NOTES SCHOENINGER, M., M. DENIRO& H. TALIBER. 1983. Stable nitrogen isotope ratios of bone collagen reflect marine and terrestrial components of prehistoric human diet, Science 220:1381-1383. R.J. In press. New AMS dates from the Lambourn SCHULTING, long barow and the earliest Neolithic in southern Britain: repacking the Neolithic package?, Oxfordlournal of Archaeology. H. & M. SCHOENINGER. 1991. Stable isotope analySCHWARCZ, ses in human nutritional ecology, Yearbook of Physical Anthropology 34: 283-321. 897 H. 1981. I7C evidence for dietary habits of prehistoric TAUBER, man in Denmark, Nature 292: 332-3. WHITTLE, A.W.R. 1996. Europe in the Neolithic: The creation of New Worlds. Cambridge: Cambridge University Press. E. 1989. Dating the introduction of food producWILLIAMS, tion into Britain and Ireland, Antiquity 63: 510-21. 1986. Foragers and farmZVELEBIL, M. & P.A. ROWLEY-CONWY. ers in Atlantic Europe, in M. Zvelebil (ed.),Hunters in transition: Mesolithic societies of temperate Eurasia and their transition to farming: 67-94. Cambridge: Cambridge University Press. The oldest metallurgy in western Europe ARTURO RUIZ-TABOADA & IGNACIOMONTERO-RUIZ * Recent excavations at the Neolithic site of Cerro Virtud (Almeria, southeast Spain) have produced new information about the development of metallurgy that m a y change ongoing research not only in the Iberian Peninsula but also in the rest of western Europe. The discovery of metallurgy in this region in the first half of the 5th millennium BC poses serious challenges to the interpretation of h o w this industry developed and spread, given that the nearest European region with similar evidence is the Balkans. This study presents the archaeological context of the discovery and the various analytical techniques (XRF, SEM, l4C)that have been applied to it. Key-words: Neolithic, metallurgy, Spain, Cerro Virtud, I4C dating, XRF, SEM Introduction Rescue excavations in 1994 at Cerro Virtud (Almeria, Spain), a site dated to the first half of the 5th millennium BC, have provided the oldest evidence for metallurgy in western Europe. This date is more than a millennium older than had previously been attested. The sole evidence consists of a sherd of an ordinary openmouthed ceramic vessel that was used to smelt copper. The sherd has a layer of slag on its inner surface. This discovery is not only important because of its date but also because of its cultural context. The study of the site and the surrounding region shows that the beginning of metallurgy is associated with the consolidation of mixed farming. Thus, metallurgical production begins during the process of Neolithicization in Iberia in an egalitarian social context showing no traces of hierarchical differences. Cerro Virtud revitalizes the traditional debate about metallurgy as a factor in social change and strengthens the hypothesis that it developed independently and autonomously in the Iberian Peninsula. Diffusionism vs autonomous development of metallurgy For years scholars believed that the most significant technological developments (such as metallurgy) had spread into Europe from the Near East (Wertime 1973; Savory 1968). Today, scholars no longer think in these terms; they generally consider that the development of these new technologies could have occurred in several parts of the world more or less simultaneously. Colin Renfrew was the first to propose the autonomous development of metallurgy in the Balkans (Renfrew 1969) and the Iberian Peninsula (Renfrew 1967; 1970). For the lat- * Ruiz-Taboada, Facultad de Humanidades, Universidad Castilla-La Mancha, Pza de Padilla 4, 45071 Toledo, Spain. taboada@arrakis.es Montero-Ruiz, Dpt de Prehistoria (CEH/CSIC), C/ Serrano 13, 28001 Madrid, Spain. imontero@ceh.csic.es Received 3 June 1997, revised 3 March 1999, accepted 15 May 1999. ANTIQLJITY 7 3 (1999): 897-903 NOTES 898 4125,6 i- + 607 0 FIGURE 1 . General location of Cerro Virtud and excavated areas (drawing by Julia Srinchez). ter, Renfrew based his hypothesis on the relative antiquity of metallurgy in Iberia and the absence of findings between this region and the Balkans. This absence has progressively been filled by a variety of metallurgical discoveries in the intermediate regions (Camps 1991; Guilaine 1991; Strahm 1994). Systematic fieldwork to investigate the role of metallurgy in the developineni of prehistoric societies in southeast Spain began in the late 1970s, but it was not until the late 1980s that such research made a significant interpretative impact (Chapman 1990). These studies did not find any evidence of metallurgy older than the mid 4th millennium BC date proposed by Renfrew. In support of an autonomous development, Montero (1993; 1994) pointed out that prehistoric metallurgy was much more archaic in Spain than in the rest of Europe. This archaism is seen as the result partly of a distinct metallurgical tradition and partly of the slight importance of metal in the development of social complexity in the region (Gilman 1996). This tradition has three main features that should be emphasized: 1 The use of vase-ovens, a special kind of smelting crucible, for reducing ores, a system not as yet documented outside the Iberian Peninsula. The discovery of several fragments kom the Bauma del Serrat del Pont (Girona) (Alcalde et al. 1998), close to the French boundary, also suggests its possible knowledge in the south of France. 2 The absence in the Iberian Peninsula of the characteristic crucibles with handles, found from the mid 4th millennium cal BC in almost all western Europe (France, Switzerland, Italy, Corsica, etc.) (Fasnacht 1991; Camps 1991). 3 The absence in southern Iberia of the clay tuyeres frequent in metallurgical contexts in the rest of Europe (Gatiglia & Rossi 1995). These only appear in the late Chalcolithic in northern Iberian regions linked to the Atlantic and central European metallurgical spheres (Alcalde et al. 1998). The discovery of metallurgical activity at Cerro Virtud changes the chronological aspect of this panorama. The Neolithic at Cerro Virtud and the origins of metallurgy The site is located (FIGURE1) in the natural depressions found near the top of a hill that stands 35-40 m above the cultivated alluvium of the river Almanzora (Almeria), 3 km from its mouth. It has a long sequence with various phases of Neolithic occupation, including the presence of a contemporary collective burial. As a result of the site’s topography, there is no stratigraphic connection between some of the excavated areas. Other details about the site may be found in previous publications (Montero & Ruiz-Taboada 1996; Montero et al. 1999). The material recovered has the typical characteristics of other sites of the so-called Impressed Ware Complex (Bernabeu 1989). The predominant decorative motifs of the pottery are parallel-line incisions, punctations, impressions, digitations and applique cordons. There NOTES decorative groups incision impression cordons comb-impressed red-slipped painted cardial Cerro Virtud 1 30.5 36.1 19.4 8.3 5.5 - 899 Cerro Virtud B2 Cerro Virtud Carigiiela 2 C 33.3 14.3 38.1 14.3 19-2 19.2 30.7 23.1 3.8 3.8 - - 31.25 14.06 35.94 10.94 - - 7.81 Nerja I1 27.57 20.76 38.98 16.01 - 6.02 Cendres Nacimiento VIII 4845 24.07 7.41 18.52 1.85 - 22 26 22 12 - 18 Cendres VIIIb 34.04 22.34 14.89 21.28 7.45 TABLE1 . Comparative distribution (%) of decorative motifs in the pottery from contemporaneous Neolithic layers in south and east of Spain. are also comb-decorations, characteristic of the Spanish Levant, and red-burnished pottery, characteristic of Andalusiaii cave sites. The presence of decorative elements typical of these Neolithic traditions is hardly surprising since the site is located between the two regions, although the absence of cardial impressions should be noted. The proportions of decorative motifs of the pottery from these levels places the occupations in the first half of the 5th millennium cal BC [phase IB2 of the Neolithic sequence defined by Bernabeu (1989)).The radiocarbon evidence from Cerro Virtud confirms this chronology, the dates falling between 6160t180 BP (5440-4690 cal BC at 95% probability, Beta-101424) and 5660k80 BP (4700-4350 cal BC at 95% probability, Beta-90884). TABLE1 summarizes the proportion of decorative motifs from phase 1, square B2 [where the vase-oven was found), and phase 2 of Cerro Virtud and from other published sites in Andalusia and the Spanish Levant. The metallurgical remains were recovered from square B2. (Three other fragments of different vase-ovens were also recovered from the trench (TR). This modern trench was filled up with prehistoric layers removed from closer FIGURE 2. BZ pit stratigraphy A modern layers; B Chalcolithic layer; C Neolithic layer; D clayey marl. FIGURE3. Fragment of the vase-oven (BZ/lO) from the Neolithic layer dated to the first half of the 5th millennium cal BC. NOTES 900 0 s tb areas.) The stratigraphy of this sector consists of various modern mixed layers, a thin Chalcolithic layer, and below this a Neolithic deposit about 60 cm thick, resting on the clayey marl of the base (FIGURE 2). The fragment appeared in the lower part of the Neolithic level about 9 to 15 cm above the marl. This undisturbed layer corresponds to phase 2 of the site’s occupation. The chronology obtained on the basis of ceramic comparisons is confirmed by a radiocarbon date obtained from organic sediments in the layer, 5830k90 BP (4915-4475 cal BC at 95% probability, Beta-118936). The vase-oven fragment (FIGURE 3) is 9 mm thick and barely 3 cm long. Remains of slag 15 2 EWW FIGURE 4. Slag from the B2llO fragment of vase-oven. Microphotograph y and spectrum from area no. 2 [SEM analysisj. adhere to its inner face. Metallography permits the identification of small metallic globules within the siliceous matrix. These indicate its function as a container for the reduction of ore. The outer face reveals no trace of having been submitted to thermic action. The slaggy inner surface is the result of copper-ore reduction with charcoal. These features are commonplace in these types of objects and have been described elsewhere (Montero 1994: 227-30; Craddock 1995: 133-4). XRF-EDS analyses were carried out in the Instituto de Patrimonio Hist6rico Espafiol. Because the metallic globules formed a very small proportion of the matrix only a qualitative analy- NOTES 901 FIG~JRE 5. Copper mapping f r o m a slaggy area of B2/10 fragment. sis was possible. A minimal amount of copper was identified, as well as lead, antimony and barium in somewhat larger proportions. A scanning electron microscope (SEM) was used for a more detailed study of the distribution of these elements. The slag texture is not very homogenous, but its basic composition was similar in the two analyses carried out with broad windows: Si, Ca, Al, K, Mg, Fe, N, S, and C1 were detected (FIGURE 4). Zone 2 was richer than zone 1 in Ca and A1 and had also Ti. Given the difficulty of aiming the microprobe at a point with copper, plus the low sensitivity of the method when it is not applied to a point, we opted for graphic mapping of the elements. 5), lead As a result we detected copper (FIGURE and antimony broadly dispersed in various microscopic fields. Although the slag is very leached, we deduce from these analyses that the vase-oven fragment was used to smelt copper ores that contained lead and antimony, among other impurities. Other very light and porous fragments of slag were found beside the vase-oven fragment, and the same kind of slag was also found in the bottom levels of square B3 (phases 1 and 2). Several samples, all of them complex silicates, were studied. Various semiquantitative analyses were conducted 011 sample B3/30-5 from phase 2 using the SEM of the Universidad Aut6noma of Madrid (TABLE2). Some analyses detected the presence of barite (barium sulphate) and celestite (strontium sulphate) (TABLE 2: no. 3), types of minerals present paragenetically in the geological formations of the Vera Basin (in particular Herrerias and Sierra 1 4 3 2 5 6 1.07 0.58 3.14 3.74 4.57 2.88 54.72 31.08 34.92 1.32 6.12 2.61 1.15 1.93 14.38 24.82 49.91 1 4 9 0.43 3.87 1.24 45.15 92.77 7.43 1.53 30.74 1.07 5.52 1.29 6.10 1.66 8.64 8.89 1.93 1.31 8.67 26.38 0.21 4-04 7.60 13.72 30.39 52.64 TABLE2. Analysis by SEM of B3/30(5) slag sample (% weight]. sample PA6354A PA6354B PA6354C Ag Sb Pb 1.63 - - 0.07 0.24 0.93 2.02 8.29 4.67 1.15 Fe Cu As 26.85 28.64 22.80 66.1 68.1 68.6 - - TABLE3. Analysis by XRF of mineral samples from Mina Guadalupe, near Cerro Virtud. Only metallic fraction (% weight) considered. Almagrera) and transferred into sedimentary deposits as a result of erosion. None of the measurements detected copper, which probably means the sample was not a metallurgical slag. It seems to be a foamy glass produced by an intense fire, the sort of ‘fuel ash slag’ (Biek & Bayley 1979) that can be formed as a result of accidental fires in clayey settings or even in wooden structures.They appear frequently in Neolithic sites in southeast Europe as the result of long-used hearths or the destruction by fire of houses built of wood and clay (Glumac & 902 NOTES Todd 1991). Their external appearance is similar to copper slags, but the latter can be distinguished by their greater specific gravity. The slags formed under these conditions share a low percentage of iron and calcium, according to Bieks (1977) analysis of specimens from a fire in a Roman granary. Their composition, however, largely depends on the characteristics of the medium that produces them. At Cerro Virtud, clayey marl composition is not like that described so far in the scientific literature. Marl is sedimentary and contains 4060% calcium carbonate, the rest being clay. It might respond to intense heat or produce reactions in a manner similar to clay, since at Cerro Virtud the marl’s texture is quite claylike, and this would explain the high proportion of calcium and barium in the slags analysed from our site, indicating that a material with these characteristics was part of their formation process. The slags only appear in the lower levels of squares B2 and B3 of the site, where they were in contact with the clayey marl. The metallic elements detected in the vaseoven correspond to the characteristics of the copper ore at Herrerias, located at the foot of the hill. Louis Siret’s early 1900s analyses of samples from that locality showed a high proportion of Pb (Montero 1994: 1 0 1). New samples obtained in 1994 near the Guadalupe mine confirm the presence of lead. Most of the metallic materials recovered from Cerro Virtud and from Bronze Age sites in the immediate vicinity of Herrerias contain lead, a circumstance that is rare at Almizaraque and other prehistoric sites in the region (Montero 1993; 1994). Thus, from the Neolithic onwards there seems to exist a direct relationship between the type of impurities found in copper ores and the metal found at particular sites. This suggests the inhabitants of sites exploited ores in their immediate vicinity. Conclusion The evidence indicates a very early start to metallurgy, with a date in the first half of the 5th millennium cal BC. This age is only exceeded in the Balkans and is similar to data drawn from northern Italy, a region closely linked geographiReferences ALCALDE G , M MOLIST, I MONTERO, LL PLANALUMA, M SANX8~ A TOLEDO1998 Producciones metaldrgicas en el nordeste dr la Peninsula Iberica durante r1 I11 cally and culturally with the Balkans (Barfield 1996).There is no chronological and geographical continuity in the first appearance of metallurgy between the Orient and the western Mediterranean, suggesting an autonomous and independent development in Iberia. The early evidence shows a full development of metallurgy involving the transformation and reduction of copper ores. The vase-oven method of ore reduction is present from the very beginning and continued in use throughout later Iberian prehistory. There are many examples, some of which even involved the use of decorated pottery (Alcalde et al. 1998). The probable use of local resources, at short distances from settlements, undoubtedly encouraged metallurgical experimentation. Knowledge of metallurgy at Neolithic Cerro Virtud cannot have been an isolated phenomenon. This discovery warrants a contextual review of material recovered from earlier excavations. Many metal objects from collective megalithic burials associated with both Neolithic and Chalcolithic materials have been assigned to the latter period because of the assumption that metallurgy was absent in the Neolithic. Clear proof that the metal belongs to the older period will be difficult. We have a clear case, however, in the copper awl from the cave of La Cocina (Valencia):it was recovered from a secure stratigraphic context, a level that because of its lithic industry cannot, according to Bernabeu (1989: 136), be younger than phase IB of the Neolithic, an age comparable to that of Cerro Virtud. This find has always been interpreted as a proof, not of an early start of metallurgy, but as persistence of geometric microliths into the Chalcolithic. In the future we no doubt will obtain other finds in reliable archaeological contexts that will confirm the evidence from Cerro Virtud. Acknowledgements. We wish to express our thanks to Antonio Gilman for his help and comments on this paper. Part of this research was undertaken with funds from the Junta d e Andalucia (Spain). Some of the ideas formulated in the article have been developed through useful discussion with, among others, Salvador Rovira and participants i n the fall seminar at the Department of Archaeology, University of Berkeley i n 1997. We are also grateful to the Department of Scientific Research (British Museum) for helping in our research about the slags. milenia cal. AC: el taller de la Bauma del Serrat del Pont (Tortelk, Girona), Trabojos de Prehistoria 55[1): 81-100. NOTES SFORZA.1984. The Neolithic AMMERMAN, A.J. & L.L. CAVALLI transition and the genetics of populations in Europe. Princeton (NJ): Princeton University Press. BARFIELD,L.H. 1996. The Chalcolithic in Italy: considerations of metal typology and cultural interaction, in B. Bagalini & F. Lo Schiavo (ed.), The copper age i n the Near East and Europe: 65-74. Forli: Edizioni Forli. UISPP 1 0 Colloquium, ABACO. BERNABEU,J. 1989. La tradicidn cultural d e las ceramicas impresas en la zona oriental de la Peninsula Iberica. Valencia: Diputacidn Provincial de Valencia. Servicio de Investigacion Prehistbrica. Serie de Trabajos Varios 86. BIEK, L. 1977. Slags and allied material, in P.A. Rahtz & E. Greenfield, Excavations at Chew Valley Lake: 357-9. London: Department of the Environment. Archaeological Report 8. BIEK, L. & J. BAYLEY. 1979. Glass and other vitreous materials, World Archaeology 11(1):3-25 CAMPS,G. 1991. Les creusets de Terrina (Aleria, Haute-Corse), i n Le Chalcolithique en Languedoc. Ses relations extraregionales: 41-9. Lattes: Archeologie en Languedoc. Colloque International Honimage au Dr Jean Arnal. CHAPMAN, R.W. 1990. Emerging complexity The later prehistory of southeast Spain, Iberia and the west Mediterranean. Cambridge: Cambridge University Press. CRADDOCK, P.T. 1995. Early metal mining and production. Edinburgh: Edinburgh University Press. ELUERE,C. & J. P. MOHEN(ed.). 1991. Dicouverte du mital. Paris: Picard. W. 1991. Analyses d'objets en cuivre du NBolithique FASNACHT, recent du bassin zurichois, in Eluere & Mohen (ed.]: 15766. GATIGLIA,A.& M. ROSSI. 1995. Les ceramiques de la mine prehistorique de Saint-VerBn (Hautes-Alpes), Bulletin de la Societ.4 Prehistorique FranFaise 92(4): 509-18. 903 GUILAINE, J. 1991. Roquemergarde et les debuts de la metallurgie en France m6diterraneenne, in Eluere & Mohen (ed.): 279-94. GILMAN, A. 1996. Craft specialization in late prehistoric Mediterranean Europe, in B. Wailes (ed.):Craft specialization and social evolution: in memory of I( Gordon Childe: 67-71. Philadelphia (PA): University of Pennsylvania. University Symposium series 6. GLUMAC, P. & J.A. TODD. 1991. Early metallurgy in Southeast Europe: the evidence for production, MASCA Research Papers in Science and Archaeology 8(1): 9-19. MONTERO, I. 1993. Bronze Age metallurgy in Southeast Spain, Antiquity 67: 46-57. 1994. El Origen de la Metalurgia en el Sudeste de la Peninsula Iberica. Almeria: Instituto de Estudios Almerienses. Colecci6n de Investigacion 19. MONTERO, I. &A. RUIZ-TABOADA. 1996. Enterramiento colectivo y metalurgia en el yacimiento Neolitico de Cerro Virtud Cuevas de Almanzora, Almeria, Trabajos de Prehistoria 53(2): 55-75. MONTERO, I., C. RHUETE & A . RUIZ-TABOADA. 1999. Precisiones sobre el enterramiento colectivo neolitico de Cerro Virtud (Cuevas de Almanzora, Almeria), Trabajos de Prehistoria 56(1]: 119-30. RENFREW, C. 1967. Colonialism and megalithismus, Antiquity 41: 276-88. 1969. The autonomy of the south-east European Copper Age, Proceedings of the Prehistoric Society 35: 12-47. 1970. The tree-ring calibration of radiocarbon: an archaeological evaluation, Proceedings of the Prehistoric Society 36: 280-311. SAVORY, H. 1968. Spain and Portugal. The prehistory of the Iberian Peninsula. London: Thames & Hudson. STRAHM, C. 1994. Die Anfange der metallurgie in Mitteleuropa, Helvetia Arqueologica 97: 2-39. WERTIME, T.A. 1973. The beginnings of metallurgy: a new look, Science 182: 875-87. Re-assessing the logboat fkom Lurgan Townland, Co. Galway, Ireland M.E. ROBINSON,D.W. SHIMWELL & G. CRIBBIN" Recent study of the prehistoric Lurgan logboat reveals m a n y details of its construction and date. Speculation on h o w the boat was used and why it was incomplete offer an insight into Irish prehistory. Key-words: logboat, Ireland, prehistory, peat bog, sea craft Logboats have been found in northwestern Europe in large numbers and it is clear that they were in use from the Neolithic to late medieval times (Lanting & Brindley 1996).Because they are relatively robust, their survival rate compares favourably with other primitive craft such as log- and bundle-rafts and hide-covered vessels of the kind depicted in the Broighter gold model (McGrail 19961. The role played by logboats in the social economy of prehistoric Europe, however, has probably been underestimated, particularly in lowland areas * Robinson & Shimwell, Palaeoecological Research Unit, Department of Geography, University of Manchester, Oxford Road, Manchester M13 9% England. m.robinson@manchester,ac.ukd.shimwell@manchester.ac.uk Cribbin, Tooraree, Ballyhaunis, Co. Mayo, Ireland. Received 3 February 1999 accepted 23 April 1999, revised 11 May 1999. ANTIQUITY73 (1999): 903-8