THE PROVENANCE OF IRON ARTEFACTS FROM MANCHING: A MULTI‐TECHNIQUE APPROACH*

Iron finds from the Celtic oppidum of Manching in southern Bavaria (Germany) are analysed in view of their possible provenance. The exceptional size and the location of Manching are usually attributed to the presence of abundant iron ores in its vicinity. After a review of previous approaches for source determination of iron artefacts, we introduce lead isotope analysis as a new approach. However, only by combining the trace element patterns of slag inclusions and iron metal with lead isotope ratios in the metal is it possible to distinguish various iron ore formations near Manching. As a result, it turns out that, indeed, the most obvious ones—namely, bog ores near the Danube—constituted the main resources for iron production at Manching. It was even possible to select one occurrence as the most likely ore source.     

INVESTIGATION ON ROMAN LEAD GLAZE FROM CANOSA: RESULTS OF CHEMICAL ANALYSES*

Five sherds of green glazed pottery excavated at Canosa (Apulia) in Italy have been analysed by scanning electron microscopy. The aim of the investigation was to determine the chemical composition of the glazes and thus obtain information on the methods used in their production. The glazes were all of the high‐lead type, coloured green by the addition of copper. Intermediate layers, observed at the interface between the glaze and body and giving the appearance of an applied white slip, were the result of the crystallization of lead feldspar from the molten glaze. Non‐calcareous clays were used in the production of the pottery bodies. Concentration profiles from the glaze exterior to the body suggested that the glazes were produced by applying a suspension consisting of lead oxide plus silica to the bodies. On the basis of the glaze and body compositions, it is suggested that the Canosa glazed pottery was produced locally.     

DISCOVERY,PRODUCTION AND USE OF TIN‐BASED OPACIFIERS IN GLASSES,ENAMELS AND GLAZES FROM THE LATE IRON AGE ONWARDS: A REASSESSMENT*

Tin‐based opacifiers (lead stannate yellow and tin oxide white) were first used in glass production for a short period in Europe from the second to the first centuries bc , and then again throughout the Roman and Byzantine Empires from the fourth century ad onwards. Tin oxide was also used in the production of Islamic opaque glazes from the ninth century ad , and subsequently in enamels applied to Islamic and Venetian glasses from the 12th century ad onwards. A selection of published analytical data for the tin‐opacified glasses, enamels and glazes is summarized, and the methods used in their production are reassessed. The phase transformations occurring when mixtures of lead oxide, tin oxide and silica are fired are investigated with high temperature X‐ray diffraction (XRD) using a synchrotron radiation source, and these results are used to explain the observed differences in the glass, enamel and glaze compositions. Possible reasons for the use of tin‐based opacifiers in the second to first centuries bc , and for the switch from antimony‐ to tin‐based opacifiers in the fourth century ad are suggested, and the possible contexts in which tin‐based opacifiers might have been discovered are considered. The introduction of tin‐opacified glazes by Islamic potters in the ninth century ad is discussed in terms of technological transfer or independent invention.     

GLAZED CERAMIC MANUFACTURING IN SOUTHERN TUSCANY (ITALY): EVIDENCE OF TECHNOLOGICAL CONTINUITY THROUGHOUT THE MEDIEVAL PERIOD (10TH–14TH CENTURIES)*

Archaeometric investigation allowed the characterization of two important classes of ceramics: ‘vetrina sparsa’ and ‘invetriata grezza’. Their archaeological peculiarity makes them particularly suited for tracing the evolution of glaze manufacturing in southern Tuscany throughout the medieval period (10th–14th centuries). These ceramics were found in different sites of historical importance, and also from a mining perspective. Local copper, lead, zinc and iron mineralizations supported the growth of several settlements in the vicinity of the mines. The many castles and different archaeological finds (ceramics, glazed ceramic, slag etc.) attest to the intense mineral exploitation of the area from at least the first millennium bc up to the modern period. In light of these geological and archaeological characteristics, archaeometric investigation was intended to provide insight into ancient technical knowledge of ceramic glazing and to determine the source area for raw materials in the medieval period (10th–14th centuries). Ceramic bodies were analysed through OM, XRDp, SEM–EDS and XRF, while coatings were investigated through SEM–EDS. Mineralogical, petrographic and chemical analyses revealed slightly different preparation and firing processes for the two classes of ceramics. These data suggest the continuity through the centuries of the ‘vetrina sparsa’ and ‘invetriata grezza’ production technology. The mineralogical phases, such as monazite, xenotime, zircon, barite, Ti oxide, ilmenite, titanite, tourmaline and ilvaite, and the lithic (intrusive and volcanic) fragments detected within the ceramic bodies suggest a source area in the vicinity of the Campiglia mining district. Lastly, the presence of Cu–Zn–Pb (Ag) and Fe sulphide mineralizations (materials used to produce glaze) in the area supports the hypothesis of local manufacture.     

TRACING THE RESOURCES OF IRON WORKING AT ANCIENT SAGALASSOS (SOUTH‐WEST TURKEY): A COMBINED LEAD AND STRONTIUM ISOTOPE STUDY ON IRON ARTEFACTS AND ORES*

Lead and strontium isotope analyses were performed by thermal ionization mass spectrometry (TIMS) on Roman to Byzantine iron artefacts and iron ores from the territory of ancient Sagalassos (south‐west Turkey), to evaluate Pb and Sr isotopes for provenance determination of ores for local iron production. It can be demonstrated that for early Roman artefacts and hematite iron ore processed in early Roman times from Sagalassos proper, as well as for magnetite placer sands and early Byzantine raw iron from the territory of the city, Sr isotopes are much less ambiguous than Pb isotopes in providing clearly coherent signatures for ore and related iron objects. Late Roman iron objects were produced from iron ores that as yet remain unidentified. Early Byzantine iron artefacts display more scatter in both their Pb and Sr isotope signatures, indicating that many different ore sources may have been used. Our study demonstrates that iron objects can be precisely analysed for their Sr isotopic composition, which, compared to Pb isotopes, appears to be a much more powerful tool for distinguishing between chronological groups and determining the provenance of raw materials.     

LEAD ISOTOPE AND TRACE ELEMENT ANALYSIS IN THE STUDY OF OVER A HUNDRED SOUTH INDIAN METAL ICONS*

Technical investigations were made on 130 South Indian statuary images and a few miscellaneous artefacts mostly sampled from the Government Museum, Madras, India, and from the Victoria and Albert Museum and the British Museum, UK. Lead isotope investigations were attempted on 60 of these, and compositional analysis for 18 elements on 115, using inductively coupled plasma optical emission spectroscopy; thus, for 40 objects both lead isotope and trace element analysis was done. From the isotopic and elemental framework, insights are obtained into some art-historical problems of images and artefacts of the Pre-Pallava, Pallava, Chola (i.e., Vijayalaya Chola), Later Chola (i.e., Chalukya-Chola and Later Pandya), Vijayanagara (and Early Nayaka) and Later Nayaka (and Maratha) dynasties, spanning the Early Christian era to the nineteenth century, along with a few other regional styles. Inferences are also made regarding provenance of the lead and the early use of zinc and brass in the early historic period (c. fourth century BC—fourth century AD).     

Model‐based Cluster Analysis of Artefact Compositional Data

Cluster analysis is the most widely used multivariate technique in archaeometry, with the majority of applications being exploratory in nature. Model‐based methods of clustering have their advocates, but have seen little application to archaeometric data. The paper investigates two such methods. They have potential advantages over exploratory techniques, if successful. Mixture maximum‐likelihood worked well using low‐dimensional lead isotope data, but had problems coping with higher‐dimensional ceramic compositional data. For our most challenging example, classification maximum‐likelihood performed comparably with more standard methods, but we find no evidence to suggest that it should supplant these.     

LEAD ISOTOPE RATIOS AS A MEANS OF SOURCING ANTHROPOGENIC LEAD IN ARCHAEOLOGICAL SOILS: A PILOT STUDY OF AN ABANDONED SHETLAND CROFT*

Results from soil chemical analysis have demonstrated enhanced concentrations of lead (Pb) associated with archaeological sites. However, interpretation of these Pb concentrations is difficult because of the multitude of possible Pb sources. This pilot study of an abandoned croft in Shetland suggests that Pb isotope ratios have the potential to identify sources of anthropogenic Pb. The results highlight two different Pb associations. The first includes hearth, house floor and house overburden soils, with end members of fuel materials and an unidentified material with a low (< 1.126) ²⁰⁶ Pb/ ²⁰⁷ Pb ratio. The second includes byre, kailyard (garden) and arable soils, with end members of hearth materials and local wind‐blown sand.     

MINERALOGY OF MEDIEVAL SLAGS FROM LEAD AND SILVER SMELTING (BOHUTÍN,PŘÍBRAM DISTRICT,CZECH REPUBLIC): TOWARDS ESTIMATION OF HISTORICAL SMELTING CONDITIONS*

Slags from the Pb/Ag medieval (14th century) smelting plant located at Bohutín, P?íbram district, Czech Republic, were studied from the mineralogical and geochemical points of view. Two types of slags were distinguished: (i) quenched slags formed mainly by Pb‐rich glass and unmelted residual grains of SiO₂ and feldspars, and (ii) crystallized slags mainly composed of Fe‐rich olivine (fayalite) and glass. The mean log viscosity value of the slags calculated for 1200°C was 2.119 Pa s. The morphology of olivine crystals was used to estimate the cooling rates of the melt, for some slags indicating rates > 1450°C/h. The projection of the bulk composition of slags onto the SiO₂–PbO–FeO ternary system was used for rough temperature estimates of slag formation, lying probably between 800 and 1200°C.