data published in two previous publications on Aiano (Cavalieri et al., 2016; Boschetti et al., 2022), this article uses an archaeometric and typological approach to glass paste materials in order to quantify the luxury displayed in the villa of Aiano during its residential phase (4^th-5^th century A.D.). As this is residual material, amounting to no more than 0.5 % of the villa’s finds, and as it was always found out of its stratigraphic context and often fragmented and accumulated for remelting, the archaeometric analytical approach and the systematization of the two types of artefacts has created a more coherent and organic whole, useful for demonstrating the refinement, including technical expertise, of the materials and ornaments with which the elite owners of the villa surrounded themselves.

Regarding the glass paste sectilia, we have used comparison to confirm the common Alexandrine origin of the craftsmanship. We prefer this hypothesis compared to the idea of a mere circulation of iconographic patterns within glassmaking workshops, as found in the examples of Kenkrai (Greece) and Italy (villa of San Vincenzino). The mounting and transport procedures of the glass panels are still unknown. Were they imported from the Eastern Mediterranean territories already completed, or —as suggested by the presence of local Tuscan elements within the glass paste of the Faragola sectile— should they be ascribed to itinerant specialized workshops that used local glassmaking facilities? Probably the Aiano sectilia too should be thought as the work of these master craftsmen, who operated between 4^th and 5^th century A.D. and served an elite clientele. Unfortunately, the decontextualization of the finds prevents us from proving a strict link between the marine themes and the use of water at the villa, as occurs, for instance, at the villa of Faragola in Apulia.

The analysis of the corpus of the Aiano glass beads clarifies their interpretation, date, and connection to the villa. First, the stratigraphic position of the beads and the lack of archaeological indicators of bead production at Aiano rule out a local production during the reoccupation of the building. The results of the typological, technological, and chemical analyses establish a direct link with the life of the villa during its main period of prosperity. The middle of the 4^th to the 5^th century A.D., to which the majority of the beads and the pendant belong, corresponds to the main period of building monumentalization of the villa. It can be reasonably assumed that the Aiano beads were ornaments worn by the people who lived in the villa before it was abandoned. A few beads can be dated to as early as the middle of the 3^rd century and can be assigned to the first phase of the building. All the drawn beads and at least half the wound beads and the pendant are imports, probably from Egypt. Rather than being local products, the Aiano beads were obtained through an exchange network operating at the Mediterranean and European level. Beads travelled exceptionally long distances between the Bronze Age and the Iron Age, and, later, during the Migration period. Contrary to previous perceptions of Roman beads as generally worthless artefacts, produced and consumed locally, the clear identification of exotic beads among the finds from Aiano leads to a necessary reassessment of the status of Roman glass beads in terms of their aesthetic, economic, and symbolic value. While the international dimension of the bead trade during the Roman and Late Antique period or the extent of its market are currently under-documented, the results obtained for the Aiano assemblage support a model of long-distance exchange. Finally, this study of typology, technology, and chemical composition of the beads and glass paste sectilia clearly excludes the possibility of a local production. Instead, two different forming techniques, four different base glasses, (Roman, HIMT, Foy 2.1 and Foy 2.1/HIMT) and numerous colouring and opacifying materials point to a well-established and extensive network of the Roman trade, in which Aiano evidently participated. The majority of the beads and glass paste sectilia can be related to the monumentalization phase of the villa in the 4^th to 5^th century A.D. and they represent a sample of the dynamics that regulated the production and circulation of glass pastes in a broad Mediterranean and European context.