The International Symposium „The Beginnings of Metallurgy" was supported by the Volkswagen-Stiftung, Hannover, by Kontron Phystech GmbH, Eching/Munich, and by Norddeutsche Affinerie, Hamburg.

The publication of the proceedings was supported by Deutsches Bergbau-Museum Bochum and Vereinigung der Freunde für Kunst und Kultur im Bergbau e.V., Bochum.

Montanhistorische Zeitschrift

= Veröffentlichungen aus dem Deutschen

Bergbau-Museum, Nr. 84

Deutsches Bergbau-Museum

Institut für Archäometallurgie

Herner Straße 45

D-44787 Bochum

Dipl.-Ing. Angelika Friedrich

DZS GmbH Essen

Die Deutsche Bibliothek - CIP-Einheitsaufnahme

The beginnings of metallurgy: proceedings of the International Conference The Beginnings of Metallurgy, Bochum 1995 / Andreas Hauptmann ... - Bochum: Dt. Bergbau-Museum, 1999

(Veröffentlichungen aus dem Deutschen Bergbau-Museum

Bochum; Nr. 84) (Der Anschnitt: Beiheft: 9)

The Illustration points to one of the basic questions of the incipient stages of metallurgy, exemplified at copper:

Was the earliest metal used by man during the Pre-Pottery Neolithic native metal as it occurs in nature or was it produced by smelting ores? To distinguish native and smelted copper, both metallographic studies and trace element analyses were successfully applied, and numerous experiments were carried out in laboratories to solve this problem.

Extractive metallurgy started in Anatolia at the end of the sixth millennium B.C., and it was commonly practiced in the Near East during the fourth millennium B.C. first by smelting ores in crucibles using blowpipes, later in furnaces. Simulation of smelting processes in crucibles are described in this this volume by Fasnacht (page 291) from where this photograph is taken.

Herausgeber:

Vereinigung der Freunde von Kunst und Kultur im Bergbau

e.V. Vorsitzender des Vorstandes:

Bergassessor Dipl.-Kfm. Achim Middelschulte

Vorsitzender des Beirats:

Assessor des Bergfachs Friedrich H. Esser, M. Sc.

Geschäftsführer:

Museumsdirektor Prof. Dr. phil. Rainer Slotta

Redaktionsleitung (verantwortlich):

Dr. phil. Werner Kroker

Wissenschaftlicher Beirat:

Dr. phil. Evelyn Kroker, M. A., Dr.-Ing. Siegfried Müller,

Dr.-Ing. habil. Dr. rer. nat. Thilo Rehren,

Prof. Dr. phil. Gerd Weisgerber

Anschrift der Geschäftsführung und der Redaktionsleitung:

Deutsches Bergbau-Museum

Am Bergbaumuseum 28 • D-44791 Bochum

Foreword 7

List of authors 8

Archaeological evidence on the early metallurgy at Çayönü Tepesi 13

Aspects of early metal use in neolithic Mesopotamia 31

Early metalworking at Çayönü 37

Frühneolithische Metallurgie von Aşikli Höyük 45

The importance of metallurgical data for the formation of Central Transcaucasian chronology 67

Zum frühen Auftreten der Zinnbronze

in Mittelasien 103

A heartland of metallurgy Neolithic/Chalcolithic metallurgical origins on the Iranian Plateau 107

A view on prehistoric Iranian metalworking: Elemental analyses and metallographic examinations 121

Early metallurgy at Arslantepe during the Late Chalcolithic and the Early Bronze Age lA-IB periods 141

The earliest Steps to copper metallurgy in the western Arabah 149

Metais and metallurgy in the Old Babylonian Period 167

Paradigms of metallurgical innovation in prehistoric Europe 175

Les minerais de cuivre et les objets metalliques en cuivre à antimoine-argent du sud de la France.

Preuves d'une exploitation minière et métallurgie du début du III⁰ millenaire av. J.C. 193

The earliest metalworking in southern Britain 211

Resource availability and metal supply in the insular Bronze Age 227

Die Anfänge der Metallurgie an der westlichen Schwarzmeerküste 237

Chemische Zusammensetzung der Kupferartefakte aus dem frühbronzezeitlichen Gräberfeld von

Jelsovce, Südwestslowakei 247

Early copper metallurgy in Austria and methods of assessing its impact on the environment 255

On the early metallurgy of gold: Some answers and more questions 267

Tracing ancient tin via isotope analyses 277

Tin isotope studies of experimental and prehistoric bronzes 285

Experimentelle Rekonstruktion des Gebrauchs von frühbronzezeitlichen Blasdüsen aus der Schweiz:

Kupferverhüttung und Bronzeguß 291

Recasting the Nahal Mishmar Hoard: Experimental archaeology and metallurgy 295

Numerical Simulation of historical casting processes for bronze 301

High temperature technology in antiquity: A sourcebook on the design and Operation of ancient furnaces 305

Naturwissenschaftliche Untersuchungen zur frühen Technologie von Kupfer und Kupfer-Arsen-Bronze 317

Bemerkungen zur Vergleichbarkeit von Analysengruppen mit archäologischen Daten 333

Farbtafeln/Plates 353

The International Symposium "The Beginnings of Metallurgy" was held at the Deutsches Bergbau-Museum in Bochum (DBM) from 24 to 26 April 1995. It was the third venue in a series after Heidelberg 1987 and Donji Milanovac 1990 dedicated to this special topic. It was jointly organized by the DBM and the Max-Planck-lnstitut für Kernphysik, Heidelberg. Archaeometallurgy as an interdisciplinary research field in its own right owes much to the impetus provided by the Volkswagen Foundation through the funding program established in 1986 for the Support of investigations in the development and spread of early metallurgy - probably the most decisive Stimulus for scientific work in this field worldwide! The proceedings of the Symposium in Bochum in 1995 may be considered as a review of the activities at the end of this funding program.The title of the Conference refers to one of the main questions of archaeometallurgy that were tackled by several research projects within this program.

The Symposium in Bochum gathered 91 colleagues from many countries all over the world who presented 43 contributions in oral lectures or posters. Many colleagues took the Chance to attend post-conference field trip over three days to the famous traditional mining region at the Erzgebirge. This excursion focussed on silver and tin mining. The group studied unique medieval silver mining in the "Reiche Zeche" of Freiberg and tin mining at the large pinge of Altenberg and at the "Tiefer Bünaustollen" near Zinnwald. The world famous mineral collection of the Bergakademie Freiberg will be remembered by many as will be the traditional miner's party in the old count house of the "Reiche Elisabeth". This volume on "The Beginnings of Metallurgy" contains 28 articies, i.e. almost 70 percent of the contributions to the Symposium. A special chapter is dedicated to "Native Copper" as perhaps the decisive Stimulus for the very beginnings of metallurgy. Two chapters contain "Regional Studies", i.e. reviews of early metal artefacts and models of developments in southwest Asia and in Europe. The last chapter, "Metallurgical - Technical Studies", contains contributions on gold and tin, on early smelting and casting processes, the evaluation of chemical data and slag investigations. All manuscripts were refereed and the authors were given the opportunity to revise their manuscripts. We gratefully acknowledge the help of many colleagues, in particular lan Freestone, Robert Heimann, Anja Herold, James D. Muhly, Volker Pingel, Edgar Pusch, Bernt Schröder, David Scott and Gerd Weisgerber.

With this volume, we present another contribution to archaeometallurgy and hope that it will receive as much popularity as the last one ("Old World Archaeometallurgy", Der Anschnitt, Beiheft 7,1989) which was soon out of print.

Bochum, September 1999

Andreas Hauptmann

Ernst Pernicka

Thilo Rehren

Ünsal Yalçin

Dr. Paul Ambert

Centre National de la Recherche Scientifique

190 rue des Genévriers Le Clos

F - 34830 Clapiers

Prof. Dr. Hans-Gert Bachmann

Wildaustraße 5

D - 63457 Hanau

Prof. Dr. Marcella Frangipane

Dip. di Scienze dell'Antichità Università La Sapienza Via Palestro 63

l -00185 Roma

Dr. Randolph Haggerty

Ancient Metallurgy Research Group

Department of Archaeological Sciences

University of Bradford

GB - Bradford West Yorkshire, BD7 1 DP

Dr. Jozef Bátora

Archeoloicky ústav Slovenskej akademie vied

94921 Nitra

SLOWAKIA

Prof. Dr. Friedrich Begemann

Max-Planck-lnstitut für Chemie (Otto-Hahn-Institut)

Saarstraße 23

D-55122 Mainz

Dr. Paul Budd

Ancient Metallurgy Research Group

Department of Archaeological Sciences

University of Bradford

GB - Bradford West Yorkshire, BD7 1 DP

Prof. Dr. Wolfgang Bunk

Lehrstuhl für Angewandte Metallkunde

RWTH Aachen

Kopernikusstraße 14

D - 52074 Aachen

Dr. Paul T. Craddock

The British Museum

Department of Scientific Research

Great Russel Street

GB - London WC1 B 3DG

M.John M. Duke

SLOWPOKE Reactor Facility

University of Alberta

CAN - Edmonton, Alberta T6G 2G6

Prof. Dr. Ufuk Esin

Prehistorya Anabilim Dali

Edebiyat Fakultesi

Istanbul Üniversitesi

TR - 34459 Istanbul

Walter Fasnacht

Schweizerisches Landesmuseum

Hardturmstraße 185

CH - 8005 Zürich

Prof. Dr. Alex N. Halliday

Department of Geological Sciences

University of Michigan

Ann Arbor Ml 48109-1063

USA

PD Dr. Andreas Hauptmann

DMT-Deutsches Bergbau-Museum

Institut für Archäometallurgie

Am Bergbaumuseum 28

D - 44791 Bochum

Dr. Karsten Hess

DMT-Deutsches Bergbau-Museum

Institut für Archäometallurgie

Am Bergbaumuseum 28

D - 44791 Bochum

Konrad Kallas

Max-Planck-lnstitut für Chemie

(Otto-Hahn-Institut)

Saarstraße 23

D-55122 Mainz

Dr. habil. Giorgi Leon Kavtaradze

Institute of History and Ethnography

Academy of Science

Melikischwili Street 10

Tbilisi

GEORGIA

Prof. Dr. Ingo Keesmann

Arbeitsgruppe Archäometallurgie

Institut für Geowissenschaften, FB 22

Johannes Gutenberg-Universität Mainz

Saarstraße 21

D-55122 Mainz

Prof. Dr. Robert Maddin

7 Highwood Way, P.O.B. 568

New Seabury, MA 20649

USA

Peter D. Marshall

English Heritage

Ancient Monuments Laboratory

23 Savile Row

GB-London W1X1 AB

Dr. Rona A.R. McGill

Ancient Metallurgy Research Group

Department of Archaeological Sciences

University of Bradford

GB - Bradford West Yorkshire, BD7 1 DP

Dr. Auxiliadora Moreno Onorato

Facultad de Filosofia y Letras

Departamento de Prehistoria y Arqueologia

Universitad de Granada Cartuja

E-18071 Granada

Prof. Dr. James D. Muhly

American School of Classical Studies

54 Souidias Street

G - 10676 Athens

Dr. Peter J. Northover

Department of Materials

University of Oxford

Parks Road

GB - Oxford 0X1 3PH

Dr. William O'Brien

Department of Archaeology

National University of Ireland

IR - Galway

Dr. Asii Özdoğan

Prof. Dr. Mehmet Özdoğan

Prehistorya Anabilim Dali

Edebiyat Fakültesi

Istanbul Üniversitesi

TR - 34459 Istanbul

Sarah L. O'Hara

Dept. of Archaeology and Archaeological Science

University of Sheffield

2 Mappin Street

GB - Sheffield, S1 4DT

Dr. Barbara S. Ottaway

Dept. of Archaeology and Archaeological Science

University of Sheffield

2 Mappin Street

GB - Sheffield, S1 4DT

Dr. Alberto M. Palmieri

Consiglio Nazionale delle Ricerche

Cas. Post. 10

l - 00016 Monterotondo Stazione/Roma

Prof. Dr. Ernst Pemicka

Lehrstuhl für Archäometallurgie

TU Bergakademie Freiberg

Gustav-Zeuner-Straße 5

D - 09599 Freiberg

Dr. Vincent C. Pigott

Museum Applied Science Center for Archaeology

(MASCA)

The University Museum of Archaeology and

Anthropology University of Pennsylvania

3rd and Spruce Streets

Philadelphia, PA 19104-6324

USA

Prof. Dr. A. Mark Pollard

Ancient Metallurgy Research Group

Department of Archaeological Sciences

University of Bradford

GB - Bradford West Yorkshire, BD7 1 DP

Markus Ratka

Lehrgebiet Wärmetechnik und Industrieofenbau

RWTH Aachen

Kopernikusstraße 16

D - 52056 Aachen

Dr. John E. Rehder

36 Castle Frank Road 309

CAN - Toronto M4W 2Z7

Dr. Karin Reiter

Benzmannstraße 10

D-12167 Berlin

Dr. Volodja D. Ruzanov

Institute of Archaeology

Academy of Science Uzbekistane Republic

3, Academic Abdullayewa St.

703051 Samarkand

UZBEKISTAN

Prof. Dr. Peter R. Sahm

Lehrstuhl für Angewandte Metallkunde

RWTH Aachen

Kopernikusstraße 14

D - 52074 Aachen

Prof. Dr. Edward Sangmeister

Institut für Ur- und Frühgeschichte

Albert Ludwig-Universität

Beifortstraße 22

D - 79098 Freiburg

Dr. Brett Scaife

Ancient Metallurgy Research Group

Department of Archaeological Sciences

University of Bradford

GB - Bradford West Yorkshire, BD7 1 DP

Ulf-Dietrich Schoop, MA

Institut für Ur- und Frühgeschichte

Jüngere Abteilung

Schloß Hohentübingen

D - 72070 Tübingen

Dr. Sigrid Schmitt-Strecker

Max-Planck-lnstitut für Chemie

(Otto-Hahn-Institut)

Saarstraße 23

D-55122 Mainz

Dr. Sariel Shalev

Weizmann Institute of Science

IL-Rehovot 76100

Dr. Tamara Stech-Smith

422 Upper Hollow Hill Road

Stowe, VT 05672

USA

Prof. Dr. Henrieta Todorova

Archäologisches Institut der Bulgarischen Akademie

der Wissenschaften

Boulv. Stambolijski 2

1000 Sofia

BULGARIA

Prof. Dr. Michael L. Wayman

Department of Chemical and Materials Engineering

University of Alberta

CAN - Edmonton, Alberta T6G 2G6

PD Dr. Ünsal Yalçin

DMT-Deutsches Bergbau-Museum

Institut für Archäometallurgie

Am Bergbaumuseum 28

D - 44791 Bochum

Dr. Wen Yi

Department of Geological Sciences

University of Michigan

Ann Arbor MI 48109-1063

USA

Suzanne M.M. Young

Archaeometry Laboratories

Harvard University

Peabody Museum

11 Divinity Avenue

Cambridge MA 02138

USA

Prof. Dr. Abdolrasool Vatandoust

Director, Research Centre for Conservation of Cultural

Relics (RCCCR)

Iranian Cultural Heritage Organization

P.O.B. 11365-4834

Tehran 11365

IRAN

The importance of metallurgical data for the formation of a Central Transcaucasian chronology

Archaeology in Georgia, as in other countries, is the science which studies human activities in the past and tries to reconstruct this past as comprehensively as possible. It was stated that the past is the main thing in our life, everything that exists belongs to it (A. France). Indeed, to reconstruct the past, archaeology needs as many ingredients based on the full range of technical and natural sciences as life itself is diverse. More and more archaeology becomes a meeting field for various sciences.

As scientific development is easily attainable in the zones of contacts and interactions between different sciences, completely new perspectives are opened for archaeology through its integration in other sciences. Archaeometallurgy is among the most important branches developed in consequence of this qualitative change - or better: the transformation of archaeology.

 In Tbilisi, the capital of Georgia, three laboratories carry out analysis of metal artefacts: the State Museum of Georgia, the Metallurgical Institute and the Centre for Archaeological Studies. The metal inventory was investigated by Josef Grdzelishvili, Ferdinand Tavadze, Tamar Sakvarelidze, Rusudan Bachtadze, Tsisana Abesadze, Tina Dvali, Givi Inanishvili, Teimuraz Mudzhiri, Natela Saradzhishvili and others.

The study of metal and other kinds of artefacts, together with chronological and environmental studies, are usually considered as three of the prime areas of modem archaeological science. At the same time chronological studies are essentially connected with artefact studies. Already in the first half of the 19th century, Christian Thomson based the first archaeological periodisation on the kind of substances used for the artefacts and classified archaeological material by the chronological order as belonging to the Stone, Bronze and Iron Ages. This correlation of time and type of material in use was known even to the old Greeks.

Among all types of artefacts, metal objects in general and tools and weapons in particular, are subjected most of all to innovations - the development of society is considerably connected with their functional abilities. Therefore metallurgical data of the ancient societies - of one and the same geographical zone - have, in contrast to the data of other archaeological sources, such as pottery, architecture, burial habits and others, which are more apt to indicate the genetical relations, a special importance in the establishment of a relative chronology.

The first and second ..radiocarbon revolutions", the use of the radiocarbon dates for the creation of absolute time-scales first and the use of calibrated ¹⁴C dates afterwards, provoked the separation of the areas dated by the ¹⁴C technique - the northern periphery of the Near East and Europe - from the areas with historical chronologies, /.e. the Near East. The separation of these two regions from each other caused something like a "geological gap" - a "fault line" between them (Renfrew 1973: 104, Figs. 20, 21). The need to fill this gap is an urgent task of the contemporary archaeological studies. Besides the further improvement of the geochronological methods, it demands an intensive stimulation of the research in the field of relative chronology on both parts of the above-mentioned gap, and, as much as it is possible, to connect them.

One of the regions along the ,,fault line" is the Caucasus. Therefore chronological problems of this region have paramount importance in the foundation of a general Near Eastern - East European chronological system; it seems that the Caucasus is an important link in the Old World's chronological chain. The inclusion of the Caucasian chronological evidence into the common Near Eastern - East European chronological system must be preceded by the formation of an all-Caucasian chronological scale.

The Great Caucasian Ridge represents a barrier dividing the Caucasus in two main parts: Transcaucasia or the South Caucasus, and the North Caucasus. At the same time, the role of the pathes crossing it permits to consider the Caucasus as one and the same geo-political zone. Among the Caucasian regions Central Transcaucasia (i.e. Eastern Georgia, old Iberia) holds a key position (Fig. 1) - it is encircled by all other Caucasian regions (Western, Southern and Eastern Transcaucasia, North-Western and North-Eastern Caucasus), and therefore it represents a basis for the elaboration of the all-Caucasian chronological scale (Figs. 2, 3).

67

87. Taşkun Mevkii

88. Nadarbazevi

89. Irganchai

90. Ozni

91. Kulbakebi

92. Medzhvriskhevi

93. Chagar Bazar

94. Hassuna

95. Ust-Jegutinski

96. Bakurtsikhe

97. Telebi

98. Sachkhere

99. Vanadzor
lOO. Iğdir

68

Fig. 2: Traditional chronological framework for the Eneolithic-Bronze Age cultures of the Caucasus.

The earliest metal artefacts of the Caucasian zone originated from this region - Central Transcaucasia. They appeared in the layers of the settlements of Shulaveri-Shomu Tepe culture of Central and Eastern Transcaucasia - in sites on the middle flow of Kura - Khramis Didi Gora and Arukhlo/Nakhiduri I in South-Eastern Georgia and Gargalar Tepesi in the western part of Azerbaijan. In the lower levels (VII, VI) of Khramis Didi Gora, in a depth of 5.09 m, a semicircular hollow object with pointed ends

and four beads were found (Kiguradze 1986: 93f.; Menabde et al. 1980: 34) with following contents: Cu - high, Sn - 0.03, Pb - 0.001, Ag - 0.002, Fe - 0.02 and Cu - high, Sn - 0.001, Ag - 0.002, Fe - 0.01 (Tavadze et al. 1987: 46). In the unstratified layers of Arukhlo/Nakhiduri I, an unidentified object of metal plate was detected (Dzhavakhishvili et al. 1987: 8). A cylindrical bead made of twisted plate was discovered in Gargalar Tepesi (Arazova et al. 1972: 435). According to R. Munchaev, it is similar to the beads excavated by him (together with N. Merpert) in the Hassuna period levels of Yarim Tepe I, in the Sindzhar

 69

 valley, Iraq (Munchaev 1982; cf. Merpert & Munchaev 1977: 157, Fig. 1.4.) (Fig. 4.1). These levels of Yarim Tepe are dated to the first half of the sixth millennium B.C. (Porada et al. 1992: vol. l: 81-83; vol. II: 94, Fig. 2).

All Transcaucasian sites mentioned above belong to the final stage of the Shulaveri-Shomu Tepe culture according to the periodization by Georgian archaeologists (Kiguradze 1986: 100). The dating of Shulaveri-Shomu Tepe culture is based on radiocarbon dates. According to the calibrated ¹⁴C dates, this culture belongs mainiy to the sixth millennium.

The ¹⁴C dates of this culture are: Shulaveris Gora, 0.2 m, 5213-4460 cal B.C. (TB-15), 2.4 m, 5677-5330 cal B.C. (TB-16). Repeated analyses of the same sample (TB-16) are: TB-72, 5588-5482 cal B.C. and SOAN-1292, 5064-4832 cal B.C. Shulaveris Gora, 1.6 m, 5621-5522 cal B.C. (LE-1099), 0.1 m, 5416-5077 cal B.C. (LE-1100); Imiris Gora, from the levels IV-I, 5332-5077 cal B.C. (TB-27); ArukhIo/Nakhiduri l, upper strata, 5473-5384 cal B.C. (TB-92), level II, 5671-5584 cal B.C. (TB-277), lower strata, 6007-5886 cal B.C. (TB-300), level VI, 5677-5585 cal B.C. (TB-309); Khramis Didi Gora, level V, 5560-5386 cal B.C. (LJ-3270), middle strata, 5433-5291

70

cal B.C. (TB-301), 5.4 m, 5446-5342 cal B.C. (TB-322); Gargalar Tepesi, lower strata, 5200-4944 cal B.C. (LE-1084), lowest stratum, 5662-5579 cal B.C. (LE-1083); Toira Tepe, 2 m, 5208-4839 cal B.C. (TF-372); Shomu Tepe, 1 m, 6413-6221 cal B.C. (LE-631). Calibrated ¹⁴C dates partially solve the discrepancy between the Near Eastern parallels dated to the seventh-sixth millennia and the uncalibrated ¹⁴C dates of the Shulaveri-Shomu Tepe culture which were largely placed in the fifth millennium (all the dates used here are obtained with one sigma, v. Stuiver & Reimer 1993: 215-230). We bear in mind the assumption about the special closeness of this culture in all stages of its existence with the Hassuna culture on the one hand and with the Umm Dabaghiah-Tell Sotto culture of the Pre-Halafian period on the other.

It seems that the decorations of the Umm-Dabaghiah pottery are not as analogous to the ornaments of the Arukhlo/Nakhiduri I, as J. Mellaart thought (Mellaart 1975: 304), but to the pottery of an earlier site - Imiris Gora (cf. Kavtaradze 1981: Table I and II). As to some Georgian archaeologists, a similarity can also be observed between small figurines of the upper levels of Khramis Didi Gora - a site which belongs to the final stage of the Shulaveri-Shomu Tepe culture - and similar figurines which were discovered in the layers of the Hassuna, Samarra and Halaf cultures (Glonti et al. 1975: 97). All these Mesopotamian cultures and sites are dated mainly to the sixth millennium, and it is obvious that the Shulaveri-Shomu Tepe culture from the point of view of typological and chronological data is quite comparable with them - the same stage of their development can be stated without any doubt.

Although all metal artefacts of the Shulaveri-Shomu Tepe culture originate from the building layers of its later stage, it seems possible to consider this culture as mainly Early Eneolithic (Chalcolithic) because of the obvious signs - observed by some specialists - the degradation of flint industry and impoverishment of the sets of stone tools, together with a lack of certain categories of artefacts, e.g. geometrical microliths as a mass series from its layers known up till now as the lowest (Chubinishvili & Chelidze 1978: 66; Chelidze 1979: 30).

We must also take into account the favourable conditions existing in Transcaucasia for metallurgical activities. There, in the mountains, accumulations of the products of oxidation zone were formed; the layers of copper on the surface could have satisfied the needs of early metalworkers. Exactly the southern region of Central Transcaucasia is supposed to be the basis of raw material for the development of initial copper metallurgy in the Caucasus.

Although there were requirements available for the development of local metalworking, nothing definite can be said about the first steps of metallurgy in the Caucasus, and particularly in Central Transcaucasia. It is supposed that metal artefacts of the Early Farming culture were made of local arsenic-copper ores, but it is not clear how the early metallurgists extracted the ore.

More than thirty years ago, in the Marneuli district, south of Tbilisi, near the village of Tsitelisopeli, traces of an ancient working place with lumps of slags and large grooved stone hammers were discovered (Lordkipanidze 1989: 104, note 11). These objects are quite similar to the well-known hammers (chisels) found in Arukhlo/Nakhiduri I, Kül Tepe I and other early sites. Such chisels are thought to have been used for the extraction of ore, but an exact dating of this working place cannot be considered as finally solved, because similar tools were also characteristic of the Kura-Araxes culture of the fourth and third millennia (Chubinishvili 1971: 30; Kushnareva & Chubinishvili 1970: 113, Fig. 5, 19).

Thirty years ago, it was thought that metalworking was introduced into Transcaucasia only at the time of the Kura-Araxes culture from the Near East. Today the same questions are raised in connection with the Early Farming culture of Transcaucasia. We must take into account the metal artefacts of the South Transcaucasian Early Farming sites (e.g., Kül Tepe, Teghut etc.) too, which mainly belong to a time rather later than the Central and Eastern Transcaucasian Shulaveri-Shomu Tepe culture, and because of that there is no sufficient reason to unite it with the latter.

In Central Transcaucasia, the culture of this period, tentatively referred by us to the Middle Eneolithic Age, is represented by Tsopi, Sioni of the Shulaveri ravine, Delisi, the lowest level of Berikldeebi, sites of the Aragvi ravine and the Alazani valley. This culture is intermediate between the Shulaveri-Shomu Tepe and the earliest materials of the Kura-Araxes culture and displays a certain similarity with the preceding and subsequent cultures.

The metal artefacts supposedly of this period were detected in Delisi (northwestern part of Tbilisi). They represent bronze pins and an awl which contain, together with other elements, up to 4% tin. Apart from the chemical composition of such early artefacts dated even by the traditional chronology to the end of the fifth - beginning of the fourth millennia (Abesadze & Bakhtadze 1987: 51), certain doubts arise concerning the circumstances of their discovery.

Unfortunately we have only one ¹⁴C date for this period from Zhinvali in the Aragvi ravine, 5206-4807 cal B.C. (TB-326). Therefore, in order to date the Middle Eneolithic culture of Central Transcaucasia, an attempt can be made by taking into consideration the data from contemporary sites of other parts of Transcaucasia, which at the same time are richer in metal inventory. E.g. in the lowest level of Kül Tepe I of Nakhichevan, considered as contemporary with the final stage of the Shulaveri-Shomu Tepe culture, seven copper artefacts: a small quatrilateral piercer, a rhomboic copper plate

71

object (maybe an arrowhead), two beads and three fragments of unidentified objects were discovered (Fig. 4.2-5). All of them, except the quatrilateral piercer, were found in the depth of 18-17 m and contained an admixture of arsenic; the piercer, which apart from arsenic also has nickel (1.6%), was discovered in a depth of 15.05 m (Abibulaev 1963: 161f, Fig. 4; id. 1982: 78f; Selimkhanov & Torosiyan 1969: 230).

It was supposed that because of the absence of nickel-copper deposits in the Caucasus this object was imported to Kül Tepe from the Near East (Kushnareva & Chubinishvili 1970: 120, 129), but nickel is characteris-tic of easternmost Georgian, Kakhetian, artefacts of later times (Abesadze 1980: 148). The fact that in the depth of 19 m of Kül Tepe l, a pot, typical of the Halaf culture, was found, is usually considered as an indication of the connection of the eneolithic population of Transcaucasia with the Near East.

We must underline the fact that in the same lower levels of Kül Tepe l, in a depth of 16.85-20.84, of all the eneolithic layers between 12.18 and 21.1 m, together with Halafian Imports, sherds of the Dalma painted wäre of the Solduz valley of North-West Iran were found (Munchaev 1975:128f). The Dalma culture undoubtedly is contemporary with Ubaid 3 (Voigt 1992: 158, 175), and it seems that the lower levels of Kül Tepe l must be dated to the period, when the end of the Halaf culture was slightly overlapped with the Early Northern Ubaid, that means to the beginning of the fifth millennium. We can consider this date as a terminus post quem for the later layers of Kül Tepe l as well as for the Middle Eneolithic period of Transcaucasia, and at the same time as a terminus ante quem for the Shulaveri-Shomu Tepe culture or the Early Eneolithic.

The ¹⁴C date for Kül Tepe l from a depth of 18.2 m is 4763-4506 cal B.C. (LE-477), somewhat later than the ¹⁴C date of the Dalma culture - 4947-4782 cal B.C. (P-503). The stratigraphy of Dalma Tepe is usefui from the point of view of the chronology of Transcaucasian sites. As it was observed, if the painted pottery typical of the lower levels of this site was represented in the Kül Tepe l and Mil-Karabagh sites, then the Impressed Ware, typical of the Late Dalma, was found in llanly Tepe and the sites of Misharchai and Guru Dere l in the steppe of Mughan, Azerbaijan (Munchaev 1975: 128f). At the same time, the Late Dalma Impressed Ware provides a chronological link with the Early Siahbid phase in the Kermanshah region and it is not represented in the Late Siahbid deposits (Voigt 1992: 158, 175). Sherds of Dalma Impressed Ware are also found at the Ubaid sites as Abada and Kheit Qasim in the Hamrin and Yorgan Tepe near Kirkuk. On the other hand, at Dalma Tepe sherds characteristic of Tepe Gawra XVI (or of the Ubaid 3 period) are represented (Voigt 1992:175). Because of that, a dating of the layers of Dalma Tepe and the Transcaucasian sites containing Early and Late Dalma Ware in the first half and middle of the fifth millennium B.C., can be proposed.

From East Transcaucasian sites metal artefacts were found in Chalagan Tepe (near Agdam) - two copper pins in the burials and an awl in a building level (Dzhavakhishvili et a/. 1987: 8). The ¹⁴C date of this site is 5411 -5259 cal B.C. (TB-318). Also in Azerbaijan, in the steppe of Mughan, in Alikemektepesi, a copper bead and an awl were found (Dzhavakhishvili et al. 1987: 8). This site is important from the chronological point of view, because in the lower levels material comparable to the Kül Tepe l was discovered, and in the upper levels pottery of the North Ubaid type, similar to the ex-amples found at the Armenian site Teghut (Narimanov 1980: 93, 102f, 271; cf. Munchaev 1975: 120). This fact has a special importance for defining the chronological Position of the Central Transcaucasian Middle Eneolithic because in Alikemektepesi, in the upper levels, aside from pottery of the North Ubaid type, sherds with combed surface and burnished interior (Narimanov 1980: 78, 208) like the pottery from Sioni, and quite un-known in Kül Tepe l, were found.

Several metal objects were discovered in Teghut, Ararat valley: a knife (Fig. 4.6), a drill and two fragments of quadrilateral awls. All of them are made of arsenical copper with an admixture of other minor components. The knife contained 5.4% of arsenic, a piece of awl 3.6% (Selimkhanov & Mareshal 1966: 145f, Table 3). The fact that certain types of copper artefacts, as weapons and tools, had for the time concerned a rather high content of arsenic, was considered as an indication of the existence of artificial alloys (Gevorkiyan 1980: 37; Narimanov & Dzafarov 1988: 22; Kushnareva 1993: 205), but recently it was suggested that in a variety of areas the real development of alloying probably only appeared with the use of tin-bronze (Northover 1989:117).

The difference of admixtures in the content of the Transcaucasian metal inventory usually is explained by various types of ores. For the southern part of Transcaucasia, in Opposition to the northern regions, the deposits of Dzhulfa-Zangezur, Madzor and Kafan were possibly used (Dzhaparidze 1989: 229).

In the opinion of some archaeologists, a new ethno-cultural element - the group of the tribes of the Ubaid culture - spread at that time to the Caucasus (cf. Narimanov 1991: 32). But in this respect we must recall H. Nissen's Suggestion about the explanation of the wide distribution of the Ubaid-like pottery with the introduction of the tournette or „slow-wheel" for the manufacture of pots (Nissen 1988: 46). We must also take into consideration the possibility of an interconnection between the manufacture of the highiy fired Ubaid pottery and the real smelting procedure of copper ore, attainable only at a temperature higher than 1100 °C (cf. Pernicka 1990: 46, 117).

72

houses appeared (Tobler 1950: Tables VI, VIII), typical of the Early Farming culture of Transcaucasia. It is interesting that the population of the XII and XI A levels used in Tepe Gawra various types of copper ores. The copper of the later level differs in the high content of arsenic together with some other components (nickel etc.) (Tobler 1950: 212).

As to Western Transcaucasia (i.e. Western Georgia, old Colchis), it is not quite clear which copper deposits were used there. The first copper artefacts of this region, hooks from Sagvardzhile, ascribed to the Eneolithic period, resemble the forms of bone examples. At the same site also a quadrilateral awl, made of pure copper by cold forging, was discovered (Lordkipanidze 1989: 67). In the eneolithic level of Tetri Mgvime (near Kutaisi) a copper dart or a leaf-like knife, which contained up to 0.7 % of arsenic, was found (Abesadze & Bakhtadze:1987: 51).

The ¹⁴C date of Machara IV (Abkhazia) is 4754-4495 cal B.C. (LE-1347).

In the following Early Bronze Age the Western Transcaucasian metal inventory was manufactured already of arsenic-copper alloys. Moulds of axes were discovered in Ispani (near Kobuleti) and also in Pichori (near Gali) of the later type; in both sites together with other signs of metallurgical activities.

The remains comprising the material of the early stage of the Kura-Araxes culture, including the Didube-Kiketi and Sioni (the lori river valley)-Gremi groups, referred to the Late Eneolithic period of Central Transcaucasia, have an extremely poor metal inventory. In Central Transcaucasia the Kura-Araxes culture is presumably dated mainly to the fourth millennium, and apart from the Late Eneolithic it comprised the first phase of the Early Bronze Age and partially, in the first quarter of the third millennium, also the second phase of the same period.

The best known sites with fixed stratigraphy of the Kura-Araxes culture of Central Transcaucasia are Khizanaant Gora, Kvatskhelebi (near Kareli) and Tsikhia Gora (near Kaspi) in the central and Amiranis Gora (Akhaltsikhe) in the south-western parts of the region. We have a few ¹⁴C dates for this period: 3644-3376 cal B.C. (LE-157) from Kvatskhelebi C 1; 3636-3356 cal B.C. (TB-831) (this date, TB-831, 2900 ±110 B.C., is published with a half-life of 5730 ± 40 year in Makharadze 1994: 61) from Tsikhia Gora, level B 2 of the final period of the Kura-Araxes culture of Shida Kartli, and 3790-3373 cal B.C. (TB-4) and 3630-3048 cal B.C. (TB-9) from Amiranis Gora. The dates from Amiranis Gora are generally in agreement with the ¹⁴C date received from the lowest layer of the same Kura-Araxes culture in Pulur (Sakyol), Eastern Anatolia: 3500-3336 cal B.C. (P-2040). Other dates of the latter site are: level IX, 2890-2409 cal B.C. (M-2173), level VIII, 3346-2888 cal B.C. (M-2172),

73

level VI, 2866-2203 cal B.C. (M-2171), level V, 3092-2669 cal B.C. (M-2170). It is interesting that the earliest „Kura-Araxes" material of Pulur (Sakyol) reveals traits typical of Amiranis Gora (Kavtaradze 1983: 89f).

The more or less contemporary Kül Tepe II ¹⁴C date should also be taken into consideration: 3766-3543 cal B.C. (LE-163). Recently three dates were received from the AMS Facility at the University of Arizona for Satkhs, the site which is situated in Dzhavakheti (8 km northeast of Nino Tsminda), i.e. in the southeast direction from Amiranis Gora and Kura-Araxes layers of which have ceramic parallels with Mokhra Blur (Ararat valley), Kvatskhelebi and Amiranis Gora: 3072-2916 cal B.C. (AA-7768), 3343-3043 cal B.C. (AA-12853) and 3301-2926 cal B.C. (AA-12854) (Isaak et a/. 1994: 26, 28f). One date was obtained from a level associated with Early Bronze Age materials of the north-west Armenian site Horom in the Shirak valley: 3371-3136 cal B.C. (AA-7767) and two dates were from a tomb of the same site: 3341-3048 cal B.C. (AA-10191) and 3990-3823 cal B.C. (AA-11130). All three vessels of this tomb reveal in the opinion of the excavators relatively early forms of the Kura-Araxes culture (Badaljan et al. 1994: 14,Table Illc).

Special attention must be paid to the first ¹⁴C date of Amiranis Gora 3790-3373 cal B.C. (TB-4), because it was obtained from the charcoal of the metallurgical workshop which belonged to the earliest building horizon of Amiranis Gora (Kushnareva & Chubinishvili 1970: 114) (Fig. 5.1). Undoubtedly this fact, together with the other data, is an evidence of a division of the metallurgical production in the extractive and processing branches.

In Amiranis Gora an arched kiln of stone had been constructed which contained heavily grounded technical charcoal, necessary for achieving high temperatures. The supplies of such charcoal, from which abovemen-tioned ¹⁴C date stems, were stored in a big clay vessel, discovered in the floor level of the same workshop. There were also clay tuyeres - a real confirmation of the smelting procedure - as well as a clay mould for a pig (Chubinishvili 1971: 57f) (Fig. 5.2-4). Clay tuyeres were found, too, in Kül Tepe II and Misharchai (Makhmudov et al. 1968: 19, Fig. 4,2) (Fig. 5.7).

A furnace of another type was discovered in Baba-Dervish II, Azerbaijan. There, on the periphery of the settlement, three oval pits - the foundation of the melting kilns with vaults of clay above - were found. Two of these furnaces had special openings with a ditch - an evidence of the technique of blowing (Fig. 5.5). This circumstance is confirmed by the fact that there also clay tuyeres were found (Fig. 5.6). Another remainder of the smelting process at that site was discovered in the form of technological waste - slags (Kushnareva & Chubinishvili 1970:114f, Fig. 40; Chubinishvili 1971; 102f).

Slags were also found in Khizanaant Gora, Kül Tepe II and Garni (Armenia). In these sites traces of metal on the walls of vessels, ladles, crucibles and moulds of f'ire-proof clay for the pouring of the liquid metal were discovered, too (Kushnareva & Chubinishvili 1970:114, Fig. 40, 17) (Fig. 6.13). In the building layers of Khizanaant Gora, a sickle was found, the forging of which was not finished (Kushnareva & Chubinishvili 1970:114).

The main part of the metal inventory of the Kura-Araxes culture was made of arsenical copper. The content of arsenic in alloys reaches on average from 2 to 8% (Tavadze et al. 1987: 45). The distribution of the lead admixtures in the greater part of the copper inventory of this period indicates in the opinion of Georgian archaeometallurgists that ores were used which contained natural arsenic admixture. Such natural alloys were possibly obtained from polymetallic ores, rich in elements (Abesadze & Bakhtadze 1987: 52). In the Caucasus the copper deposits are of the polymetallic type which contain from 3 to 10% and more the so-called usefui components - arsenic, antimony, tin, lead, zinc, iron, nickel, silver etc. Among these elements arsenic is mainly associated with copper, which occurs in the territory of Central Transcaucasia in the form of realgar and auripigment types (cf. Abesadze & Bakhtadze 1987: 52).

It is supposed that at the time of the Kura-Araxes culture, comparatively easily smeltable polymetallic-arsenical ores were used. The metal artefacts of this period did not contain sulphur, a fact which possibly indicates the use of the carbonized and oxidized upper layers of copper deposits (cf. Rapp 1989:107-110). The spectral analyses of the artefacts of the Kura-Araxes culture and their correlation with the composition of the copper ores of the Great and Little Caucasus demonstrate that at that time metal was gained by the process of straight reduction (Tavadze et al. 1987: 45).

74

Together with arsenic, various amounts also of other ore admixtures were transmitted in the metal, since the recovery of them was conditioned by the process of smelting. Therefore it was supposed that all the content of these admixtures were originally in the ore (Abesadze & Bakhtadze 1987: 51 f). But it is quite obvious that ancient metallurgists in Transcaucasia were intentionally choosing copper with high or low content of arsenic, according to the functional destination of the artefacts; e.g. for ornaments from 7 to 22.7%, for tools and weapons much less, otherwise their functional abilities would be significanly low (cf. Kushnareva 1993: 235). In this connection it is worthwhile to note that the copper inventory of the late fourth millennium hoard of Nahal Mishmar (Palestine) with a very high concentration of arsenic and antimony is explained by the smelting of selected ores and not of artificially manufactured alloys (cf. Pernicka 1990: 48, 50). At the same time, copper alloys with an arsenic content ranging from 3 to 10 % are considered as to be gained by the direct addition of arsenic (Palmieri et al. 1993: 574).

It is interesting that the metal artefacts of various functional use do not differ from each other by their composition in the Kura-Araxes culture. Such a divergence began to exist only in the later part of this period, when some ornaments were manufactured out of high arsenic copper, with the content of arsenic exceeding 10 % and up to 20%: e.g. biconical beads and a rhomboic pendant from Urbnisi, curl-rings from Dzagina (Tavadze et al. 1987: 45; Abesadze & Bakhtadze 1987: 52). In such cases it is possible to suppose that arsenic was intentionally added to increase the melting ability of copper and to obtain an alloy with a silver-like colour. Some artefacts have high amounts of antimony and lead (Tavadze 1987: 45).

Genuine silver ornaments (pendants, spirals, curl-rings) from the territory of Georgia were found in the earliest layers of the advanced stage of Kura-Araxes culture at Amiranis Gora and Kvatskhelebi (Dvali 1974: 62).

In the later period of the Kura-Araxes culture, partially contemporary with the second phase of the Early Bronze Age or the Early Kurgan period of Central Transcaucasia, together with the primitive types of artefacts rather complicated shapes and of various types began to appear in the inventory: flat axes and shaft-hole axes with downward directed butts, spearheads, daggers, bayonet-like weapons, chisels, awls, tools for farming, sickles, pins with T (crutch)-, loop- and double-spiral-shaped heads, axe-adzes, earrings, pendants, bracelets, rings, curl-rings, beads, a diadem (Figs. 6, 7,10) (see below).

From the metallurgical point of view particular interest was given to tools of combined type - an axe-adze which was uniting a wedge-shaped axe with a long beak stretched on the butt. This artefact was found in sites of south-eastern Georgia - in the region rich in metal deposits. Therefore these artefacts were considered as tools used for the obtaining of ore (Lordkipanidze 1991: 50). The clay casting mould for a flat axe from the level C of Kvatskhelebi is similar to the axe from Sachkhere and has early parallels in Near Eastern and East European sites (Dzhavakhishvili & Glonti 1962: 58, Table 4, N 489; Kavtaradze 1983: 85).

It is a widespread view that the metals from the Transcaucasian ore deposits together with certain types of metal artefacts were distributed in many regions of the Ancient World from the early stages of metallurgical activities. Impulses coming from the Transcaucasian metallurgical centre through the northern Caucasus penetrated wide territories, from the river Volga to the Dniepr and even farther, reaching the Carpathian mountains (Chernykh 1992: 91, 159). In the southern direction metal of Transcaucasian provenance was widely distributed in Anatolia and Syria-Palestine. In the opinion of archaeometallurgists, the research on Anatolian metallurgy should be integrated with the location of both - copper ore deposits and arsenic occurences in the Caucasian regions (Palmieri et al. 1993: 591). It seems that Caucasian metallic ores and metallurgical traditions were used in the Near East at the time when the Transcaucasian population, bearers of the Kura-Araxes cultural traditions, were spread there (e.g. in Geoy Tepe, north-western Iran (cf. Burton Brown 1951)). They migrated in most cases to the south, west, south-west and south-east, from the Transcaucasian homeland of this culture, to southern Palestine, central Anatolia and central Iran. Along with the pottery, metal, obsidian and characteristic architecture and graves, a strong indicator of this culture, a peculiar type of hearth, was distributed in far off lands.

This event chronologically is placed in the early third millennium, but it seems that the second half of fourth millennium was the time of initial penetration of certain elements of the Transcaucasian Kura-Araxes culture and possibly also that of the population - bearers of this culture - in the northern part of the Near East.

It was stated that the metal artefacts from the hoard of the East Anatolian Late Chalcolithic site of Arslantepe (Malatya) VI A hoard (from A 113 Room of Building III) do not belong to the local copper deposits, as they have high arsenic admixtures (up to 4 %) and no trace of nickel, but might rather be of a northern provenance (Burney 1993: 314f). Besides the tradition of arsenic-bronze metallurgy, the northeastern direction is indicated also by twelve poker-butted, leaf-shaped spearheads with a cylindroid mid-rib and at the same time containing 1.3-4.3% of arsenic (Palmieri 1981: 108f, Fig. 4), found together with nine swords and a quadruple spiral piaque of the same hoard. They have dose parallels in the northwestern part of Central Transcaucasia by the similar copper spearhead found in Tsartsis Gora (Sachkhere) (Kuftin 1949: 74, Table LIX) (Fig. 7.3)

76

which likewise contained no nickel but 4.66% of arsenic and 0.30% of antimony (Table 1, no. 21) and which must be dated by the period following to the Central Transcaucasian Kura-Araxes or contemporary with its latest levels. To the character and typology of the Arslantepe hoard, important chronological implications are given for the evaluation of the origin of Anatolian and North Syrian metallurgies (Palmieri 1981: 109; Yakar 1985: 276). Some scientists believe that metal-work was a major item of trade through Arslantepe (e.g. Burney1993: 314).

It is quite probable that the economical importance of Arslantepe VI A as well as of such Late Uruk enclaves and outposts as Hassek Höyük 5, Habuba Kabira-Tell Qanas, Jebel Aruda and Tepecik 3 was the reason of their violent destruction by the intruders from the north - the bearers of the Kura-Araxes culture. This phenomenon has a parallel in the western part of central Iran by the destruction of the Late Uruk colony in Godin Tepe V, which ceased its existence as the result of the invasion of the Kura-Araxes population east of the site, in the Hamadan valley, cutting off commercial routes to the east. After a short interval of time Godin IV emerged, with the material of the Kura-Araxes culture of the Yanik Tepe I type (Weiss & Young Jr. 1975: 15).

It seems that in the second half of the fourth millennium in the northern part of the Near East one and the same phenomenon - the destruction of the sites, revealing traits typical of Late Uruk period by a population of northern provenance, characterized by the red-black, hand-made burnished pottery, the high-arsenic copper metallurgy and certain types of metal artefacts, the ,,wattle and daub" houses and the particular type of hearthes - took place.

The intrusive character of the Kura-Araxes culture in this area became quite obvious after the exposure of the stratigraphical sequence, documented at Arslantepe, where layers containing the material of this culture interrupted the preceding and following local development of the horizons with the Reserved-Slip pottery (Palmieri 1985: 208).

At Kurban Hoyuk (Karababa basin, north-west of Urfa, on the left bank of the Euphrates), in the Late Chalcolithic VI period, which equates with Tell Judeidah (Amuq) phases F-F/G sequence, three fragments of Kura-Araxes pottery (,,Karaz Ware") were discovered. They all are diagnostic and consist of a dense brownish clay with varying amounts of fine grit and chaff tempering. One of them is uniformly black, but two have bichrome surfaces, with orange interior and black exterior (Algaze 1990: 260, 268, Table 42, G,F,H; Helwing 1996: 75). Also all of them resemble by their shape the Kura-Araxes pottery (cf. Sagona 1984: Forms 81, 82 Fig. 36, 2,5,6, Form 34 Fig. 21, 6). There are indications of the long existence of the Karaz Ware in the neighbourhood of the Karababa region, because in the subsequent Early Bronze Age (V and IV) levels of Kurban Höyük a few fragments of the same ware were also discovered (Algaze 1990: 289, 333, Table 90, J,K.). These finds agree with the long existence of a population of Transcaucasian origin in the regions adjacent to the upper flow of the Euphrates.

79

ize also other Kura-Araxes sites, as Geoy Tepe, Pulur (Sakyol), Kvatskhelebi (Fig. 8.12-19 cf. 11) (Sagona 1984: Fig. 122). Pots with a rounded body and a slightly flaring high neck were found in Amiranis Gora, Nakhidrebis Chala, Ghrmakhevistavi, Keti (Fig. 8.5-8) etc. (Chubinishvili 1971: Table XV, 5; Table XVII, 2; Kushnareva & Chubinishvili 1970: Fig. 21, 6; Petrosyan 1989: Table 30, 4; Kushnareva 1993: Fig. 19, 6; Abramishvili et al. 1980: 70, Table V, Fig. 41, 390). In Nakhidrebis Chala and Ghrmakhevistavi the pots were presumably with handles. A similar pot, but with a wider, spherical body and decorated with cord Impression was found in the Ukraine, in the Mikhailovka l settlement (on Pidpilna, a tributary of the lower Dniepr) of the late fourth millennium. This settlement shows affinities on the one hand with the Maikop culture of the North Caucasus and with the Usatovo barrows near Odessa on the other (Gimbutas 1992: 403f).

It must be emphasized that in Tepecik 3 a similar Uruk type red-slipped pot with four handles was also found, together with bevelled rim bowls of Uruk tradition and Early Karaz pottery (Behm-Blancke 1983: 167; Behm-Blancke 1984: 38; Hoh 1984: 72). The Karaz Ware was common at that site during the following Early Bronze period, as well as at Hassek 4, where it obviously represented a part of the spectrum of pottery. According to specialists, the metal of Hassek Höyük came from the area located between Erzurum and the southern coast of the Black Sea (Schmitt-Strecker et al. 1992:122).

Arslantepe VI A, Tepecik 3 and Hassek 5 are thought to be contemporary and, like Kurban Höyük, roughiy coeval with Habuba Kabira-South (8 km downstream from Jebel Aruda). Hence, they must fall somewhere in the middle Hama K levels and the transitional Amuq F/G, revealed at Teil al-Judaidah and Çatal Hüyük (Amuq) (Trentin 1993: 184). Despite the substantial similarity between Arslantepe VI A, Tepecik 3 and Hassek 5, the links between Tepecik and Hassek seem to be stronger than those with Arslantepe, essentially due to their greater affinities with Habuba Kabira and with the south (Frangipane & Palmieri 1987: 298). It is possible that Hassek, Tepecik and Habuba Kabira were important members of a foreign enclave and Arslantepe a local center of power (Trentin 1993: 197). But in spite of the characteristics of the sites mentioned, it seems that the first appearance of the Transcaucasian Kura-Araxes culture to the north, as well as to the south of the Tarsus range, must be dated to the time of the Late Uruk period.

If we take into account the absolute date of the Late Uruk period, placed in the middle of the second half of fourth millennium, the necessity of pushing back the traditional low date of the Central Transcaucasian Kura-Araxes culture will be without doubt. The dates obtained for the Near Eastern layers, characterized by the appearance of Transcaucasian elements, represent a good possibility to date the Kura-Araxes culture of Transcaucasia - the latter being earlier than the Near Eastern sites with material of Kura-Araxes provenance. From the point of view of comparative chronology of regional variants of the Kura-Araxes culture, it must be taken into account that the material discovered in the oldest Kura-Araxes XI level at Pulur (Sakyol), as it was stated above, seems contemporaneous with the middle layers of Amiranis Gora in south-western Central Transcaucasia (Kavtaradze 1983: 89f). At the same time Pulur (Sakyol) XI has dose parallels with Arslantepe VI B as to the forms and incised decorations of pot Stands (Palmieri 1981: 112, Fig. 7,6,8).

As an additional possibility to date the initial penetration of the Kura-Araxes population in the Near East, the evidence of the growing Mesopotamian sea commerce in the Arabian Gulf of the Jamdat Nasr period can be used. This event seems to be caused by changed political circumstances in Eastern Anatolia, Northern Syria, Western Iran and the desertion of the Uruk sites in these areas and as a consequence the passing of the distribution of traded ores and artefacts to local control (Moorey1982:15).

The determination of the chronological position of the Kura-Araxes culture is of a paramount importance for the establishment of a common chronological System for the Ancient World, considering the intermediary area of distribution of this culture; between regions dated by historical chronologies of the Near East, based on the literary sources, and regions dated mainly by the use of geochronological methods.

l cannot agree with the point of view that, before receiving the large series of radiocarbon dates from the Georgian and the adjacent sites of the Kura-Araxes culture, it is premature to consider the reliability of the calibrated ¹⁴C dates for this culture (Munchaev 1994: 17). First of all, the „widely accepted" absolute chronology of the Kura-Araxes culture in the third millennium as well as of the preceeding eneolithic culture in the fifth-fourth millennia and of the subsequent Trialeti culture in the first part of the second millennium B.C. is based mainly on uncalibrated „traditional" radiocarbon dates (Munchaev 1994:16; cf. Kushnareva & Chubinishvili 1963:16f). This fact makes by itself necessary to re-consider the „widely accepted" chronological framework. Also the proposal to re-calculate the ¹⁴C dates by the new period of half-life, which would make dates 200 years older (Munchaev 1994: 16), has any sense from the chronological point of view because of the variations in concentration of radiocarbon with time (cf. Kavtaradze 1983:18f).

Secondly, the statement that the calibration curves and tables based on the dendroscales of the Californian pine have not still received the füll acknowledgement, and that therefore it is better to be refrained from their use (Munchaev 1994: 17), after the publication of the calibration curves based on the joint American and Eu-

80

ropean data (the real witnesses of the simultaneous fluctuation of the content of carbon-14 in the northern hemisphere), must be considered as completely obsolete. The calibration curve officially recommended for the correction of the ¹⁴C dates was published in the journal Radiocarbon, 1993 (Stuiver & Reimer 1993), but also already in 1981, at the symposium in Groningen, the use of the available calibration curves for the preliminary correction of ¹⁴C dates was suggested (Burleigh 1982: 139).

Thirdly, the fact must be taken into account that, as the data of relative chronology for a long time indicated, there was a need to revise the traditional chronological position of the Transcaucasian Kura-Araxes culture even independently from the results of geochronological studies. I mean the dates obtained for those Near Eastern layers which contained the remains of Kura-Araxes provenance (Arslantepe/Malatya, Godin Tepe etc.), the cultural ties pointing at the Late Uruk period as to the time of the initial distribution of the Kura-Araxes culture or the penetration of its bearers in the Near East and the stadial proximity between the Georgian Kura-Araxes and Early Kurgan metalworking (and even of some artefacts) and those of the Near East of the Late Uruk - Early Dynastic periods (Kavtaradze 1983: 85-104, 109-115; id. 1987: 12-15; id. 1992: 46-50; cf. Munchaev 1994: 17). (Uncertainty caused by different approaches to the problems of the chronology of the Palaeometallic Age are in the extreme form reflected in some publications concerning the Caucasian archaeology of this period. E.g. in two volumes of the ,,Archaeology of Georgia" (published recently in Tbilisi) some authors are operating with calibrated ¹⁴C dates, others based themselves on the uncalibrated ones).

The second phase of the Early Bronze Age of Central Transcaucasia comprises the final levels of the Kura-Araxes culture, including the final layers of level B of Kvatskhelebi-Khizanaant Gora, the bulk of the Early Bronze Age material of Sachkhere and the latest burials of Amiranis Gora. In this phase, it is possible also to include the Early Kurgan culture of Central Transcaucasia, in which two groups are distinguishable: The first group, comprising the kurgans (barrows) of the Martqopi/Ulevari and Samgori valleys (east of Tbilisi) and the earliest among the so-called ,,Early Bronze Age kurgans of Trialeti"; the second, chronologically subsequent group, represented by the kurgans of the Bedeni plateau (near Trialeti) and the Alazani valley (in Kakheti, eastern part of East Georgia), as well as by the later kurgans among the early group of Trialeti and later group of Martqopi kurgans with pit graves (Dzhaparidze et al. 1980: 40; Dzhaparidze 1994: 75, 77).

This phase seems to be contemporary with the particularly wide diffusion of the Kura-Araxes culture in the NearEast; thus it should be dated to the first half and the middle of the third millennium. Such a date must find corroboration in the typological parallels of the metal inventory of this phase (Kavtaradze 1983:109-116).

While the pottery of the first group of kurgans is close to the Kura-Araxes culture, the pottery of the second, later group is characterized by the so-called "pearl-like" ornaments which is typical of the Novosvobodnaya (Tsarskaya) stage of the North Caucasian Maikop culture and Early Bronze Age north-east Iranian sites (Tureng Tepe IIIC, Shah Tepe III, Tepe Hissar IIB, Yarim Tepe); two such sherds were found in the Late Chalcolithic levels of Alisar (Central Anatolia) (Kavtaradze 1983: 108n.341).

Radiocarbon dates of the first group are: TB-317, Martqopi kurgan no.3, 2279-2050 cal B.C.; four dates are obtained for the Martqopi kurgan no.4: 2611 -2457 cal B.C. (TB-325), 2035-1934 cal B.C. (LE-2198), 2459-2207 cal B.C. (Bin-291) and 2877-2405 cal B.C. (GX-9252); two dates of Zeiani (near Sagaredzho) kurgan no.1 are quite distinct from each other: 2452-2138 cal B.C. (TB-328) and 3497-3131 cal B.C. (TB-329); cereals from the settlement layers of Berikldeebi (near Kareli) gave 3692-3547 cal B.C. (LE-2197). We must take into account also the dates received for the Uch Tepe kurgans (in the steppe of Mil, Azerbaijan), 3364-2925 cal B.C. (LE-305) and 3930-3356 cal B.C. (LE-300). The dates of North Caucasian Ust-Jegutinski graves of the post-Maikop period are: 2866-2507 cal B.C. (LE-693), 2615-2468 cal B.C. (LE-687) and 2464-2284 cal B.C. (LE-692). We have the following ¹⁴C dates for the second group of the Early Kurgans of Central Transcaucasian Early Bronze Age: Alazani valley kurgans, 2566-2458 cal B.C. (TB-243), 2317-2137 cal B.C. (LJ-3271) and 2875-2500 cal B.C. (UCLA-?); Khramebi (near Gurdzhaani), 2587-2468 cal B.C. (TB-242). The date of the Bedeni kurgan, 1680-1520 cal B.C. (TB-30), seems to be anomalous.

In this period most of the artefacts are the result of a complicated production. Simultaneous casting with subsequent hot and cold forging was employed. The composition of such artefacts contains admixtures - determined by excellent smelting abilities of alloys used (Tavadze et al. 1987: 47). After the exhaustion of copper rich oxide ores, it was necessary to exploit deeper sulfidic deposits, chalcopyrite, which required preliminary roasting for the removal of the sulfureous minerals and the oxidation of ore. This caused a significant reduction of arsenic in the cast metal. Therefore, in the opinion of Georgian archaeometallurgists, the artefacts with high arsenic content were won from oxidized arsenic-copper ores, and after the beginning of the use of sulfide ores, the ,,soft" copper, smelted from sulfureous minerals, needed the admixture of other alloying elements to increase its melting and mechanical characteristics (Abesadze & Bakhtadze 1987: 52f).

81

  Fig. 8: 1 - Hassek Höyük (Hoh 1984: Fig. 12, 4); 2 - Hassek Höyük (ibid.: Fig. 12, 5); 3 - Tepecik (Esin 1979: 61, Fig. 12); 4 -Tepecik (Esin 1982: 74, Fig. 11); 5 - Amiranis Gora (Chubinishvili 1971: Table XVII, 2); 6 - Nakhidrebis Chala (ibid.: Table XV, 5); 7 -Keti, grave 5 (Petrosyan 1989: Table 30, 4); 8 - Amiranis Gora, Level III (Kushnareva & Chubinishvili 1970: Fig. 21, 6); 9 -Kvatskhelebi (Sagona 1984: Fig. 1, 3); 10 - Samshvilde (southern part of Eastern Georgia) (ibid.: Fig. 40, 2); 11 - Geoy Tepe K 1 (Chubinishvili 1971: Table XII, 6); 12 - Kvatskhelebi (ibid.: Fig. 105, 1); 14 - Geoy Tepe K 1 (Chubinishvili 1971: Table XII, 7); 15 - Pulur (Sakyol) (Sagona 1984: Fig. 122, 242); 16 - Pulur (Sakyol) (ibid.: Fig. 122, 243); Geoy Tepe (ibid.: Fig. 122, 244); 18 - Pulur (Sakyol) (ibid.: Fig. 122, 245); 19 - Pulur (Sakyol) (ibid.: 122, 246).

82

The typical artefacts of this period are various types of pins, earrings, tubes, beads (Fig. 6.39,40), bracelets (Fig. 6.27), rings (Fig. 7.10), awls, chisels (Fig. 7.8), sickles, arrowheads, "standards", daggers and spearheads (Figs. 6.37,38; 7.3,6,12,15,16), flat (Fig. 7.9) as well as shaft-hole axes. The chronological value of the pins of this period - with loop- (Fig. 7.7,11,14), toggle- (Fig. 6.30), double-spiral-heads (cf. Sagona 1981, 152-155) (Figs. 6.28,29,31; 7.5) - is rather diffuse. Because of that, only their earliest appearances in the various areas of the Near East and the Eastern Mediterranean region need to be mentioned.

(Helms 1973:116,120, Fig. 10,70/4) and a plaque of the quadruple-spiral shape in Arslantepe VIA (Palmieri 1981: 110, Fig. 3,5).

The pins with T-, crutch- or hammer-shaped heads (Fig. 7.1,2,13) have the most parallels in East European sites (north Pontic and north Mediterranean regions), but, as stated above, weapons and tools are of a higher chronological value, since, because of their greater functional possibilities than those of the ornaments, the need of their improvement was much more important.

The shaft-hole axes with slightly cut butts and wide blades, except from Kulbakebi, Marneuli and Medzhvriskhevi (Fig. 6.15,16,18), were discovered in the Bedeni kurgan no. 2, Martqopi kurgan no. 3, Kvemo Sarali (near Marneuli) kurgan no. 9 and Nadarbazevi (Tetritsqaro region). Among the shaft-hole axes a peculiar type with a long, thin, very curved blade and a narrow tubular shaft-hole blade, found mainly in Sachkhere burials, can be distinguished (Fig. 7.4). The shaft-hole axes from the Martqopi kurgans nos. 4 and 5 (Abesadze & Saradzhishvili 1989: 49, 53, fig. 15-18; Dshaparidse 1995b: 230 (no. 69), 232 (no. 76)) are of an intermediate type between the above mentioned two types.

It has been supposed that the diffusion of the technology of casting shaft-hole axes in Asia Minor and Western Asia in the Middle Bronze Age was connected with the spread of traditions of East European and North Caucasian "pastoralists" (Chernykh 1992: 300). Without excluding such an explanation concerning one particular type of shaft-hole axes, we must at the same time take into consideration the finds of the shaft-hole axes in the Near Eastern Chalcolithic and Early Bronze Age sites. They are known from Susa I, the second half of the fifth millennium, and Tepe Ghabristan, a late fourth millennium North Iranian site (Pernicka 1990: 37, Fig. 10, Table 9, 1). Shaft-hole axes were discovered also in Khazineh, Susiana (Khusistan, south-western Iran) in the layers contemporary with the Early Dynastic period in Mesopotamia (Maxwell-Hyslop 1949: 91, 94f, Table XXXIV, 4) and Central Anatolian Early Bronze Age site Kalinkaya (De Jesus 1980: 147). A casting mould, presumably of a shaft-hole axe, was found in Poliochni ,,Azzuro" (Branigan 1974: 79, 82, Fig. 4 (M89)). It must be also taken into account that Transcaucasian axes usually are considered as weapons which, more than other copper-bronze artefacts of palaeometallic age, reveal exact and close stadial, morphological features common to the whole Transcaucasian region: hence this region can be considered as one and the same metalworking center (Gogadze 1990: 89).

At the same time, it is quite obvious that small workshops which used ores with different chemical compositions sometimes existed in one and the same region, e.g. metal artefacts from Kvemo Kartli, southern part of Eastern Georgia, despite their morphological similarity

 83

contained different chemical elements: the weapons from Kvemo Sarali and earlier items from Ghait-mazi (near Marneuli) - zinc, but artefacts from Nadarbazevi and Bedeni - nickel and antimony. The difference of the contents among bronze artefacts was also detected in Shida Kartli of the latest stage of the Trialeti culture (Abesadze 1974b: 56).

The date of the bayonet-like weapon from Tsartsis Gora (Fig. 6.36) is usually connected with the years of reign of the Akkadian king Manishtushu (23th Century) or the ruler of Elam Pusur Shushinak (22th Century) because of the existence of similar weapons with the inscriptions of their names (Kuftin 1949: 74). We must take into account that similar weapons were already discovered in the Royal Graves of Ur and they have, like the spearhead from Tsartsis Gora, an octalateral section of the foundation of blade (Woolley 1934: 303, Table 227, la (U-7925), Ib (U-7930)). Earlier bayonet-like weapons were also found in Transcaucasia itself: at Kül Tepe (Nakhichevan) (Fig. 6.34) and the Tvlepias-Tskaro cemetery (ca. 200 m north from Kvatskhelebi) (Abibulaev 1982: 161, Table 4, 5; Dzhavakhishvili & Glonti 1962: 43, Table XXXVI) (Fig. 6.35). In the C 1 level of Kvatskhelebi a weapon of the same bipyramidal shape was found, but made from bone (Dzhavakhishvili & Glonti 1962: 28f). Local production of this type of weapon in Transcaucasia seems quite possible.

The sickles known from Khizanaant Gora B (Kushnareva & Chubinishvili 1970: Fig. 42,31) and Amiranis Gora (Fig. 6.22), together with sickles from the other sites of the Kura-Araxes culture (Fig. 6.20,21) (Khanzadyan 1964: 94, Fig. 1,3), have more developed parallels, e.g. from Tarsus of the second level of Early Bronze Age (Goldman 1956: 281). Curved points, possibly from sickles, were found in Tarsus in the first level of Early Bronze Age (Goldman 1956: 281).

For the manufacture of jewellery all the existing methods were already used - casting, forging, embossing, plating etc. It has been suggested that since the decorations applied to the jewellery are repeated on the huge pots and other objects of household, they must be products of local craftsmen (Pitskhelauri 1987: 24) (e.g., cf., Fig. 9.13 and 23).

Two daggers from Sachkhere (from Tsartsis Gora and Koreti) with decorated hilts (Fig. 6.45,46) must be considered as the earliest examples of the artistic metalworking. Their blades were forged, but the hilts were cast. The dagger from Tsartsis Gora has a flat hilt, decorated with the broken line and wattled relief patterns. The other dagger from Koreti has also similar wattled ornamentation, but in the form of spiralic tendrils. This decorations of the Sachkhere daggers were obtained by casting in wax mould, a sign of the high Standard of metalworking in this period (Abesadze et al. 1958: 22).

On the diadem from the latest burials of uppermost Kvatskhelebi B period the representations of birds, deer and astral signs were made by punching on the thin plate of copper (Fig. 6.41). The closest analogues to this diadem are a golden diadem from the Ur Royal Graves and a silver one from Chalandriani (Kavtaradze 1983: 115).

The Early Kurgan period in Central Transcaucasia is marked by the introduction of tin bronze. However, the chemical-technological analyses of the two stages of this period show a considerable difference between them. The bronze artefacts of the first, so-called „Martqopi-Ulevari stage", except some artefacts of pure copper, are mainly manufactured of arsenic-antimony bronzes with a high content of alloying elements. The admixture of arsenic in the artefacts from Martqopi kurgans is from 1 to 4.6%, lead 0.001-0.5%, iron 0.001-0.6%, zinc 0.001-0.32%, antimony 0.001-0.4%, silver 0.001-0.03%, nickel 0.001-0.05% (Abesadze & Saradzhisvili 1989: 54, 59f). Only in two cases tin bronzes were found there with 3.2 and 6 % of tin with the addition of zinc and lead. (Both artefacts, the curl-ring and the standard, were found in the Martqopi kurgan no. 3 (Inanishvili 1989:127, note 25; cf. Abesadze & Saradzhishvili 1989: 54). Another standard with 7.25% Sn is from the Martqopi kurgan no. 5. The kurgans nos. 3 and 5 are the latest among the Martqopi kurgans and must be factually synchronous with the Bedeni stage (Dzhaparidze 1994: 77; Schillinger 1997: 27). At the same time, in A. Schillinger's opinion the earliest tin bronze artefact of Transcaucasia is the quatrilateral awl detected in Telebi (on the right bank of the Alazani, near Telavi), the East Georgian site of the late Kura-Araxes times, and it contained already 11.3% Sn (Schillinger

84

1997: 24f). On the other hand, in the artefacts of the second, Bedeni-Alazani stage, the content of tin is from 8 to 15%. At that time the metal inventory is represented mainly by tin bronze (Inanishvili 1989: 127, note 25. Four artefacts from Bakurtsikhe (Eastern Georgia, near Gurdzhaani) kurgans: two spearheads, dagger and spike contained 10-13.7% Sn (Abesadze & Saradzhishvili 1989: 55)).

In the subsequent, so-called Middle Bronze Age Trialeti culture of the „brilliant barrows", tin bronze also prevails with the same percentage of tin content (Pitskelauri 1987: 24) as during the second stage of the Early Kurgan period, its amount in most cases fluctuates between 4.30 and 14%. Only few items are without any trace of tin; two of them are kettles, one with 9.85% of arsenic and another with 1.5%. In this period, except other elements mainly lead (0.2-0.7%), silver (0.05-0.15%), arsenic (0.2-0.5%), iron (till 0.12%) and nickel (till 0.03%) were used. Zinc was not detected (Abesadze 1974b: 50).

According to the conclusion of Georgian archaeometallurgists, the artefacts of tin bronze were made by smelting the local copper ores with the addition of im-ported tin. At the same time, in certain cases the import of separate metal objects was supposed. It is interesting that in the western part of Georgia twenty deposits with tin content were discovered by geologists, but unfortunately without any trace of ancient exploitation. Georgian specialists emphasize the fact that tin is genetically connected with granitoid intrusions, characteristic of the Caucasus (Abesadze 1980: 148-156; Abesadze & Bakhtadze 1987: 54; Abesadze & Saradzhishvili 1989: 58).

Taking the dates of the Early Kurgan culture of Transcaucasia into consideration it appears possible to place the beginning and initial stages of the Trialeti culture of the „brilliant barrows" in the second half of the third millennium. Even from the point of view of stadiality, the culture of the „brilliant barrows" of Trialeti is essentially a typical product of the Early Bronze Age of the Near East and its periphery. Numerous parallels can be found in the materials dating back to the third millennium for the early materials of these kurgans, noted already by B. Kurtin (cf. Kuftin 1941; Gogadze 1972; Kavtaradze 1983).

The typical trait of the Trialeti Middle Bronze Age kurgans is their poverty of weapons. There were only isolated spearheads, rapiers, daggers and knives discovered, among them a silver dagger. It must be taken into account that there is no difference between weapons and decorations as to their chemical composition (Abesadze 1974b: 50f). This fact can be considered as an evidence of a simultaneous production of both types of artefacts.

The Trialeti kurgans are very rich in gold. An exceptionally high Standard of the goldsmith is witnessed by a well-known golden cup with curls and friezes outlined in the double filigree and set with turquoises and carnelianes (Fig. 9.13) and by pierced silver pins with golden heads set with the same precious stones and granulations (Fig. 9.8,9). The central piece of agate mounted in gold (Fig. 9.11) of the big golden beads is similar to the jewellery of the Ur III period from Uruk (Kuftin 1941: 94, Fig. 98; Maxwell-HysIop 1971: 75), but by its shape more similar to the agate pendant from the Ur burials of the Akkad period (cf. Woolley 1934: 371 f, Fig. 79; Kavtaradze 1981: 104f., Table VII). The silver cup from the Trialeti barrow no. 5 decorated with mythological scenes (Fig. 9.1) has a recently discovered analogue in the Karashamb kurgan of Armenia (on the bank of the Razdan, in the Ararat valley). Both these cups are dose to the silver and golden ones of Troy llg by their shape (cf. Schliemann 1885: 586, 594, no. 840, 841, 858).

From Irganchai, in the western part of Kvemo Kartli, near Dmanisi, recently series of ¹⁴C dates for the various periods of the Trialeti culture were received; for the later stage: kurgan no. 1 of Irganchai, 2122-1910 cal B.C. (TB-475), kurgan no. 2, 1932-1749 cal B.C. (TB-476), kurgan no. 3, 2284-1984 cal B.C. (TB-545), kurgan no. 4, 1886-1706 cal B.C. (TB-477), kurgan no. 5,1512-1406 cal B.C. (TB-478), kurgan no. 9,1872-1679 cal B.C. (TB-496), kurgan no. 18, 1678-1208 cal B.C. (TB-548); for an earlier stage, the Middle Bronze Age material of which reveals some traits typical even of the Bedeni culture: kurgan no. 21, 2132-1951 cal B.C. (TB-546), kurgan no. 25, 2460-2138 cal B.C. (TB-811), kurgan no. 26, 2200-1934 cal B.C. (TB-812), kurgan no. 27, 2856-2409 cal B.C. (TB-817), kurgan no. 28, 2578-2285 cal B.C. (TB-818), kurgan no. 30, 2582-2328 cal B.C. (TB-835), kurgan no. 32, 2033-1749 cal B.C. (TB-834), kurgan no. 37, 3336-3036 cal B.C. (KN-4499). Two dates gained from the II kurgan of Aruch (Armenia), typical of the later stage of Trialeti culture: 2112-1772 cal B.C. (Bln-2727) and 2032-1890 cal B.C. (Bln-2801). The dates obtained for layer II of Dikha Gudzuba in Anaklia, in Western Georgia, which is dated more or less contemporary with the Trialeti culture, are: 2273-2044 cal B.C. (TB-274), 2391-2146 cal B.C. (TB-275) and 2117-1930 cal B.C. CTB-276).

The necessity of the pushing back the Transcaucasian dates was also recently demonstrated by the finds of the kurgan of Karashamb. This unique complex (with the copious golden, silver and bronze artefacts) of the second group of the kurgans of the Trialeti culture has some traits characteristic of the third dynasty of Ur (21th-20th centuries B.C.), but at the same time it reveals connections with the earlier Central Anatolian culture of the „Royal Tombs" of Alaca Höyük (Golovina 1990: 230; Oganesian 1992: 84, 100 note 1).

At the time of the later part of Trialeti culture, mainly in Western Transcaucasia, a new type of alloy, namely arsenic-antimony bronze appeared. If tools and weapons were made of firmer and better forgeable arsenic and

86

tin bronzes, ceremonial and cult inventory were cast from less firm, although easily handled arsenic-antimony bronzes which at the same time had a silver-like appearance. Often beads of antimony were found. Casting with the wax mould was widely used. Such a development of metallurgy was possible because of the existence of local resources.

On the upper stream of the Rioni (the ancient Phasis, the main river of Western Georgia) copper and arsenic-antimony deposits (in Zopkhito, Sagebi, Kvardzakheti etc.) are known. At the same time, there (near the village of Ghebi, in Racha) were found traces of ancient exploitation - nearly 100 copper and 30 antimony mining places with waste heaps of more than 100 000 tons (Mudzhiri et al. 1987: 235f; Kushnareva 1993: 245).

Apparently this is the reason why in Georgia (mostly in the western part) and the Northern Caucasus arsenic-antimony bronzes were used so widely in the Middle and Late Bronze Ages. Radiocarbon dates of the antimony mines of Racha are: Zopkhito, 1379-1167 cal B.C. (TB-335) and 1517-1083 cal B.C. (TB-302); Sagebi, 1895-1748 cat B.C. (TB-310) and 1882-1698 cal B.C. (TB-334). In the opinion of specialists they were exploited from the beginning of the second millennium B.C. (cf. Kushnareva 1993: 245). The traces of ore melting workshops and ore-threshing stone hammers found there were dated to the beginning of the Bronze Age (Abesadze 1980: 27). In Ghebi copper ores have 8% of copper. There specimens of antimony ore from a mining place situated 6 km from Ghebi contained following elements: I, antimony - 27.26%, lead - 0.66%, arsenic -4.64%, copper - 0.11%, iron - 2.08%, sulphur -12.37%; II, antimony - 40.47%, arsenic - 0.56%, lead -0.87%, iron -1.90%, sulphur - ; III, antimony - 27.31 %, arsenic - 1.41 %, lead - 0.87%, iron - 5.66%, sulphur -15.83% (Abesadze 1980: 26).

By taking into account the fact of the existence of high antimony copper objects from the Early Bronze Age materials of Sachkhere (south of Racha), which are sometimes also characterized by a high content of arsenic (Abesadze 1969: Table III-V, NN 76-117; Kushnareva 1993: 234) (Table I), it seems to be possible to date the initial exploitation of such ores already to the early third millennium B.C. For the dating of the earliest materials from Sachkhere, importance should be given to the abovementioned spearhead found in Tsartsis Gora, as well as to similar spearheads discovered in the Late Chalcolithic VI A level of Arslantepe-Malatya.

An awl from Ozni (Kvemo Kartli) containing 2.7 % of antimony and a curl-ring from Kvatskhelebi B with 5% of antimony are dated to a rather early time (Abesadze 1969: 99,101). At the same time, while materials of Dzagina and Kvatskhelebi (Shida Kartli) has a high content of antimony in the artefacts from Koreti, of the site from the outskirts of Sachkhere, no traces of antimony were detected (Abesadze et al. 1958: 19; Abesadze 1969: 102,104).

In North-Western Georgia, Abkhazia, the watershed mountain of the ravines of Kodori and Bzyp, 20 ancient copper mining places were discovered (Abesadze 1980: 35). We have following ¹⁴C dates from mine no. 4 of Bashkapsaara: eastern part, 3015-2148 cat B.C. (LE-4198), northern part, 1518-994 cal B.C. (LE-4197), western part, 1750-1318 cal B.C. (LE-4196), central part, 2906-2137 cal B.C. (LE-4199) (Kushnareva 1993: 244, 280).

In the territory between Abkhazia and Racha, in mountanous Svaneti, in Zaargash, on the upper flow of the Enguri, a polymetallic deposit was discovered which was exploited at the times contemporary with the mines of Racha (Chartolani 1988; Kushnareva 1993: 245). Svaneti is exeptionally rich in lead-zinc and arsenic ores (Abesadze 1980: 28).

In the Middle Bronze Age the cultures of Eastern and Western Georgia (Central and Western Transcaucasia) had different metallurgical sources, but it seems that they had strong ties. As the consequence of such an interrelationship, tin was spread to Western Georgia from the Trialeti culture, but antimony from the western part of Georgia to Eastern Georgia (Abesadze 1980: 24). The presence of zinc and lead, beside antimony, is typical of the contents of West Georgian bronze artefacts. There, at the same time, zinc (till 1.25 %) is detected in arsenic-antimony bronzes of Abkhazia and lead (1.8-6.3 %) in tin bronze artefacts of Racha and also sometimes in the Trialeti culture (1-3.5 %, in one case - 8.38 %). Because zinc and lead were already found in Kura-Araxes artefacts of Georgia (1.2-2.5 %), it was supposed that they must represent natural admixtures, typical of the copper ores of Georgia (Abesadze 1980: 24).

For the dating of the common Transcaucasian Middle Bronze Age certain importance can be given to the obsidian of south Transcaucasian provenance revealed in Tal-i-Malyan in the Iranian province of Fars in deposits of the Kafteri phase (2100-1800 B.C.) and determined by the Conservation Analytical Laboratory of the Smithsonian Institution. If one part of them was similar to the obsidian used in Alikemektepesi (Azerbaijan), another part, coming from the Gutansar complex of Armenia (western slope of Gegam), was found in great quantity in the sites of the Ararat valley, i.e. south of its "birthplace" - the Gegam mountain. At the same time, in the eight kurgans of the Karmirberd culture, necklaces were found dated to the time of the Babylonian king Samsu-iluna, 1806-1778 B.C. Among the necklaces some consisted of shells of sea molluscs which were obtained either at the estuary of the Persian Gulf or on the south Iranian coast (cf. Simonyan 1984).

The discovery of the south Transcaucasian obsidian in the southern Iran and of south Iranian or south

 87

Mesopotamian ornaments in Transcaucasia can be considered as the reflection of one and the same phenomenon - the existence of trade connections between southern Transcaucasia on the one hand and south-western Iran and southern Mesopotamia on the other which determines the coexistence of the late Karmirberd and early Sevan-Userlik cultures of southern Transcaucasia and of the final part of the Trialeti culture in the 18th Century B.C. (Kavtaradze 1992: 51f;cf. Kushnareva 1994: 117).

The upper date of the Trialeti culture, and together with it the Middle Bronze Age of Central Transcaucasia, was assigned to the middle of the fifteenth Century B.C. on the basis of the date of the shaft-hole spearhead with a ferrule at the end of the kurgan no. 15 of Trialeti (Scha-effer 1948: 512) (Fig. 9.15). But since in the Near East shaft-hole spearheads seem to have come from the Early Dynastie period (Thomas 1967: 73) and the spec-imens with ferrule appear from the end of the third millennium there must be a reason to put into doubt the correctness of the above date for the Trialeti spearhead and to shift it back, together with the end of the Trialeti culture, to the middle of the first half of the second millennium (cf. Kavtaradze 1983: 130-134). The same can be stated concerning the Aegean parallels of the seventeenth century B.C. to the silver bowl with a cotton-reel handle from the Vanadzor (Kirovakan) kurgan of the Trialeti culture because of the existence of a similar bowl in Upper Egypt, among the treasure of Tôd, dated back to the time of Amenemhet II (1929-1892 B.C.), Pharaoh of the XII Dynasty, and the findings of similar cotton-reel handles in the Anatolian sites of the end of the third millennium B.C. (Vandier 1937: 174; Maxwell-Hysiop 1995: 243-245, 250).

The latest group of Trialeti barrows revealing also some traits peculiar to the Late Bronze Age, together with other sites contemporary with them, can be united in the latest part of the Middle Bronze Age and dated by the post-Trialeti times - approximately in the middle of the second millennium B.C. Up to now the only ¹⁴C date of this period is known from the Metekhi burial (near Kaspi): 1526-1426 cal B.C. (TB-31).

The dating of the Transcaucasian metal artefacts and complexes containing them is in many cases possible by the consideration of the dates of materials from well-dated Near Eastern strata. The chronological conclusions received by this way, that is by correlation with the data of other archaeological materials and geochronological analyses, represent the decisive factor for the formation of relative and absolute chronologies of Central Transcaucasia of the Palaeometallic Age.

l am very much indebted to G. Burger, E. Schalk and E. Pernicka for reading the text and giving usefui advices, as well as to K. Kakhiani and A. Paghava for the possibility to publish ¹⁴C dates of Irganchai. Discussions with E. Pernicka, R. Gläser and D. Sürenhagen have been of highiy important value. l would also like to thank B.B. Helwing and A. Schillinger for allowing me to cite their theses. l want to express my gratitude to the organisers of the very interesting Conference in Bochum and to the Volkswagen-Foundation which made my participation possible.

88

89

90