REVISITING “TIN IN SOUTH-EASTERN EUROPE?”

UDC: 902.62"633/634"(497.11) 903'1"633/634"(497.11) https://doi.org/10.2298/STA1969009B Original research article WAYNE POWELL, Department of Earth and Environmental Science, Brooklyn College OGNJEN MLADENOVIĆ, Institute of Archaeology, Belgrade STEFFANIE CRUSE, Colorado School of Mines, Golden, Colorado H. ARTHUR BANKOFF, Department of Anthropology and Archaeology, Brooklyn College RYAN MATHUR, Department of Geology, Juniata College REVISITING “TIN IN SOUTH-EASTERN EUROPE?” e­mail: wpowell@brooklyn.cuny.edu Abstract. – The important role of the Balkans in the origin and development of metallurgy is well established with respect to copper. In addition, Aleksandar Durman, in his 1997 paper “Tin in South-eastern Europe?”, essentially initiated studies into the role of the Balkans in Europe’s Bronze Age tin economy. He identified six geologically favourable sites for tin mineralisation and associated fluvial placer deposits in the former Yugoslavian republics, and suggested that these may have added to the tin supply of the region. The viability of two of these sites has been confirmed (Mt Cer and Bukulja, Serbia) but the exploitation potential for the other locations has remained untested. River gravels from these four sites (Motajica and Prosara in Bosnia and Herzegovina; Bujanovac in Serbia; Ogražden in North Macedonia) were obtained by stream sluicing and panning. The sites of Prosara and Bujanovac were found to be barren with respect to cassiterite (SnO2). Streams flowing from Motajica and Ogražden were both found to contain cassiterite, but in amounts several orders of magnitude less than at Mt Cer and Bukulja. Although it is possible that minor tin recovery occurred at Motajica and Ogražden, it is unlikely that they could have contributed meaningfully to regional tin trade. This is supported by the fact that the isotopic signature (δ124Sn) of cassiterite from Motajica is highly enriched in light isotopes of tin compared to that associated with Late Bronze Age artefacts of the region. Key words. – Cassiterite, Placer, Tin, Bronze Age, Balkans, Sn Isotopes T he quest for the origin of the earliest metallurgy in southeast Europe dates to the first half of the 20th century when O. Davies published descri­ ptions of the remains of two mining shafts in Jar­ movac in south­western Serbia.1 The period after the Second World War was marked by papers that discus­ sed the prehistoric mining in the context of geology and possible exploitation.2 The discovery and excava­ tions of the prehistoric copper mines of Rudna Glava in eastern Serbia and Ai Bunar in Bulgaria in the late ’60s, and Mali Šturac in the ’80s,3 resulted in several publications that addressed various aspects of prehisto­ ric copper mining, such as the material culture, chro­ nological positioning, and technological processes4, primarily based on stylistic and typological analyses of archaeological material (e.g. potsherds, crucibles, 85 copper beads and malachite remains) from Neolithic and Eneolithic sites.5 Papers focused on physical­ chemical analyses followed in the ’90s.6 While the 1 Davies 1937. Simić 1951; Simić 1969. 3 The excavations at the site of Mali Šturac are still ongoing. For previous results refer to: Antonović, Vukadinović 2012, with cited literature; Antonović 2018, with cited literature. 4 The most complete overview of the history of research and current data and progress is provided in Богосављевић-Петровић 2005 and Antonović 2018. 5 cf. Jovanović 1971; Черних, Радунчева 1972; Јовановић 1974; Jovanović, Ottaway 1976; Јовановић 1978; Černych 1978; Черньх 1978; Jovanović 1985. 6 Pernicka et al. 1993; Begeman et al. 1995. 2 Manuscript received 28th April 2020, accepted 13th October 2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Fig. 1. Potential tin sources identified by A. Durman (1997) Сл. 1. Потенцијални извори калаја према А. Дурману (1997) search for the source of tin, necessary for the produc­ tion of tin­bronze, was ongoing in western Europe throughout the 20th century, it was essentially neglected by local authors in south­eastern Europe. In 1997, Aleksandar Durman published the article “Tin in South-eastern Europe?”. This breakthrough paper proposed that the long­sought source of tin dur­ ing the Bronze Age of south­eastern Europe need not have been the large Variscan deposits of the Erzgebirge or Cornwall, nor an exotic source in Central Asia. Rather, he suggested small placer cassiterite (SnO2) deposits in the Balkans may have provided tin for re­ gional bronze production. Based on communications with Serbian geologist Dr. Antonije Antonović, he identified two Serbian sites as the most likely sources for prehistoric exploitation, Mt Cer7 (20 km east­north­ east of Loznica) and Bukulja8 (8 km west­southwest of Aranđelovac) (Fig. 1). Mining companies and national geological sur­ veys have conducted feasibility studies of these two placer deposits. As reported by Durman (1997), Mi­ hajlović (1978) estimated the gravels associated with Bukulja’s Cigankulja stream contain sufficient cassite­ rite to produce approximately 225 tons of tin metal. Tin placer deposits of Mt Cer were found to be signifi­ cantly larger than those of Bukulja, and this was con­ firmed by Tomić (1991), who projected that the alluvial deposits of the Lešnica and Cernica rivers contain suf­ ficient cassiterite to produce 2,700 tonnes of tin. Thus, 86 only 0.1% of Cer’s current ore reserves could have produced 30 tonnes of bronze, sufficient to supply the entire central Balkan bronze production of the Late Bronze Age. Subsequently, highly disturbed archaeological sites yielding predominantly fragments of Late Eneolithic and Bronze Age pottery were documented adjacent to the richest tin gravels at Mt Cer, and these include rare fragments of technical pottery with tin­bearing vitre­ ous surfaces.9 The Sn isotopic composition of Late Bronze Age Serbian artefacts defines a regional clus­ ter south of the Danube River and west of the Morava River (coincident with the location of Mt Cer) that is consistent with those of the ores of Cer’s Milinska, Cernica, and Lešnic rivers.10 Accordingly, Mason et al. (2020) concluded that mining activities at Mt Cer likely contributed significantly to tin metal produc­ tion of the region at that time. In addition to Mt Cer and Bukulja, Durman (1997) identified four other locations of interest.11 He repor­ ted an unpublished account of detrital cassiterite in rivers at Bujanovac in Serbia, and a report by Tućan 7 8 9 10 11 Durman, 1997, 10. Durman, 1997, 8. Huska et al., 2014, 487–488. Mason et al., 2020. Durman, 1997, 9–10. СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Fig. 2. Sample Location Maps (Geology of Ogražden is based on Boev et al. 2002; Geology of Motajica and Prosara is based on Ustaszewski et al. 2010; Geology of Bujanovac is based on geological map of Vranje K34–56) Сл. 2. Локације на којима је вршено узорковање (геолошке подлоге засноване су: Ограждена – на Boev et al. 2002; Мотајице и просаре – на: Ustaszewski et al. 2010; Бујановца – на геолошкој карти Врања, исечак К34–56) (1957) of cassiterite in streams that flow from the Ogražden granite near Strumica in North Macedonia (Fig. 1). In addition, he noted that the granites that lie beneath Motajica and Prosara in northern Bosnia are similar in composition to those of Mt Cer and Bukulja, and so could potentially host similar mineralisation. However, none of these four sites had been investigated in detail to confirm whether detrital cassiterite is pres­ ent, and if so, whether sufficient quantities would have allowed for production of a significant mass of tin. The purpose of this study was to examine each of these sites and evaluate their viability as potential pre­ historic placer tin mining sites. Sampling and Methods The bedrock geology and river drainage pattern of each site was examined to identify the most likely sites for placer tin accumulations (i.e., confluences of 87 rivers with larger watersheds that cross­cut granite bodies). For each stream that had sufficient water flow to allow a sluice to operate, river gravels were sieved (<2 mm) to produce approximately 30 litres of sand, which was then fed through a portable sluice box. The sluice output was panned on site to a “black­sand” concentrate. An Olympus Delta portable x­ray fluo­ rescence device (pXRF) was used to determine the tin content of the “black­sand” concentrate on site. Initial sampling was undertaken at Motajica (M1 through M3) and Prosara (P1 through P3) in June 2017, and reported on by Cruse et al. (2017). The remaining samples, including a resampling of Motajica, were collected in May 2018 when streamflow was more conducive to sluicing. A total of 22 samples were taken from 17 streams at the four study locations (Fig. 2): four streams at Bujanovac (Bogdanovačka, Dragučica, Krševačka and Ljiljanovačka rivers), five at Ogražden СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) (Jazga, Štučka, Suva, Vasilica, unnamed stream), five at Motajica (Brusnički, Ina, Lepenica, Stojkovica, un­ named stream), and three at Prosara (Busovača, Gasnica, Jablanica). In addition, one sample was taken from both Mt Cer (Milinska) and Bukulja (Dugačko) for comparison. At each sample site one bulk sample was taken. Samples were subsequently dried and the heavy mineral concentrate was further purified by extracting light minerals (<2.9 g/cm³) using flotation separation with a solution of sodium polytungstate (3Na2WO4• 9WO3•H2O). The tin content of this concentrate was determined by pXRF. The heavy mineral assemblages were then fed through a Frantz Isodynamic Magnetic Separator after magnetite was removed using a hand magnet. Subsamples were taken at 0.5A, 0.7A, 1.0A, 1.75A, and the remaining non­magnetic fraction, the non­magnetic fraction being that into which cassiterite accumulates. The Sn content of the non­magnetic fraction was determined by pXRF and then mounted on adhesive carbon stubs for SEM examination and mineral identification. A Hitachi TM3030Plus scanning electron micro­ scope operating at 15kV and an Oxford Instruments AZtec energy dispersive spectrometer with the AZtec One software platform were used to identify all heavy mineral grains mounted on each stub. These quantitative analyses were used to define the major element compo­ nents of each mineral grain so that mineral formulae/ identities could be deduced stoichiometrically, in con­ junction with the physical features observable under the SEM (form and cleavage) and binocular microscope (colour, lustre). Emphasis was placed on the identifi­ cation of cassiterite and other Sn­bearing minerals. Isotopic analyses of cassiterite from Cer and Bu­ kulja were presented recently in Mason et al. (2020), and, so, analyses were not repeated in this study. Of the samples from the four remaining sites, only sample M6 from the Lepenica at Motajica contained a suffi­ cient mass of cassiterite to allow for the separation of a cassiterite concentrate that could be used for isotop­ ic analysis. Grains of cassiterite (0.3–0.5 mm) were identified by SEM­EDS analysis and then hand­ picked to form a cassiterite separate of >100 grains. The cassiterite sample was digested following the procedure of Mathur et al. (2017, p17): 0.1 g of ­100 mesh cassiterite powder was mixed with 0.5 g of KCN and heated at 850°C for one hour in graphite crucibles contained within capped alumina crucibles. The re­ sulting reduced Sn metal beads were dissolved in 88 heated ultrapure 11N HCl overnight. A small aliquot of this solution was removed and dried and redigested in ultrapure 1M HCl. This solution was purified using the ion exchange chromatography described in Ballia­ na et al. (2013, 2981–2982) and employed by Mason et al. (2016) and Mathur et al. (2017). Analysis was conducted on the Neptune MC­ ICPMS at Rutgers University. Solutions were measured at 50ppb Sn with 50ppb Sb ICP­MS standard, as de­ scribed in Mathur et al. (2017). Mass bias was correc­ ted for using Sb doped solutions and an exponential mass bias correction defined in Mathur et al. (2017). The corrected values were then bracketed with the NIST 3161A Sn standard (Lot# 07033). One block of 30 ratios was collected. Data is presented relative to the NIST 3161A Sn standard (Lot# 07033) in per mil notation defined as: Instrumentation 2σ error for δ124Sn is 0.02%. Whole procedural 1σ errors for analysis (ample varia­ bility, reduction, dissolution, purification, and analy­ sis) are δ120Sn= 0.08‰ and δ124Sn= 0.16‰. Note that the δ124Sn and δ120Sn are reported relative to 116Sn, with a difference of 8 amu and 4 amu, respectively. Results The results are summarised in Table 1. In the pan­ ned black sand concentrates, tin was detectable in only the Milinska at Mt Cer (1.4 wt%) and the Dugačko at Bukulja (0.9 wt%). Tin was detected in post­flotation heavy mineral separates at Cer (2.6 wt%), Bukulja (1.2 wt%), two streams at Motajica (Brusnički and an unnamed stream, each with 0.33 wt%), and two streams at Ogražden (Jazga 0.31 wt% and Vasilica 0.24 wt%). In the non­magnetic mineral fractions tin was detect­ ed in samples from two additional sites, the Lepenica at Motajica and an unnamed stream at Ogražden. No tin was detected in samples from either Prosara or Bujanovac In addition to Cer and Bukulja, cassiterite was identified in the non­magnetic fraction of one sample from Motajica (M6 Lepenica) which contained 0.47 wt% Sn, and two samples from Ograzden (O3 Suva, O5 unnamed stream), with 0.05 and 0.12 wt% Sn, re­ СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Panned Heavy Liquid Non Magnetic Minerals of Interest nd nd Sample Stream B1 Ljiljavancka nd B2 Krsevacka nd nd nd B3 Bogdanovacka nd nd nd B4 Dragucica nd nd nd B5 Dragucica nd nd nd B6 Dragucica nd nd nd B7 Dragucica nd nd nd M1 Ina nd nd nd Sn­bearing rutile, euxenite, bastnaisite M2 Stojkovica nd nd nd Thorite, scheelite, euxenite, bismuth, brookite M3 Lepenica nd nd nd Bastnaisite M4 Brusnički nd 0.03 M5 Unnamed nd 0.03 M6 Lepenica nd nd 0.47 M7 Stojkovica nd nd nd O1 Jazga nd 0.03 0.04 O2 Štučka nd nd nd O3 Suva nd nd 0.05 O4 Vasilica nd 0.02 nd O5 Unnamed nd nd 0.12 P1 Busovaca nd nd nd Scheelite, wolframite P2 Gasnica nd nd nd Scheelite, wolframite P3 Jablanica nd nd nd Scheelite, wolframite Cer Milinska 1.4 2.6 18.7 Cassiterite, euxenite, microlite Bukulja Dugačko 0.9 1.2 11.6 Cassiterite, wolframite, thorite Cassiterite, ixiolite, euxenite Cassiterite, euxenite, scheelite Cassiterite, thorite, scheelite Table 1. Tin concentrations in various sample fractions in percent, as determined by pXRF, along with key minerals identified through SEM-EDS analysis Табела 1. Концентрације калаја у узорцима различитих фракција, изражени у процентима, концентрације су утврђене путем pXRF анализе, а главни минерали путем SEM-EDS анализе spectively. Although these are the most Sn­rich sam­ ples found in this study, their tin concentration is two to three orders of magnitude less that that found at Mt Cer and Bukulja. Although several other samples from Motajica and Ogražden (M4, M5 and O4) had detectable Sn based on pXRF analysis, no cassiterite was found. This suggests that the Sn occurs as a component in other minerals. This has been documented at Mt Cer, where Sn is present within a microlite­series mineral ([Ca,Sn,U]2[Ta,Nb]2O6(OH,F])12 and within rutile 89 [(Ti,Sn)O2] in sample M1 from the Ina at Motajica. All samples from Bujanovac and Prosara were found to be barren with respect to tin in any form. The cassiterite concentrate from sample M6 from the Lepenica at Motajica yielded mass­dependent iso­ topic compositions of δ120Sn=-0.44‰ and δ124Sn= ­0.83‰, relative to NIST 3161A. This contrasts with the more heavy isotope weak to moderate heavy isotope 12 Powell et al., in press. СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Fig. 3. δ124Sn values for Late Bronze Age bronze artefacts from Serbia and Bosnia and Herzegovina, as reported in Huska et al. (in press) after a -0.2‰ shift to correct for smelt induced fractionation, as compared to cassiterite from Mt Cer, Bukulja, and Motajica. (Cassiterite data for Mt Cer from Huska et al., in press; Bukulja data from Huska et al., in press; Berger et al., 2019) Сл. 3. Вредности δ124Sn за предмете позног бронзаног доба из Србије и БиХ (Huska et al., in press), кориговане са -0.2‰ како би се одстранили ефекти фракционације проузроковане топљењем, у поређењу са вредностима добијеним за каситерит са Цера, Букуље и Мотајице. (Вредности за каситерит преузете су из: за Цер – Huska et al., in press; за Букуљу – Huska et al., in press; Berger et al. 2019) enrichment displayed by similar multi­grain placer samples from Mt Cer and Bukulja (δ124Sn of 0.3 to 0.8‰ at Cer; δ124Sn of ­0.1 to 0.4‰ at Bukulja).13 Conclusions Of the six potential tin placer deposits identified by Durman, Mt Cer and Bukulja are by far the richest, with both yielding percent­level concentrations of tin in the panned black­sand concentrate. In addition, cassiterite was documented to occur in stream sedi­ ments at Ogražden and Motajica, but at concentra­ tions several orders of magnitude lower than Cer and Bukulja. Although containing other rare elements, in­ cluding tungsten, Prosara appears to be barren with re­ spect to tin. The Vranje geological map (K34–56) notes the presence of Sn in the Dragučica River, along the south contact of the Bujanovac pluton. However, four separate samples taken along the same stretch of this river failed to yield any trace of tin, as was the case for the other three rivers samples at Bujanovac. This may indicate that Bujanovac hosts very minor tin minerali­ sation, certainly less than would have been necessary for the development of a mineable placer deposit. Mason et al. (2020) demonstrated through isotopic composition and geographic correlation that ore from Mt Cer likely supplied much of the tin economy of 90 Late Bronze Age Serbia south of the Danube. Although Bukulja is a viable source of tin ore, there remains no archaeological evidence that these deposits were ex­ ploited in prehistory. Nor is there a clear correlation between the Sn isotopic composition of Bukulja ores and local artefacts, particularly when the 0.2‰ cor­ rection associated with smelt­related enrichment of heavy isotopes in metal products is taken into account, as recommended by Berger et al. (2019), Powell et al. (2019) and Mason et al. (2020) (Fig. 3).14 Thus, it ap­ pears that Bukulja either was not mined in the Late Bronze Age, or was a minor contributor to the tin eco­ nomy of Serbia relative to Mt Cer and the Erzgebirge at that time. The Sn istotopic composition of the multi­crystal composite sample of cassiterite from Motajica is far more enriched in light isotopes of Sn than that of Late Bronze Age metal artefacts from the region (Fig. 3). The full range of Sn isotopic composition of cassiter­ ite from Motajica cannot be determined from a single analysis. However, the placer sample was composed of at least a hundred individual sand­ to silt­sized detrital 13 14 Mason et al. 2020 Mason et al., 2020. СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) mineral grains. Thus, the composition of this sample represents a multi­crystal average and, as such, is like­ ly to approach the central tendency of the deposit’s isotopic composition. To support this assertion, the mean and standard deviation of the set of nine placer samples from the Erzgebirge reported by Haustein et al. (2010) is 0.56±0.12‰, as compared to 0.59±0.38‰ for the total of 43 samples. Accordingly, it is reasona­ ble to conclude that the Motajica mineralisation is highly enriched in light isotopes of Sn. Given that none of the 336 artefacts analysed by Mason et al. (2020) yielded δ124Sn values < ­0.04 (Fig. 3), Late Bronze Age artefacts with light isotope enrichment appear to be absent from the region. Therefore, if the low­grade placers from Motajica were exploited at all at that time, they did not contribute to the regional tin economy. Based on the results of on­site sluice­ and pan­ based sampling, of the six potential Balkan tin ore sources identified by Durman, only Cer, Bukulja and Motajica appear to have sufficient cassiterite concen­ tration to have allowed for the potential exploitation of placer gravels. Furthermore, based on the isotopic composition of Sn from these sources compared to Late Bronze Age artefacts, Cer is the only site that ap­ pears to have likely contributed substantially to the regional tin economy at that time Starinar is an Open Access Journal. All articles can be downloaded free of charge and used in accordance with the licence Creative Commons – Attribution­NonCommercial­NoDerivs 3.0 Serbia (https://creativecommons.org/licenses/by­nc­nd/3.0/rs/). Часопис Старинар је доступан у режиму отвореног приступа. Чланци објављени у часопису могу се бесплатно преузети са сајта часописа и користити у складу са лиценцом Creative Commons – Ауторство-Некомерцијално-Без прерада 3.0 Србија (https://creativecommons.org/licenses/by­nc­nd/3.0/rs/). 91 СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) BIBLIOGRAPHY: Antonović, Vukadinović 2012 – D. Antonović, M. Vukadi­ nović, Eneolithic mine Prljuša-Mali Šturac: archaeological and geophysical investigations. Старинар LXII, 2012, 95–106. Antonović 2018 – D. Antonović, Eneolitski rudnici bakra na Balkanu, in: Povratak u prošlost. Bakreno doba u sjevernoj Hrvatskoj, (ed.), J. Balen, I. Miloglav, D. Rajković, Za­ greb 2018, 187–209. Balliana et al. 2013 – E. Balliana, M. Aramendía, M. Resa­ no, C. Barbante, F. Vanhaecke, Copper and tin isotopic analysis of ancient bronzes for archaeological investiga­ tion: development and validation of a suitable analytical methodology. Analytical and Bioanalytical Chemistry 405, 2013, 2973–2986. Begemann et al. 1995 – F. Begemann, E. Pernicka, S. Schmitt­Strecker, Searching for the Ore Sources of Eneo­ lithic and EBA Copper Artefacts from Serbia. Ancient Mining and Metallurgy in Southeast Europe, B. Jovanović (ed.), (Papers from the International Symposium, May, 20–25 1990, Belgrade­Bor 1995, 143–149. Berger et al. 2019 – D. Berger, J. Soles, A. Giumlia­Mair, G. Brügmann, E. Galili,N. Lockhoff, E. Pernicka, Isotope systematics and chemical composition of tin ingots from Mochlos (Crete) and other Late Bronze Age sites in the eastern Mediterranean Sea: An ultimate key to tin prove­ nance?. PloS one, 14(6), 2019, e0218326. Boev et al. 2002 – B. Boev, S. Lepitkova, G. Petrov, Grani­ toid formations in the Republic of Macedonia, Geologica Carpathica 53, Proceedings of XVII Congress of CarpathianBalkan Geological Association, September 1–4 2002, Bra­ tislava 2002. Богосављевић-Петровић 2005 – В. Богосављевић-Пе­ тровић, Праисторијски рудници на централном Балка­ ну. Зборник народног музеја XVIII–1, 2005, 80–113. (V. Bogosavljević-Petrović, Praistorijski rudnici na centralnom Balkanu. Zbornik Narodnog muzeja XVIII–1, 2005, 80–113). Черних, Радунчева 1972 – E. Черних, А, Радунчева, Старите медни рудници около гр. Стара Загора. Aрхеология XIV(1), 1972, 61–66. (E. Černych, A. Raduncheva, Starite medni rudnici okolo gr. Stara Zagora. Arheologiya XIV(1), 1972, 61–66). Černych 1978 – E. N. Černych, Aibunar – a Balkan copper mine of the fourth millennium BC (Investigations of the years 1971, 1972 and 1974). Proceedings of the Prehistoric Society 44, 1978, 203–217. 92 Черньх 1978 – E. Н. Черньх, Горное дело и металлургия в древнейшей Болгарии. София 1978. (E. N. Černych, Gornoe delo i metalurgia v drevnitei Bolgarii. Sofia 1978). Cruse et al. 2017 – S. Cruse, W. Powell, A. Huska, H.A. Bankoff, V. Filipović, Geochemical Prospecting at Mt. Pro­ sara and Motajica, Northern Bosnia: Examining Potential Sites of Bronze Age Placer Tin Ore Mining. Geological Society of America Abstracts with Programs, 2017, 49(6). Davies 1937 – O. Davies, Prehistoric copper­mine at Jarmo­ vac near Priboj na Limu. Glasnik Zemaljskog muzeja u BiH XLIX(1), 1937, 1–3. Durman 1997 – A. Durman, Tin in South­eastern Europe? Opuscula Archaeologica 21, 1997, 7–14. Haustein et al. 2010 – M. Haustein, C. Gillis, E. Pernicka, Tin isotopy: a new method for solving old questions. Archaeometry 52, 816–832. Huska et al. 2014 – A, Huska, W. Powell, S. Mitrović, A.H. Bankoff, A. Bulatović, V. Filipović, R. Boger, Placer Tin Ores from Mt. Cer, West Serbia, and their Potential Ex­ ploitation during the Bronze Age. Geoarchaeology 29, 2014, 477–493. Jovanović 1971 – B. Jovanović, Metalurgija eneolitskog perioda Jugoslavije, Beograd 1971. Јовановић 1974 – Б. Јовановић, Технологија рударства у раном енеолиту Централног Балкана. Старинар XXIII, 1974, 1–13. (B. Jovanović, Tehnologija rudarstva u ranom eneolitu Centralnog Balkana. Starinar XXIII, 1974, 1–13). Jovanović, Ottaway 1976 – B. Jovanović, B. S. Ottaway, Copper mining and metallurgy in the Vinča group. Antiquity L, 1976, 104–113. Jovanović 1985 – B. Jovanović, Rudna Glava: najstarije rudarstvo bakra na Centralnom Balkanu, Bor–Beograd 1985. Јовановић 1988 – Б. Јовановић, Прљуша – Мали Шту­ рац: Праисторијски рудник бакра и горског кристала на Руднику. Зборник радова Народног музеја Чачак XVIII, 1988, 5–12. (B. Jovanović, Prljuša-Mali Šturac: Praistorijski rudnik bakra i gorskog kristala na Rudniku. Zbornik radova Narodnog muzeja Čačak XVIII, 1988, 5–12). Mason et al. 2016 – A. Mason, W. Powell, A. H. Bankoff, R. Mathur, A. Bulatović, V. Filipović, J. Ruiz, Tin isotope characterization of bronze artifacts of the Central Balkans. Journal of Archaeological Science 69, 2016, 110–117. Mason et al. 2020 – A. Mason, W. Powell, A. H. Bankoff, R. Mathur, M. Price, A. Bulatović, V. Filipović, Provenance of tin in the Late Bronze Age Balkans based on probabilistic and spatial analysis of Sn isotopes. Journal of Archaeological Science 122, Article 105181. СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Mathur et al. 2017 – R. Mathur, W. Powell, A. Mason, L. Godfrey, J. Yao, M. Baker, Preparation and measurement of cassiterite for Sn isotopic analysis. Geostandards and Geoanalytical Research 41, 2017, 701–708. Mihajlović 1978 – K. Mihajlović, Aluviajalno ležište kasiterita. Cigankulja. Radovi sa IX Kongres Geologa Jugoslavije, Sarajevo 1978, 620–624. Pernicka et al. 1993 – E. Pernicka, F. Begemann, S. Schmitt­ Stecker, G. A. Wagner, Eneolithic and Bronze Age copper artefacts from the Balkans and their relation to Serbian cop­ per ores. Prehistorische Zeitschrift 68(1), 1993, 1–57. Powell et al. 2019 – W. Powell, R. Mathur, J. John, M. Price, H.A Bankoff, M. Tisucká, L. Godfrey, Unearthing Europe’s Bronze Age Mining Heritage with Tin Isotopes: A Case Study from Central Europe. European Geologist, 48, 58–62. 93 Simić 1951 – V. Simić, Istoriski razvoj našeg rudarstva, Beograd 1951. Simić 1969 – V. Simić, Istorijski osvrt na rudarstvo bakar­ nog rudišta u Boru i okolini. Zbornik radova Rudarsko-metalurškog fakulteta i instituta u Boru VIII, 1969. Tomić 1991 – R. Tomić, Godišnji izveštaj o izvršenim tehnološkim ispitivanjima kalaja i retkih metala u aluvijalnim ležistima Lešnice i Cernice u području Cera u 1990. godini, Beograd 1991. Tućan 1957 – F. Tućan, Specijalna minerologija, Zagreb 1957. Ustaszewski et al.2010 – K. Ustaszewski, A, Kounov, S. Schmid, U. Schaltegger, E. Krenn, W. Frank, F. Fügen­ schuh, Evolution of the Adria­Europe plate boundary in the northern Dinarides: From continent­continent collision to back­arc, Tectonics 29, 2010, TC6017. СТАРИНАР LXX/2020 Wayne POWELL, Ognjen MLADENOVIĆ, Steffanie CRUSE, H. Arthur BANKOFF, Ryan MATHUR Revisiting “Tin in South-eastern Europe?” (85–94) Резиме: ВЕЈН ПАУЕЛ, Одељење за геонауке и природну средину, Бруклин колеџ ОГЊЕН МЛАДЕНОВИЋ, Археолошки институт, Београд СТЕФАНИ КРУЗ, Институт за рударство Колорадо Х. АРТУР БАНКОФ, Одељење за антропологију и археологију, Бруклин колеџ РАЈАН МАТУР, Одељење за геологију, Јуниата колеџ ЈОШ ЈЕДНОМ О „КАЛАЈУ У ЈУГОИСТОЧНОЈ ЕВРОПИ?” Кључне речи. – сводови, цеви за сводове, керамичке цеви, tubi fittili, Timacum Minus, технике грађења, римска архитектура, касноантичка архитектура, северијански период, југоисточна Европа Обновом археолошких ископавања античког кастела Тима­ кум Минус 2019. године створиле су се нове могућности за тумачења његових грађевина које су истраживане пре више деценија. Међу остацима грађевина око античког кастела Timacum Minus-a посебну пажњу привлачи делимично ис­ тражен „објекат са хипокаустом”, нарочито у погледу ње­ гових конструктивних карактеристика. Поред иначе честих античких конструкција хипокауста и зидног грејања, међу остацима ове грађевине уочена је и посебна врста грађе­ винских елемената – керамичке цеви за сводове. Велика количина откривених цеви указала је на то да је ова грађе­ вина заиста имала сводове израђене од њих. Иако је појава цеви за сводове приликом истраживања античких локалитета на тлу југоисточне Европе регистро­ вана, она није довољно документована, као што ни сама функција цеви често није препозната. Један од разлога за то јесте недовољна упућеност истраживача у специфичне карактеристике цеви за сводове и њихову функцију, услед чега се оне мешају са водоводним цевима, тубулусима или калемовима везаним за зидно грејање – будући да сваки од тих елемената припада керамичким производима који су намењени грађевинарству. У раду су разматране карактеристике цеви за сводове на Тимакум Минусу, као и контекст у коме су пронађене унутар „објекта са хипокаустом”. На основу налаза печата кохорте Аурелије II Дараданорум одређено је да „објекат са хипокаустом” и конструкција сводова од керамичких цеви потичу из III века – у коме је и иначе појава тих сводо­ ва широм Римског царства била честа. Приликом систематизациjе врста керамичких цеви на Тимакум Минусу посебно је издвојена она које је било нај­ више у „објекту са хипокаустом”. У склопу ње је препознат и сасвим специфичан централни елемент који је омогућавао да се два низа цеви на истом правцу, али из супротних сме­ рова, међусобно споје. Тај елемент је дефинисао облик сво­ 94 да којим су биле покривене просторије чију је реконструк­ цију основе било могуће извршити. Архитектонске анализе „објекта са хипокаустом”, као и карактеристике уочене на самим цевима указале су на то да су просторије биле покривене полуобличастим сводом, изграђеним од лучних вертикалних низова цеви које су у темену биле ,,закључане” централним елементом. Реконструкција изгледа цеви и начина њиховог ређања уклапа се у хронологију извођења објекта и свода током III века. Даљим статичким анализама дошло се до још неколико сазнања. Показало се да је преко свода морао бити нанесен одређен слој малтерне масе да би дебљина свода досегла oптималну вредност у опсегу 20–30 cm. На основу пропорција објекта које су одређене у његовој основи испитана је висина објек­ та, где је група случајева такође дефинисана пропорцио­ нално. Према нашим анализама, зидови просторија „објек­ та са хипокаустом” у којима су цеви регистроване могли су досезати висину до 3,08 m, док је висина просторија у теме­ ну свода могла бити 6,16 m. Овим истраживањем покушали смо да укажемо на ве­ лики значај појединачних архитектонско-грађевинских еле­ мената, а међу њима и керамичких цеви за сводове, којима се често не придаје довољна пажња. Налази керамичких цеви за сводове у Тимакум Минусу, уз извршене архитек­ тонске анализе, употпуњују слику откривеног „објекта са хипокаустом” из више аспеката. Посебно је значајно дефи­ нисање његове висине, које је веома тешко за античке грађе­ вине профане архитектуре на нашем тлу будући да су нај­ чешће сачуване у приземној или темељној зони. Значај налаза керамичких цеви за сводове у Тимакум Минусу ве­ лики је стога што је он омогућио како конкретно дефини­ сање контекста њиховог налаза тако и реконструкцију облика одређених делова грађевине помоћу тог елемента, што до сада није истраживано приликом анализа античке архитектуре на тлу југоисточне Европе. СТАРИНАР LXX/2020