Pliny the Elder and Sr–Nd isotopes: tracing the provenance of raw materials for Roman glass production

In Roman and Byzantine times, natron glass was traded throughout the known world in the form of chunks. Production centers of such raw glass, active from the 4th to 8th century AD, were identified in Egypt and Syro-Palestine. However, early Roman primary glass units remain unknown from excavation or scientific analysis. The ancient author Pliny described in 70 AD that besides Egyptian and Levantine resources, also raw materials from Italy and the Gallic and Spanish provinces were used in glass making. In this study, the primary provenance of 1st–3rd century AD natron vessel glass is investigated. The use of combined Sr and Nd isotopic analysis allows the distinguishing and characterizing of different sand raw materials used for primary glass production. The isotope data obtained from the glass samples are compared to the signatures of primary glass from known production centers in the eastern Mediterranean and a number of sand samples from the regions described by Pliny the Elder as possible sources of primary glass. Eastern Mediterranean primary glass has a Nile dominated Mediterranean Nd signature (higher than −6.0 ? Nd), while glass with a primary production location in the western Mediterranean or north-western Europe should have a different Nd signature (lower than −7.0 ? Nd). Most Roman glass has a homogeneous ⁸⁷Sr/⁸⁶Sr signature close to the modern sea water composition, likely caused by the (intentional) use of shell as glass raw material. In this way, strontium and neodymium isotopes now prove that Pliny's writings were correct: primary glass production was not exclusive to the Levant or Egypt in early Roman days, and factories of raw glass in the Western Roman Empire will have been at play.     

δ18O measurements of archaeological glass (Roman to Modern age) and raw materials: possible interpretation

The present paper reports results from a systematic study of oxygen isotopic compositions for glass samples from various archaeological sites (i.e., Iulia Felix, Grado, Vicenza, Pozzuoli and Modena in Italy, and Derrière Sairoche in Switzerland) and dated from the Roman period to the 18th century AD, as well as of some raw materials that may have been used for their production. The analysed samples differ essentially in the type of flux, using Roman and high Medieval glass natron and late Medieval and modern glass plant ash, soda and potash, respectively. The aim of this study was to amplify the database of oxygen isotope data for various archaeological glasses and to identify isotopic trends indicating different raw materials, production technology, and/or provenance. Results indicate that natron glass samples of various provenance and age have consistently higher δ¹⁸O values than plant ash ones (about 15.5‰ vs 13.0‰), probably due to the different flux, highly ¹⁸O-enriched in the case of natron. Isotopic data on Belus and Campanian sands, the types mentioned by Pliny for glass production, show that they have similar isotopic composition. Taking into account the oxygen isotopic composition of Roman glass, the “positive natron effect”, and the negligible influence of small amounts of manganese and antimony containing decolourisers, the suitability of both sources for glass production is verified, supporting the hypothesis of multiple sand sources. Notwithstanding this, the isotopic similarity between Belus and Campanian sands prevents us from identifying the starting material from the δ¹⁸O of the final product. In the case of plant ash used as flux, it is not possible to distinguish between soda and potash plant ash, because the addition of ash did not contribute isotopically heavy oxygen and the silica source is presumed to be comparable in the analysed samples. The isotopic data of the present study are also compared with those already published in the literature, and possible interpretations on their analogies and differences are discussed.     

Glassmaking using natron from el-Barnugi (Egypt); Pliny and the Roman glass industry

Pliny the Elder describes the discovery of a process for making natron glass, which was widely used for much of the first millennium bc and ad. His account of glassmaking with natron has since been corroborated by analyses of archaeological glass and the discovery of large-scale glass production sites where natron glass was made and then exported. Analyses of Egyptian natron have shown it to be a complex mixture of different sodium compounds, and previous experiments to make glass with Egyptian natron have been unsuccessful. Here, natron from el-Barnugi in the Egyptian Nile delta, a site which also probably supplied Roman glassmakers, is used to produce glass. The experiments show that high-quality glass, free of unreacted batch or bubbles, could have been produced from natron in its unprocessed form in a single stage, that larger quantities of natron would be required than has previously been anticipated, that the presence of different sodium-containing compounds in the deposit aided melting, and that negligible waste is produced. The implications for the identification of glass production sites, for the organisation of trade and for the supply of natron within and outside Egypt are discussed in the light of Pliny’s accounts.     

THE PROVENANCE AND TECHNOLOGY OF NEAR EASTERN GLASS: OXYGEN ISOTOPES BY LASER FLUORINATION AS A COMPLEMENT TO STRONTIUM*

The ability to make rapid measurements on small samples using laser fluorination enhances the potential of oxygen isotopes in the investigation of early inorganic materials and technologies. δ¹⁸O and ⁸⁷Sr/⁸⁶Sr values are presented for glass from two primary production sites, four secondary production sites and a consumer site in the Near East, dating from Late Antiquity to the medieval period. δ¹⁸O is in general slightly less effective than ⁸⁷Sr/⁸⁶Sr in discriminating between sources, as the spread of measured values from a single source is somewhat broader relative to the available range. However, while ⁸⁷Sr/⁸⁶Sr is derived predominantly from either the lime‐bearing fraction of the glass‐making sand or the plant ash used as a source of alkali, δ¹⁸O derives mainly from the silica. Thus the two measurements can provide complementary information. A comparison of δ¹⁸O for late Roman – Islamic glasses made on the coast of Syria–Palestine with those of previously analysed glasses from Roman Europe suggests that the European glasses are relatively enriched in ¹⁸O. This appears to contradict the view that most Roman glass was made using Levantine sand and possible interpretations are discussed.     

ROMAN AND MEDIEVAL GLASS FROM THE ITALIAN AREA: BULK CHARACTERIZATION AND RELATIONSHIPS WITH PRODUCTION TECHNOLOGIES*

Compositional investigations were performed on 81 Roman and medieval glass fragments (first to 14th centuries ad ) from four Italian archaeological sites. The samples were soda–lime–silica in composition, with natron as flux for the Roman and early medieval glass samples, and with plant ash as flux for the late medieval ones. The varying colours are due to the differing FeO, Fe ₂ O ₃ , MnO and Sb ₂ O ₃ contents. Hierarchical cluster analysis identified six compositional groups related to age, which were compared with those found in the literature. In this way, technological continuity from the Roman to the early medieval period and the appearance of plant ash technology in the ninth century, 200 years in advance of the period previously believed, are demonstrated.     

MEDIEVAL GLASS FROM ROCCA DI ASOLO (NORTHERN ITALY): AN ARCHAEOMETRIC STUDY

An archaeometric study was performed on 33 medieval glass samples from Rocca di Asolo (northern Italy), in order to study the raw materials employed in their production, identify analogies with medieval glass from the Mediterranean area and possible relationships between chemical composition and type and/or production technique, contextualize the various phases of the site and extend data on Italian medieval glass. The samples are soda–lime–silica in composition, with natron as flux for early medieval glasses and soda ash for the high and late medieval ones. Compositional groups were identified, consistent with the major compositional groups identified in the western Mediterranean during the first millennium AD . In particular, Asolo natron glass is consistent with the HIMT group and recycled Roman glass; soda ash glass was produced with the same type of flux (Levantine ash) but a different silica source (siliceous pebbles, and more or less pure sand). Cobalt was the colouring agent used to obtain blue glass; analytical data indicate that at least two different sources of Co were exploited during the late medieval period. Some data, analytical and historical, suggest a Venetian provenance for the high/late medieval glass and a relationship between type of object (beaker or bottle) and chemical composition.     

The colourless glass of Iulia Felix

The present paper focuses on 85 colourless glass samples selected among the recycling cullet found in the Roman ship Iulia Felix, wrecked off the town of Grado (province of Udine, North Italy) in the first half of the 3rd century AD, with the aim of examining their chemical composition, investigating similarities within and between the three established group types (cups, plates, bottles), and comparing the chemical compositions of the Iulia Felix glass samples with groups reported in the literature for colourless glasses. Chemical analyses, performed by X-ray fluorescence and electron microprobe, indicate that the samples were all soda-lime–silica in composition, with natron as flux, according to typical Roman production technologies. Chemical data are treated by multivariate statistical tools, such as cluster analysis and the NPC test. Statistical analyses allow us to conclude that the Iulia Felix colourless glasses were produced by at least two different production technologies related to group type, confirming the trend recently identified in the literature. In addition, the trace element patterns (particularly those of Sr, Zr and Ba) suggest that beach sand with differing amounts of alkali feldspars was predominantly used as raw material, like those specifically mentioned by Pliny for glass production.     

Radical changes in Islamic glass technology: evidence for conservatism and experimentation with new glass recipes from early and middle Islamic Raqqa,Syria*

The chemical analysis of excavated glass fragments from dated archaeological contexts in Raqqa, Syria, has provided a detailed picture of the chemical compositions of artefacts deriving from eighth to ninth and 11th century glassmaking and glassworking activities. Evidence for primary glass production has been found at three excavated sites, of eighth to ninth, 11th and 12th century dates; the first two are discussed here. The 2 km long industrial complex at al‐Raqqa was associated with an urban landscape consisting of two Islamic cities (al‐Raqqa and al‐Rafika) and a series of palace complexes. The glass fused and worked there was presumably for local as well as for regional consumption. Al‐Raqqa currently appears to have produced the earliest well‐dated production on record in the Middle East of an Islamic high‐magnesia glass based on an alkaline plant ash flux and quartz. An eighth to ninth century late ‘Roman’/Byzantine soda–lime recipe of natron and sand begins to be replaced in the eighth to ninth century by a plant ash – quartz Islamic soda–lime composition. By the 11th century, this process was nearly complete. The early Islamic natron glass compositional group from al‐Raqqa shows very little spread in values, indicating a repeatedly well‐controlled process with the use of chemically homogeneous raw materials. A compositionally more diffuse range of eighth to ninth century plant ash glass compositions have been identified. One is not only distinct from established groups of plant ash and natron glasses, but is believed to be the result of experimentation with new raw material combinations. Compositional analysis of primary production waste including furnace glass (raw glass adhering to furnace brick) shows that contemporary glasses of three distinct plant ash types based on various combinations of plant ash, quartz and sand were being made in al‐Raqqa during the late eighth to ninth centuries. This is a uniquely wide compositional range from an ancient glass production site, offering new insights into the complexity of Islamic glass technology at a time of change and innovation.     

The Production and Circulation of Carthaginian Glass Under the Rule of the Romans and the Vandals (Fourth to Sixth Century ad): A Chemical Investigation

Fifty‐seven glass samples from Carthage dating to the fourth to sixth century ad were analysed using the electron microprobe. The results show that these samples are all soda–lime–silica glass. Their MgO and K₂O values, which are below 1.5%, suggest that they were made from natron, a flux that was widely used during the Roman period. The major and minor elements show that these samples can be divided into four groups, three of which correspond to the late Roman period glass groups that were found throughout the Roman Empire: Levantine I, and ‘weak’ and ‘strong’ HIMT. Of particular interest is our Group 2, which is technologically and compositionally similar to HIMT glass and the CaO and Al₂O₃ values of which are similar to those of Levantine I. Glass of similar composition has been reported by several authors and is predominantly found dating from the late fifth to seventh century. This could represent a ‘new’ glass group; therefore further study is needed to determine its origin. Also, this study suggests that the Vandal invasion in North Africa did not disrupt the glass trade between Carthage and the Levantine coast.     

Strontium Isotopes in the Investigation of Early Glass Production: Byzantine and Early Islamic Glass from the Near East*

⁸⁷Sr/⁸⁶Sr ratios have been determined for glasses from four production sites, dated to between the sixth and the 11th centuries, in the Eastern Mediterranean region. On the basis of elemental analyses, the glasses at each location are believed to have been melted from different raw materials. Two glass groups, from Bet Eli‘ezer and Bet She‘an, in Israel, are believed to have been based upon mixtures of Levantine coastal sands and natron, and have ⁸⁷Sr/⁸⁶Sr ratios close to 0.7090, plus high elemental strontium, confirming a high concentration of modern marine shell (⁸⁷Sr/⁸⁶Sr ~ 0.7092) in the raw materials. The isotopic compositions of these two groups of glasses differ slightly, however, probably reflecting a varying ratio of limestone to shell because the sands that were utilized were from different coastal locations. Natron‐based glasses from a workshop at Tel el Ashmunein, Middle Egypt, have ⁸⁷Sr/⁸⁶Sr values of 0.70794–0.70798, and low elemental strontium, consistent with the use of limestone or limestone‐rich sand in the batch. High‐magnesia glasses based on plant ash, from Banias, Israel, have ⁸⁷Sr/⁸⁶Sr values of 0.70772–0.70780, probably reflecting the isotopic composition of the soils that were parental to the plants that were ashed to make the glass. Strontium and its isotopes offer an approach to identifying both the raw materials and the origins of ancient glasses, and are a potentially powerful tool in their interpretation.     

The composition of colourless glass: a review

A total of 1496 investigated colourless glass analyses have been collected with the aim of achieving a clear geographical, typological, chronological and compositional overview on this particular type of glass. Based on manganese and antimony contents, four main groups were characterised: naturally colourless, Mn-decoloured, Sb-decoloured and Mn/Sb-decoloured. Main achievements relates to the chronological distribution of manganese and antimony technologies, the former being associated to a long lasting technology which culminated during the Late Antique period while the latter being practically absent after the 8^th century AD, being at its acme during the Roman imperial period. Except for naturally colourless glass, glass-making technology mostly implied the use of impure sands and natron, relegating the other components to a virtually irrelevant ?presence?, except during the  Late Antique and Medieval periods.     

COMPOSITIONAL VARIATION IN ROMAN COLOURLESS GLASS OBJECTS FROM THE BOCHOLTZ BURIAL (THE NETHERLANDS)*

We investigated the major and trace element composition and Pb and Sr isotope characteristics of a series of about 20 colourless glass objects from a single high‐status Roman burial from the Netherlands (Bocholtz). The major elements show a relatively homogeneous group, with one outlier. This is corroborated by the Sr isotopes. Based on the Sb and Pb content, three major groups can be discerned, with two other outliers. This grouping is corroborated by the contents of the trace elements Bi, Sn, Ag, As and Mo, and by variations in lead isotopic ratios. On the basis of these results, we conclude that the glass of all objects was probably made with sand and lime from the same source. The variation in trace elements and lead isotope composition is most likely the result of variations in the composition of the sulphidic antimony ore(s) that were used to decolourize the glass. The composition of the Bocholtz glass is compared with that of other Roman glass, and implications for production models, trade and use of colourless glass objects are discussed. On the basis of isotopic and major element variation, we conclude that the antimony ore presumably originated from different mines.     

New insights into Byzantine glass technology from loose mosaic tesserae from Hierapolis (Turkey): PIXE/PIGE and EPMA analyses

This study focuses on the Byzantine glass tesserae from Hierapolis (Phrygia, central Turkey). Fifty-seven samples of loose tesserae from two sites in the town (the theatre and the church of St. Philip) are analysed by particule-induced X-ray emission and particule-induced gamma ray emission and electron probe X-ray microanalysis to obtain the chemical composition and identify the colourants and opacifiers. The aims are to add new information to the scant knowledge of the Byzantine glassmaking technology, to constrain the chronology of the mosaics and to trace the supply routes of the tesserae. In the destruction layers of the theatre, tesserae produced following the Roman glassmaking technology (natron glass opacified by calcium and lead antimonate) were found. They were made using a Levantine 1 raw glass, generally attributed to the early Byzantine period (fifth to sixth c.). In the church, the samples attest a technological change from Roman tradition, and a complex pattern according to building history (two phases are attested, probably in the sixth and eighth to ninth c.), and a multiplicity of supply. Three glass types and some recipes not attested before in this chronological range for the production of tesserae are documented, such as the use of a local low-chlorine natron glass for the production of black and red tesserae, the blue colouring by a source of cobalt with zinc in a natron glass tessera and the opacification with tin oxide (both in a lead-free and in a high-lead natron glass), as well as with quartz.     

THE SECTILIA PANELS OF FARAGOLA (ASCOLI SATRIANO,SOUTHERN ITALY): A MULTI‐ANALYTICAL STUDY OF THE GREEN,MARBLED (GREEN AND YELLOW), BLUE AND BLACKISH GLASS SLABS

Analyses at the Cu–K, Fe–K and Mn–K edge were performed to study the green, marbled (green and yellow), blue and blackish (deep greyish olive green) glass slabs decorating three sectilia panels from the archaeological site of Faragola. Results indicate that all slabs were made by mixing siliceous sand with natron, sometimes probably mixed with small percentages of plant ash. Cu²⁺ and Pb antimonates should be responsible for the opaque green colours. The dark green and yellow portions of the marbled slabs are respectively comparable to the slabs comprising only one of these colours. Cu²⁺ together with Ca antimonates probably produced light blue slabs, whereas cobalt was used to produce dark blue slabs. We consider it possible that the abundance ratio of Fe²⁺/Fe³⁺ and the complex Fe³⁺S^2? would have an effect on the blackish slabs. The contribution of Mn cannot be ascertained even if it could have played a role in darkening glass colour. The comparison between the chemical composition of Faragola samples and several glass reference groups provided no conclusive evidence of provenance; whereas, the presence of a secondary local workshop can be hypothesized.     

The coloured glass of Iulia Felix

Research on the cargo of glass in the Roman ship Iulia Felix, wrecked off the town of Grado (province of Udine, North Italy) in the first half of the 3rd century AD and composed of recycling cullet carefully selected for colour and type, provided much information on Roman glass production technology. A combined approach, involving analytical, statistical, and archaeological evidence, included chemical analyses (X-ray fluorescence, electron microprobe), which indicated that the coloured samples were all soda-lime-silica in composition, with natron as flux, although cluster analysis identified many compositional groups. Comparisons among the compositional groups of the colourless glass, previously studied, and those of coloured glass showed that the production technologies of the colourless glass vessels constitute two well-defined technological end-members, also related to group types, into which those of the coloured glass samples fall. In particular, coloured glass samples—all bottles, low-status vessels—were produced with mainly beach siliceous-feldspar-calcareous sand. However, unlike the colourless type, strict control of raw materials and decolourising processes was not adopted, and recycling was also practised, as demonstrated by the Sb₂O₃, Cu and Pb patterns. In this context, a “recycling index” (RI)—[(Sb₂O₃)_X/(Sb₂O₃)_Ref] * 100—is proposed here for the first time, in order to quantify the extent of recycling of antimony colourless glass in the batch. RI is valid for glass containing abundant Sb from an end-member of colourless glass. In conclusion, although it cannot be stated unequivocally that the identified compositional trends are related to different production centres or different raw materials, the strong evidence of compositional variability among all the Iulia Felix glass samples, both colourless and coloured, supports the dispersed production model for Roman glassware and the common practice of recycling in Roman imperial times, especially for low-status vessels.     

WHITE,VEINED MARBLE FROM ROMAN AMMAIA (PORTUGAL): PROVENANCE AND USE

A multimethod approach using petrography and strontium (Sr) isotopic analysis was applied to determine the geological source of 17 marble artefacts from the Roman town of Ammaia (Portugal). All samples are calcitic, with dolomite, quartz and muscovite as accessory minerals. The marbles are characteristically medium‐grained with a maximum grain size (MGS) between 0.98 mm and 1.82 mm, have a heteroblastic texture, and have curved to embayed calcite grain boundaries. ⁸⁷Sr/⁸⁶Sr values of marble leachates range from 0.708488 to 0.708639. Comparison with Hispanic and Mediterranean marbles suggests the Estremoz Anticline as the most likely source for the Ammaia marble, especially for architectural marble. This hypothesis is supported by the geographical proximity of the Estremoz marble district and the long and expensive overland transport required for other marbles to reach Ammaia.     

Analyses of colourless Roman glass from Binchester,County Durham

Forty samples of Roman colourless glass tableware from Binchester, dating from the 1st to mid-3rd centuries AD, were analysed using ICP spectrometry and parallels were sought with similar studies of Roman glass from Colchester and Lincoln [C. Mortimer, M.J. Baxter, Analysis of Samples of Colourless Roman Vessel Glass from Lincoln, Ancient Monuments Laboratory Report 44/1996, 1996; H.E.M. Cool, J. Price, Roman vessel glass from excavations in Colchester, 1971–1985, Colchester Archaeological Report 8, Colchester Archaeological Trust Ltd. and English Heritage, Colchester, 1995; M.P. Heyworth, M.J. Baxter, H. Cool, Compositional Analysis of Roman Glass from Colchester Essex, Ancient Monuments Laboratory Report 53/1990, 1990]. Some samples from the Binchester, Colchester and Lincoln (BCL) groups were re-analysed using energy dispersive spectrometry (SEM–EDS) and the results were used to compensate for the differences in reproducibility between the ICP data sets, so that these could be directly compared. The majority of the glass from all three sites was similar but some distinct compositional characteristics were identified that were specific to certain types of ware. There were differences in the concentration of lead in the samples, which appear to be to some extent chronological. The compositional data for the glass from Binchester, Colchester and Lincoln were also compared to literature data for various types of natron glass produced in the 1st millennium AD.     

WASTE GLASS,VESSELS AND WINDOW‐PANES FROM THAMUSIDA (MOROCCO): GROUPING NATRON‐BASED BLUE–GREEN AND COLOURLESS ROMAN GLASSES

A collection of window‐panes, vessels and alleged waste from Thamusida has been investigated by OM, SEM–EDS, ICP–MS, ICP–OES and XAS at the Fe–K and Mn–K edges. Glass samples have been characterized as natron‐based soda–lime–silica glasses, with low magnesium and low potassium. The results have been compared with 43 reference groups available for ‘naturally coloured’ and colourless glasses of both Roman and later ages. Two main types were distinguished: RBGY 1 (R oman B lue–G reen and Y ellow 1) and RBGY 2 (R oman B lue–G reen and Y ellow 2). Given their compositional similarity to the Levantine I or, to a lesser extent, HIMT glasses, the Syrian–Palestinian coast for RBGY 2 and Egypt for the RBGY 1 have been suggested for their provenance. Most Thamusida samples have been assigned to the RBGY 2 type. A small group of Thamusida colourless vessels was included into the RC (R oman‐C olourless) compositional field; the latter still being defined. The alleged waste pieces may define a local production that should have been of secondary type. The investigations performed on local raw materials seem to discount the possibility of a primary glass‐making site. XAS measurements determined that Fe²⁺ contents ranging between 30 and 52% are able to assure an aqua blue colour; below 30%, the glasses turn light green or light yellow.     

An archaeometric study of Hellenistic glass vessels: evidence for multiple sources

In the present study, 53 glass fragments from core-formed vessels and 3 glass beads are investigated using SEM/EDX, EPMA and LA-ICP-MS. All samples were excavated in the Latin settlement of Satricum in central west Italy and apart from two, were found in the so-called fourth–third c. BC Hellenistic Votive deposit, also known as Votive Deposit III, discovered in front of the sanctuary of Mater Matuta on top of the acropolis. The analytical results indicate that the glass from Satricum is a typical soda-lime-silica type with natron used as a flux. Its chemical compositions display a relatively low compositional variation. Small differences in the concentrations of major and minor oxides (SiO₂, Al₂O₃, CaO and Fe₂O₃) and in trace elements (Sr, Zr and Nd) between individual samples suggest the use of different types of raw materials, especially sand. In turn, this suggests that the glass derived from more than one glass making centre. The combined investigation of colourants (Co, Cu and Mn) reinforces and confirms the idea that glass from Satricum was made using different manufacturing traditions during the Hellenistic period.     

A non-invasive study of Roman Age mosaic glass tesserae by means of Raman spectroscopy

Ancient mosaic tesserae are a range of materials of very varied and complex nature, including pottery, stone and glass. Raman spectroscopy is a powerful tool for the analysis of all these kinds of materials. In the particular case of glasses, this technique can be used both for a study of surface weathering and for the characterization of bulk structure, but it has not yet been extensively used for the characterization of mosaic glass tesserae. We carried out Raman analyses on a set of Roman and Late Antiquity period mosaic glass samples, which allowed a good characterization of both the glass matrix and the crystalline inclusions. All the samples show the typical Raman signatures of soda-lime-silicate glasses. Several crystalline phases were also identified, being relics of raw materials used during the glass manufacturing process, such as quartz and feldspars, or linked to the glass color/opacification, such as bindheimite and cuprite. The analyses also led to the identification in some blue, turquoise and green tesserae of calcium antimonate, whose Raman signature has only recently been recognized in the scientific literature on mosaic glasses. Some emphasis is given to the analysis of red lead-containing tesserae, colored with Cu⁺ ions or even Cu⁰ (or Au⁰) metal nanoparticles. Samples with peculiar compositions, as well as “modern” (and restoration) samples, could quite easily be distinguished from the ancient ones by their Raman spectra.