Version classiqueVersion mobile

On salt, copper and gold

 | 
Catherine Marro
, 
Thomas Stöllner

Early mining in the Caucasus and beyond

The Paravani calculation

The early gold‑mining complex in Sakdrisi and its societal impact

Thomas Stöllner, Irina Gambashidze, Nicole Boenke et Werner H. Schoch

Résumé

Lorsque le complexe de Sakdrisi a été découvert il y a plus de douze ans, et que son importance est apparue, personne ne se rendait compte à l’époque de l’importante moisson d’informations que livrerait son étude, que ce soit sur le terrain ou en laboratoire. On admet désormais que les industries minières, la métallurgie, le site d’habitat et les structures sociales sont des éléments contemporains liés les uns aux autres : ceci ouvre diverses perspectives permettant de mieux saisir la société minière de la fin du IVe millénaire avant notre ère.
Cet article se propose de décrire de façon succincte la chaîne de production kuro-araxe dans son entièreté, en commençant par l’extraction et la préparation du minerai, sa mise en poudre et enfin l’opération de réduction. Ces données précieuses nous ont aussi permis de comprendre l’importance revêtue jadis par la production de l’or dans la vie quotidienne et son influence sur les rituels et le système social. Une des questions posées est de savoir si la production d’or était à l’époque une activité permanente ou saisonnière, notamment parce qu’une ambiguïté demeure en ce qui concerne le mode de vie des groupes impliqués dans la production de l’or à l’origine. Ces groupes étaient-il mobiles ou sédentaires ? Cette question doit être approchée de façon quantitative et non seulement qualitative : quelle force de travail était nécessaire pour produire un gramme d’or depuis les riches dépôts aurifères de Sakdrisi ? Notre équipe a abordé cette question à travers une série d’expérimentations qui lui ont permis d’établir un calcul quantitatif détaillé des étapes de la production (le « calcul de Paravani », nommé ainsi d’après un site où l’on a découvert une boucle d’oreille en or). Ces données quantitatives seront utilisées pour modéliser les différentes étapes du travail effectué à Sakdrisi et Dzedzebi et nous aider à comprendre les conséquences qu’elles eurent jadis.

Note de l’éditeur

In collaboration with : Antoine Courcier [contributor] Tobias Skowronek [contributor]

Remembering Antonio Sagona († 29th June 2017)

Texte intégral

I am grateful for fifteen years of mutual collaboration within the research team, especially with my co-director Dr Irina Gambashidze (Tbilisi). The research at Sakdrisi-Dzedzvebi is financially supported by the Volkswagen Foundation (Hannover) and the DFG (since 2013); Dr Levermann (Hannover), Dr Bienert and Dr Kümmel (Bonn) are thanked for their continual interest in our research. For this article I have to thank especially A. Hornschuch; S. Senczek, MA; F. Klein, MA (Bochum); Dipl. Eng. T. Sachvadze and K. Tamashashvili, MA (Tblisi). I also thank Roselyn Campbell, Los Angeles, USA, for her corrections of my English. Some aspects of this article are already published: Stöllner 2016, 2017; Stöllner, Gambashidze 2014.

Introduction

1The rise of inequality in early societies is a matter of long-lasting debates within research. From the beginning of the 5th millennium on we see many cases of social inequality from the very western shores of Eurasia to the eastern parts. In all of these case samples we learn about ritual complexity that configures social inequality. In the Caucasus such conditions can be studied when comparing the Maikop phenomenon in the 4th millennium (Munchaev 1975; Korenevskij 2004; Govedarica 2002) with Transcaucasian late Chalcolithic (Leila Tepe/Berikldeebi culture; Javakhishvili 1998; Akhundov 2004), different production spheres and even social complexity displayed in settlements and some early kurgans (Soyuq Bulaq: Lyonnet et al. 2008; Boyuk Kesik N4: Akhundov, Makhmudova 2008; Kavtiskhevi: Makharadze 2007). This Late Chalcolithic ritual and societal complexity seems to be followed by more standardized ritual displays during the Kura-Araxes phase from the second quarter and the mid of the 4th millennium onwards (Sagona 1984; Kushnareva 1997; Palumbi 2007; Kohl 2009). This societal rules and orders changed anew in its latest phase when the lavishly arranged tumuli of the Martkopi-Bedeni stage occurred (fig. 1).

Fig. 1 – Rise and decline of the ritual display of social inequality in Transcaucasia (DBM, RUB, T. Stöllner [graphics]; Lyonnet et al. 2010; Gambashidze et al. 2010; Gambashidze et al. 2001 [O. Lordkipanidze]; T. Rabsilber, DBM [photos/drawings]).

Fig. 1 – Rise and decline of the ritual display of social inequality in Transcaucasia (DBM, RUB, T. Stöllner [graphics]; Lyonnet et al. 2010; Gambashidze et al. 2010; Gambashidze et al. 2001 [O. Lordkipanidze]; T. Rabsilber, DBM [photos/drawings]).

2These phenomena show that the display and construction of social complexity could change over time more or less rapidly. The reasons are as manifold as the societal formation and its material shapes are. Social structures can appear in more firm structures such as clear rules of participation or in more imbalanced dispersal of wealth distribution and social expression. Objects and especially goods that represent such a societal construction take an important role in such a manifestation. Therein the consumption of metals and prestigious goods plays a prominent role (metals, gold, and jadeite axes also; Hansen 2013). But seldom we interwove the role of production of these goods and their producers, nor are the complexity of work steps considered. It can simply be asked if complex social and economic practices as mining enterprises required or even favored the development of hierarchical organizations. This often was claimed; simplified models even strengthen the idea of some kind of dependency and dichotomy between “ruling elites” and a producing “under” (Gilman 1996). But mining as a social practice is more complicated and might have inherited a social complexity right from the very beginning (Stöllner 2016): expeditions to procure special or prestigious materials reveal examples of a complex interrelation between social role and personal expert knowledge.

  • 1 Grain mills are still missing as a category of artefact at the crafts-shop part of Dzedzvebi settl (...)

3The Sakdrisi-Dzedzvebi prehistoric gold-mining and production complex belongs to the most important examples of prehistoric gold mining in the world. In a total of ten excavation campaigns (2004‑2015), the gold mine and an associated gold district were further investigated. After a 12‑year field project, the German-Georgian research team has collected a plethora of data about many aspects of the Kura-Araxes period gold production and its technology and also about its economic and social impact (Gambashidze et al. 2010; Gambashidze, Stöllner 2016; Hauptmann et al. 2010; Stöllner et al. 2010; Stöllner et al. 2014; Stöllner, Gambashidze 2011; Stöllner, Gambashidze 2014; Stöllner 2016; Stöllner 2017). The project also included extensive surveys in the area around the gold mine with the aim of investigating more closely the question of an associated settlement landscape (fig. 2). We discovered an associated settlement together with contemporary groups of burials and were able to link these directly with the gold extraction. Evidently, a large part of the time-consuming labour of processing the gold ore, that is, the arduous milling of the gold ore, was carried out in the settlement. New discoveries also suggest that the groups of craftspeople engaged in working the gold were largely provided from outside and did not pursue any agrarian activities.1 Gold mining and the processing of gold ore, including the gold metallurgy that goes with it, indicate a clear division of the technical processes and also a society that practised division of labour (see below). The gold production lasted over a time span of about 600 years between roughly the late 4th and the early 3rd millennium BC (Stöllner et al. 2014, pp. 79‑81). It was a time during which the Kura-Araxes phenomenon spread to large parts of the near eastern regions and became a driving force within the cultural and historical development between the northern Levant and northern Iranian plateau as well as Transcaucasia and eastern Anatolia (e.g. general aspects: Munchaev 1975; Sagona 1984; Kushnareva 1997; Kohl 2009; Batiuk 2013; recently various articles in Chataigner, Palumbi 2014). And it was also a time during which metallurgy and use of metals were appropriated to a larger extent by these communities (e.g. recently Japaridze 2013; Stöllner, Gambashidze 2014).

Fig. 2 – The plateau of Balitchi-Dzedzvebi and its Kura-Araxes settlements, workshops and graves in relation to the gold-mining site at the Kachagiani hillock/Sakdrisi (ArcTron/DBM, F. Klein).

Fig. 2 – The plateau of Balitchi-Dzedzvebi and its Kura-Araxes settlements, workshops and graves in relation to the gold-mining site at the Kachagiani hillock/Sakdrisi (ArcTron/DBM, F. Klein).

The Sakdrisi-Dzezdvebi gold-production complex: new insights into the chaîne opératoire of the gold-mining process

  • 2 The single operation steps are currently studied by various colleagues: While F. Klein (MA) is inv (...)

4The reconstruction of a chaîne opératoire is an ongoing process requiring a transdisciplinary approach of material sciences, archaeology, ethnology and experimental archaeology. The process reconstructed so far can be compared with techniques described for antique and younger periods (Agricola 1556; Tylecote 1987; Bachmann 1999). Despite this empirical approach, many questions remain open that pertain to technical solutions and quantitative aspects such as the time span and the yield of different production steps. Together with the estimate of the gold once produced, these questions are essential to understand the social impact that the gold exploitation once had (see in detail Stöllner et al. 2014; Stöllner 2016). The ore production and the gold production consisted of several steps that had been documented by archaeological observations, by artefacts and by features (Stöllner, Gambashidze 2011, p. 196; Stöllner et al. 2012; Stöllner et al. 2014). Besides those observations the research team has carried out studies in experimental archaeology (Stöllner et al. 2012); a detailed study of the tool inventory has made it possible to demonstrate the individual steps of the extraction process right through to the washing of the ore and the associated smelting in crucibles.2 Important new data and observations especially concerning milling, smelting and the panning process will be presented here.

5The operation stages can be summarized as follows. The mining of the gold-bearing quartz-haematite veins took place above and below ground in Sakdrisi. Fire-setting and further accompanying techniques, such as hammering and wedge work, were the most important techniques in advancing the drifts down to a depth of nearly 30 m (Stöllner et al. 2014, esp. pp. 75‑89). The galleries remained nearly as narrow as the spatial extension of hardened and siliceous veins and were completely emptied by the mining work at first. Later, parts of the mines had been refilled by the mining with mining debris and partly also by crushing debris from the above-ground beneficiation work. Our excavations revealed the careful efforts in refilling parts of the gallery with such “debris”. The first stages of beneficiation and sorting the country rock from the yield were done below ground: crushing dimples at the rock surface indicate crushing and sorting work, perhaps also in order to minimize the size of gangue and ore pieces. As the miners could not decide beforehand whether the ore contained high or low amounts of gold, they might have had to transport all the prepared pieces to the surface. The next steps of beneficiation took place there, during which ores were crushed to a coarser grain size of about 0.5 to 0.1 cm. This work took place at the edges alongside all the opencast entries of the mines – proven by dimpled stone and dimpled rock faces discovered there. The ores might have been tested at a milling and washing station near the entry of mine A, where a cistern and milling tools were found exclusively in a specific tailing. As milling was not carried out to a large extent, it had been assumed that it took place predominantly in the settlement site of Dzedzvebi nearby. Special workshops were discovered there, in which the grinding work was carried out on a professional level (Stöllner et al. 2014, pp. 97‑99; Gambashidze, Stöllner 2016, pp. 71‑77). Such grinding activities could also be evidenced, besides in workshop houses, on open-space areas, obviously where grinding was carried out rather sporadically or temporally to a limited extent. The recently excavated site III.5 has provided ample evidence for grinding and smelting (fig. 3).

Fig. 3 – Dzedzvebi, III.5, map of the gold-grinding and smelting site, sections through the working and smelting pits (DBM, I. Luther, F. Klein).

Fig. 3 – Dzedzvebi, III.5, map of the gold-grinding and smelting site, sections through the working and smelting pits (DBM, I. Luther, F. Klein).

6Smelting activities also included the smelting and manufacturing of pure copper and arsenic copper, perhaps at or near site III.5 but certainly in the metal workshop in houses 1‑2 at site II.2 (see below). But it has to be stressed that gold-smelting crucibles are still difficult to identify for certain: even the position of a crucible bottom in situ in a probable smelting pit at site III.5 was not unambiguous.

  • 3 Pers. comm. from Dr I. Gambashidze and Dr G. Mindiashvili.

7One unsolved question concerning the chaîne opératoire pertained to the matter of the wet mechanical concentration of mineral concentrates to be smelted in the next step. As the experiments of 2011 and 2013 (Stöllner et al. 2012) have shown impressively, it was necessary to select the heavier magnetite minerals from the lighter gold flakes at the end of the panning process. This would have certainly needed gold-panning installations near the grinding areas: one of these panning installations appears to have been present in house 4, which was excavated in 2015 (fig. 4). The pit 38046 was discovered south-east of a deposition of millstones and rocks. The feature consisted of a pit that was covered by a lightly burnt clay basin with an outer elevated rim of 5 to 20 cm in width, while the inner, deeper basin had a diameter of 25 cm. What made the feature outstanding was a ceramic pan-shaped bowl discovered inside the inner basin (fig. 5) and the deposition of a smaller storage vessel on the western edge of the basin. The flat ceramic bowl has an indented rim with a knob. This depicts most likely a handle that might have been used for washing activities. Vessels of this kind have been discovered only seldom in Kura-Araxes contents3 but whether those vessels also resemble a possible connection with gold panning remains open. However, the distinct gold levels found in the sediments inside this feature provide clear proof of this activity. It may therefore be asked if the accompanying vessel originally contained mill concentrates. There is a likely but indirect connection to the milling process within the chaîne opératoire of gold processing: a millstone deposition was found nearby and can be related to the activities once carried out within this house. The deposition itself certainly was laid down at the end of the activities during ritual practices.

Fig. 4 – Dzedzvebi, area II.8, house 4, map of upper level with ceramic/stone deposits and of lower level with gold-panning installation, central pit and pigment deposit (GNM; DBM/RUB, F. Klein).

Fig. 4 – Dzedzvebi, area II.8, house 4, map of upper level with ceramic/stone deposits and of lower level with gold-panning installation, central pit and pigment deposit (GNM; DBM/RUB, F. Klein).

Fig. 5 – Dzedzvebi, area II.8, house 4; left: gold-panning installation; right: ceramic pan from gold-panning installation (GNM; DBM/RUB, H.-J. Laufer, T. Sachvadze, T. Stöllner).

Fig. 5 – Dzedzvebi, area II.8, house 4; left: gold-panning installation; right: ceramic pan from gold-panning installation (GNM; DBM/RUB, H.-J. Laufer, T. Sachvadze, T. Stöllner).

New evidence on the gold processing: gold content and possible gold crucibles at the Dzedzvebi settlement area

8Ore grinding as well as metallurgical activities can be regarded as a central craft practice in some areas of the Dzedzvebi settlement (Stöllner et al. 2014): all of the Kura-Araxes houses so far excavated delivered indications of such a practice (houses 1‑4: II.2‑3 and II.8). The same is true of a craft site discovered in 2014 in area III.5. As this area has exclusively delivered grinding stones and a pit filled with ashes and crucible pieces (see above), gold production was a probable explanation. Evidencing gold production is difficult, as the crucible smelting of gold concentrates does not produce slag but sometimes some smelting dross or tiny splashes of gold (Rehren, Temme 1994; pers. comm. from A. Hauptmann). As a rule, gold prills are not found and therefore it is difficult to identify the precious metal’s smelting crucibles (a rare exception for instance is Inzersdorf-Walpersdorf: Ramsl 1998, pp. 42‑43). During the recent years of research on the Dzedzvebi plateau, soil sediments have been investigated from features that could be regarded as related to debris of gold grinding (tab. 1). The soils were tested for their gold content with the help of standard wet chemistry, a method that has proven to be successful in detecting elevated gold contents. These contents were significantly higher than in soil in the surrounding area that basically derived from erosion of the basalt of the Dzedzvebi basement rock.

Tab. 1 – Gold content of various soils analysed (data after DBM, laboratory, T. Skowronek).

DBM-Lab. no. Find no. Area ppm Au
3362_16 25549 II.8 6.5
4169_16 18205 III.5 3.3
4170_16 18204 III.5 12
  Dz.II/3 H-2 II.3 0.61
  Dz.II/3 H-1 II.3 0.63

9There were significant connections with archaeological contexts, such as the soil sample that was harvested in the surroundings of a ceramic pan found in feature 38046. The gold content average was 6.5 ppm of gold, which is not only significantly elevated but also at a level that could be regarded as debris of an originally much higher gold content (for gold content in the Sakdrisi mine, see Stöllner et al. 2014, pp. 89‑91, 105‑106). The average gold content is less high in soils sampled in house 3 of site Dzedzvebi II.3: in contrast to the specific features in Dzedzvebi II.8 and III.5, the soil tested was rather unspecific debris found inside the grinding house. The sampling area was not related to any specific work installation, which may explain the rather low gold enrichments.

10The opposite is true for the features of area III.5 (fig. 3): two layers have been found below a deposit of grinding rubbers, which points to the extraction of gold-containing educts. It is interesting to note that the darker layer 33038 (find 18204) has delivered a four times higher gold elevation than the lighter layer 33039 (find 18205) (tab. 1). Regarding the high relevance of the data, 21 blank values of the beakers used for analysis have been investigated. But all their gold contents were below 1 ppm, so a contamination by the sampling process could be excluded. Thus, there is no reason to doubt the gold enrichment and its obvious interpretation as debris from gold grinding.

11As a pit with crucibles found inside was discovered nearby, it has to be asked if these crucibles can also be related to gold processing. Two crucibles were investigated by A. Courcier at the Bochum laboratory: the sample 4034/15 (finding 18209, feature 33046) was found inside a pit filled with an ashy layer. The pit was situated at the northern edge of the grinding area. A first cross section was realized in the bottom of this technical ceramic. The microscopic study reveals no highly heated part on its surface. The ceramic, heated in oxidized conditions, is composed of finely ground minerals (quartz, diopside, calcite, cristobalite, albite, and microcline), vegetal temper and grog. Two other cross sections were carried out on the edges. But they unfortunately did not permit proving a metallurgical use. The SEM/EDS analyses were also in concurrence with these results. There is a different result for sample 4036/15 (finding 18185) (fig. 6.1‑2), discovered not in situ in the lighter cultural layer 33039. The ceramic of this smelting crucible was heated in reduction conditions and composed of a few vegetal temper fine grinding minerals and one large black mineral inclusion of anorthite/augite. A small fragment of quartz with a gold inclusion was detected by SEM/EDS in the ceramic, so it is likely that the crucible was manufactured near a gold grinding area where some gold-bearing quartz had been mixed with the clay used. On the surface, thin dark grey slaggy dross is visible (0.2 cm thick). The fully liquefied bullous slag is composed of two phases: magnetite clusters or needles precipitated into siliceous matrix, and amorphous vitrified phase. The slag contains also small fragments of gangue mineral (quartz). Fine dispersed precipitation of cuprite and small malachite fragments is observed in this second phase. Numerous tiny (1‑4 μm) and few small metallic prills (40‑70 μm) were trapped in the slag (fig. 6.1). The SEM/EDS analyses permit determination of two types of prills: unalloyed copper and arsenical copper (1‑8% As) (fig. 6.3). According to the very small size of the splashes and their dispersal within the smelting dross, it seems questionable at first hand if arsenical copper was the desired product.

Fig. 6 – 1: Dzedzvebi III.5, 18185, crucible fragment, photomicrograph of slag, sample no. 4036/15, taken under plane polarized light, magnification×2.5, the slag contains magnetite clusters precipitated into undetermined siliceous matrix and small metal prills in copper and in arsenical copper spread out everywhere in the slag; 2: 18185, crucible fragment, macro-scale cross section views of smelting crucible from Dzedzvebi, sample no. 4036/15, a large black mineral inclusion of anorthite/augite is noticed on the right cross section, the thin grey bullous slag is characterized by cuprite zones, red, and small malachite fragments, green; 3: SEM‑EDS compositional data for the slag and the metal prills of crucible fragment, sample no. 4036/15, results normalized and expressed in weight percentage (DBM, A. Courcier, T. Skowronek).

Fig. 6 – 1: Dzedzvebi III.5, 18185, crucible fragment, photomicrograph of slag, sample no. 4036/15, taken under plane polarized light, magnification×2.5, the slag contains magnetite clusters precipitated into undetermined siliceous matrix and small metal prills in copper and in arsenical copper spread out everywhere in the slag; 2: 18185, crucible fragment, macro-scale cross section views of smelting crucible from Dzedzvebi, sample no. 4036/15, a large black mineral inclusion of anorthite/augite is noticed on the right cross section, the thin grey bullous slag is characterized by cuprite zones, red, and small malachite fragments, green; 3: SEM‑EDS compositional data for the slag and the metal prills of crucible fragment, sample no. 4036/15, results normalized and expressed in weight percentage (DBM, A. Courcier, T. Skowronek).

12In comparison to the crucible fragments found at the metallurgical workshop of house 1‑2 at Dzedzvebi II.2, the metallurgical evidence is different: arsenical copper droplets found in slag and dross provide a much sounder basis for interpreting them as smelting crucibles (Courcier 2015). But there is the question of how to explain the high amount of finely dispersed arsenic copper splashes within the dross as arsenic minerals have not been described as common in Sakdrisi deposits (Omiadze 2007). We must either assume a mineral anomaly such as domeykite (arsenic copper) that stood in paragenesis with the haematite-quartz deposit, or admit that the sample does not give a conclusive picture at all. Regarding the slaggy phases, it rather seems that a magnetite-rich concentrate containing also copper and arsenic-copper minerals to a smaller extent was smelted. Whether gold was the goal of the smelting practice cannot be decided yet: further investigations taking into account the ore mineralogy from all the Sakdrisi prospects is needed since the whole ore-field exceeds the mineral compositions from the Kachagiani hillock. However we have to interpret this crucible fragment, it was manufactured in relation to gold grinding perhaps at a workshop nearby and later displaced to the grinding site. Considering the free gold levels, the grinders and the smelting pit, there is no reason to doubt the general interpretation of the site as one for gold-processing activities.

An estimate of the amount of gold extracted from the ore-deposit of Sakdrisi-Kachagiani: the “Paravani” calculation

13In order to understand the social impact that the gold production once had, we had to calculate as well the general gold production and the societal and technical efforts of the gold production. Let us start with the estimation of the gold production. As the mines were refilled again with mining debris, we were calculating the amount of gold once left behind and summing it in comparison to the amount of gold that we found still in the single veins that have been exploited. According to our investigations several data could be achieved. First of all there was information about debris ores above ground and refilled mining debris underground. The gold grade of the ore allows insight into the cut-offs of the gold productions. A general mean of 1 ppm Au can be regarded as a minimum number of which could not be removed from the deposit (calculations in detail: Stöllner et al. 2014, tab. 2a‑b; tab. 2). On the other hand there is the average gold amount per each of the gold bearing quartz-hematite veins. A case example will be given here for the lowest mining part of mine 1/2 (called the north extension). Besides the volume and the general ore-content (approximately a third of the volume: 32 m3 with 96.6 t of rock-ore results in roughly 40.4 t of pure ores), we were also able to measure the gold content in some parts of the deposit. Au ranges between 10 ppm and 180 ppm at various parts of this area (in the mean rate: ø 130 g/t Au; ø 70 g/t Ag). Such an estimate would end in a yield of 5.27 kg of gold and roughly 2.8 kg of silver as a natural alloy (tab. 2). In general and according to the partly very high gold amounts, we estimate at the moment a general output of something between 500 to 1,000 kg of gold that once had been produced in Sakdrisi. According to about 600 to 700 years of production in Sakdrisi the calculation would end with a very general annual output of something around 1 kg.

Tab. 2 – Mine 1/2, north extension, calculation ore volume, time of ore extraction, yield of Au/Ag alloy.

  Mining in ancient technique (operation time) (h) Volumes (m3) Specific weight (kg/m3) Weight mined (t) ᴓ Au/Ag yield (kg)
Ore/rocks 4902 (h) ryolithe/dazite, 21.6 m3 2.6 kg/m3 56.2 t 0
  hematite/goethite, 5.65 m3 ᴓ 4.5 kg/m3 25.4 t 5.27 kg Au
2.8 Ag
  quartz, 5.65 m3 ᴓ 2.65 kg/m3 15 t
      ᴓ 3.25 kg/m3 Σ 96.6 t Σ 8.07 kg
Backfill (?) 32 m3x1.3 = breaking = 41.6 (compacted, ca 40 m3) ᴓ 3.25 kg/m3>130 t 130 t Minus 0.130
          Σ 7.94 kg

14Further investigations have been performed by fire-setting experiments which allowed estimating all the efforts that were made with the single steps of gold production from the preparation of the ore mining to the panning and smelting of the gold concentrates (Stöllner et al. 2012; Stöllner et al. 2014). If we take the earring of Paravani with a weight of 9.4 g of Au/Ag as an example, we could calculate its production efforts (further called the “Paravani” calculation; named after the earring of Paravani-Khulgumo: Kvavadze et al. 2007, p. 100; Gambashidze et al. 2010, Nr 580). According to our experiments we could estimate 3.46 min for the production of 1 kg of ore-rock; for the separation and beneficiation work, the value is much higher and can be estimated as 224 min for 1 kg (this is the real effort!). All these working steps take the same time whatever gold enrichment is in the ore-concentrate. But none the less the gold grade influences directly the amount of time to produce 1 g of gold: if one calculates the amount of time according to the enrichment of Ag/Au found in mine 1/2-N extension, one could end up with 42 h to produce 1 g of gold. By considering these data we also get an impression of how many people might have been involved to the production. If we take the Paravani earring as an example, the heavy ring (9.4 g) would have needed at least 3 days by a group of 16 persons to be produced (fig. 7). Without integrating other data, one could argue that expeditions could have sporadically reached the mining site to produce such items on special occasions (e.g. funerals, festivals, gifts and social processes of negotiation). But if one looks to the general estimate of an annual yield of 1 kg of gold, one would estimate the need of 16 workers to produce this amount within 330 working days with eight hours of permanent work. This is not little in a way that other parts of the community also had to provide the subsistence of the mining people. Such a sedentary group perhaps in a size of a larger village can be evidenced at the plateau of Balitschi-Dzedzvebi (Stöllner, Gambashidze 2011; Stöllner et al. 2014; Gambashidze, Stöllner 2016). But looking at the current evidence we do not know if the gold-milling and smelting activities were carried out there seasonally or permanently and in which way of labour division this was done.

Fig. 7 – The “Paravani” calculation shows the time-consumption for producing 1 g of gold in the ore-deposit of Sakdrisi in regard to the technical processes reconstructed by excavations and experiments and the Paravani earring (after Stöllner et al. 2014).

Fig. 7 – The “Paravani” calculation shows the time-consumption for producing 1 g of gold in the ore-deposit of Sakdrisi in regard to the technical processes reconstructed by excavations and experiments and the Paravani earring (after Stöllner et al. 2014).

Simply the other side of the coin: the ritual interplay of the chaîne opératoire

15During our field work it was astonishing to detect so many features at all work areas that could not be explained simply by technical processes (Stöllner 2017). Let us briefly look at these features, especially at the aspect of their temporal relations with the gold-working operations. Were they carried out prior to the work processes or afterwards? It is an important question whether the rituals were carried out to express hope (i.e. to find enough gold) or to express gratitude at the end of the process (i.e. to have found enough gold).

16Let us start with the mining process: the most obvious observations were tool depositions in the galleries, often discovered at the end of drifts and galleries. During the underground excavation between 2005 and 2011, several hammer-stone depositions were discovered and were conspicuous as special depositions. The first and most obvious one was discovered inside the north-western niche of mine 2 (feature 25011) (fig. 8.1‑2): four complete hammer stones, to which a fifth, fragmented one was found nearby, had been put on the bottom rock before the niche was filled with debris. The hammer tools themselves display a special “unused” and well-kept image. Such deposits have been found in different areas, at the bottom of working edges (mine 1/2: features 24022 and 24069‑5, and mine 1/3: 23005) (fig. 8.3‑4, 8.7) and at ends of drifts, such as in the northernmost part of mine 1/2 where the mine ended after a last fire-setting (feature 24043) (fig. 8.6). In each of the cases at least one bigger hammer tool was included and again such peculiar pieces seem unused or scarcely used (fig. 8.1). It is rather difficult to understand the motivations behind such depositions, but if we look for possible narratives we can assume that the deposition of used hammers was probably connected with the work once done in the locations where the hammers were deposited. A piece of broken antler which was discovered at the southernmost end of mine 1/2 in a small layer of fire-setting debris at the bottom of the rock can be included in such a context. This position and stratigraphic relation is conspicuous because of the fact that all the other debris were filled in later, so the piece could have been broken during usage on site and left deposited nearby.

Fig. 8 – Sakdrisi, deposits of hammer-stones and an antler fragment underground; 1: hammerstones found in the NW niche of mine 2, 25011 (DBM/RUB, T. Rabsilber); 2: find situation of deposit 25011 (DBM/RUB, T. Stöllner); 3‑7: find situations of hammer tool deposits at mine 1/3, feature 23005 (7) in mine 1/2 north extension, features 24069/5, 24022 (3‑4), of an antler (5) and in the northernmost gallery (6) (DBM/RUB, T. Stöllner, P. Thomas).

Fig. 8 – Sakdrisi, deposits of hammer-stones and an antler fragment underground; 1: hammerstones found in the NW niche of mine 2, 25011 (DBM/RUB, T. Rabsilber); 2: find situation of deposit 25011 (DBM/RUB, T. Stöllner); 3‑7: find situations of hammer tool deposits at mine 1/3, feature 23005 (7) in mine 1/2 north extension, features 24069/5, 24022 (3‑4), of an antler (5) and in the northernmost gallery (6) (DBM/RUB, T. Stöllner, P. Thomas).

17Some comments should finally be made in respect of the filling of the mines after they had been exploited: this refilling work has no technical meaning (e.g. for stabilizing the mine, like packing with debris would have had), as the mine has stable rock faces and packing would have hindered any further exploitation alongside the veins. So, one should discuss another meaning: when looking at the filled layers themselves, it is clear that some packing was done deliberately and with great attention (Stöllner et al. 2014, pp. 80‑82, fig. 10), while other, coarser layers developed by mining the ceiling in a second working step.

  • 4 The ceramic vessel from the Kura-Araxes period found in Sakdrisi comprises only fragments of big s (...)

18If we leave the mining work, we certainly may ask if the crushing stage has left further examples of ritual activity: as the crushing itself certainly was going on monotonously for hours (e.g. proven in experiments: Stöllner et al. 2012, pp. 72‑74, tab. 2), it is likely that this work was paired with social activities such as singing and storytelling. Unfortunately no remnants of such activities can be found, and drinking and eating activities cannot be traced unambiguously either (through, for example, a fireplace or a deposition of ceramic findings).4 What was even more interesting was the finding of an idol stand at the millstone and tool dump already described (Gambashidze, Stöllner 2016, pp. 34, 43). As pottery fragments were also found within deposit 10042 to a larger extent (about 77 fragments from storage vessels and handled mugs and goblets), it seems that the ceramic stand was somehow connected with drinking practices during the milling and panning work (Otchvani et al., this volume). We can only guess whether the treacherous gold content within the veins (they could vary to a great extent; for analyses see Stöllner et al. 2014, pp. 89‑92, 105‑106) would have also triggered religious practices, especially when the moment came to see a larger, smaller or even no golden “face side” within the pan. This is still today one of the “magic moments” in manual gold processing. This moment, however, decided the fortunes of the mining group, and might therefore be connected with the influence of numinous powers.

19As the Sakdrisi mine was probably not operated constantly the whole year round, it is not surprising to find evidence of only minor traces of ritual activities: as the surface parts were heavily eroded and destroyed by late antique to medieval activities, even more traces might have been destroyed over time. The situation is better when looking at the settlement of Dzedzvebi (fig. 1).

  • 5 All the collective burials investigated so far contained male and female adults but also one juven (...)

20When looking at the Kura-Araxes settlement of Dzedzvebi, there are obvious ritual interconnections with the gold-milling and gold-panning processes as well as with other metallurgical activities: one of the most conspicuous examples was detected in the copper workshop area (fig. 9.1, 9.3) of a house where a skull grave (no. 7: fig. 9.2) of a juvenile was discovered beneath a grinding station consisting of two large grinding plates. It is obvious that the area was of “meaning” as the accompanying hearth as well as the grinding plates were used over a longer period and renewed in a second phase (fig. 9.3). The connection with the grave is undeniable, as can also be seen from a similar skull grave in the form of a stone cist attached to the large milling house 3/Dzedzvebi II.3 (grave 5). While the prior status of the burial is evident for house 2, this is not the case with house 3, where the slab cist seems attached to an older wall construction. What is striking anyway is the choice of parts of a juvenile to display ritual concepts. It is worth looking in general at the Kura-Araxes graves discovered nearby and surrounding the workshop areas: most of them can be characterized as collective burials and apart from grave 2 all of them date to the second Kura-Araxes stage (e.g. graves 1‑3, 6: Gambashidze, Stöllner 2016, pp. 80‑85). What is interesting is that all the collective burials contained one juvenile besides a male and some female persons.5

Fig. 9 – The Sakdrisi-Dzedzvebi complex near Kazreti-Balitchi; 1: Dzedzvebi area II.2, house 2, workshop, vertical photography; 2: Dzedzvebi area II.2, grave 2, 3D SFM model of the skull and vessel deposition; 3: Dzedzvebi area II.2, house 1 and 2 plus workshop, drawing (GNM; DBM/RUB, T. Rabsilber, F. Klein [3-D models], T. Sachvadze [map], T. Stöllner).

Fig. 9 – The Sakdrisi-Dzedzvebi complex near Kazreti-Balitchi; 1: Dzedzvebi area II.2, house 2, workshop, vertical photography; 2: Dzedzvebi area II.2, grave 2, 3D SFM model of the skull and vessel deposition; 3: Dzedzvebi area II.2, house 1 and 2 plus workshop, drawing (GNM; DBM/RUB, T. Rabsilber, F. Klein [3-D models], T. Sachvadze [map], T. Stöllner).

21Milling must have been an important stage, particularly when regarding the time that was needed to sort and mill the crushed ores that had been transported from Sakdrisi. As with the crushing and testing work in Sakdrisi, it is difficult to appraise the contemporaneous social activities during the work, but that feasting activities could have played a role can be evidenced by two depositions in the excavation area II.7-8 excavated during the campaigns 2014 and 2015. In pit 37006, a deposition of four ceramic vessels was discovered that was associated with burnt hearth fragments (or a very coarse storage vessel) and a milling rubber (fig. 10.1‑3). The whole feature might have been deposited during a feasting ritual after which the tool was laid down with the feasting crockery. According to the drink set, one may assume a smaller group of people had shared the event before depositing the remnants in the pit. It seems the vessels were deliberately destroyed before being deposited in the pit: at the end only the drinking bowl was put upside down into the pit. We can only assume that the event took place during the gold-milling process but one cannot be sure whether the ritual took place prior to or after the milling process.

Fig. 10 – Dzedzvebi, area II.8, house 4; 1: map of upper level with ceramic/stone deposits; 2: ritual deposit of vessel with red pigment and two ceramic stands beside; 3: artefacts of the ritual pigment deposit; 4: decorated plaster fragment 25521, found with burned clay debris on top of the vessel (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [maps], T. Sachvadze [drawings], T. Stöllner [photos of features]).

Fig. 10 – Dzedzvebi, area II.8, house 4; 1: map of upper level with ceramic/stone deposits; 2: ritual deposit of vessel with red pigment and two ceramic stands beside; 3: artefacts of the ritual pigment deposit; 4: decorated plaster fragment 25521, found with burned clay debris on top of the vessel (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [maps], T. Sachvadze [drawings], T. Stöllner [photos of features]).
  • 6 As the pit was heavily destroyed and overlaid by the ceramic deposition and its cultural layer, no (...)
  • 7 No crucible fragments were found and a fired hearth near the panning installation was heavily dest (...)

22This is different with the features of house 4, in which several hearths were uncovered, most likely from different occupation levels inside the house. As the house had been partly destroyed by downhill erosion and an animal den, only the upper parts were well preserved. At least three phases can be reconstructed according to hearths and occupation layers: it appears to have been a round- to oval-shaped house of about 7 to 8 m in diameter (Gambashidze, Stöllner 2016, pp. 72‑75, 92‑97). The western uphill part was the best preserved: a clay bench was detected at the westernmost part of the round-shaped house and a central- to western-lying stone tool and rock deposition (38011) was found on top of a ceramic deposition (38017) (fig. 4). Below the stone tool and rock covering 38011 were discovered not only regularly deposited ceramics but also a pit with the bottom part of a storage vessel (38057). The most conspicuous feature was the deposition of a three-handled vessel and two ceramic stands still in situ on top of the western clay bench (fig. 11). The feature was covered with lightly fired daub fragments, which had perhaps fallen down from a roof or a side wall. The vessel was filled partly with coarsely ground haematite (for the analyses see T. Skowronek in Stöllner 2017), most likely derived from Sakdrisi gossan ores but containing almost no gold. It is undoubtable that the pigment played a ritual role within this house, even more so if one considers the ceramic deposition found beneath the stone tool/rock deposition and the gold-panning installation south-east of it. According to the analyses of the sediments from the 38046 feature, gold might have been washed there. But there are no indications of the grinding work, although grinding rubbers and even one tool set of grinding plate and rubber were deposited on top of the ceramic deposition 38017. The ceramic deposition (38017) consisted of several drinking mugs, nicely decorated storage vessels, two ceramic idol stands and ceramic lids. So, the deposition might be connected with feasting activities at the end of the gold processing, before its remnants were covered by a rock and work-stone deposition, some of which were put in a row outside the deposition. If we take the high amount of rubbers, one could assume that they once belonged to the working gang who laid them down after finishing their work. A close connection with the panning installation 38046 is given also by the central pit 38057 in which the milled powder originally might have been collected.6 Although house 4 might have had a central role for producing gold concentrates, there is no evidence that gold was smelted here.7

Fig. 11 – 1: Dzedzvebi area II.7, pit 37006, photograph; 2: Dzedzvebi area II.7, sections through pit 37006; 3: pottery and stone tool from the pit (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [drawings], T. Stöllner [photo]).

Fig. 11 – 1: Dzedzvebi area II.7, pit 37006, photograph; 2: Dzedzvebi area II.7, sections through pit 37006; 3: pottery and stone tool from the pit (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [drawings], T. Stöllner [photo]).

23However the rock and stone tool deposition 38011 was deposited on the top and at the end of all the activities in house 4 similar depositions of work tools on top of working areas were observed several times. One case is the gold-grinding and smelting site III.5, where the grinding rubbers were deposited on top of a presumably temporarily used work area. The more elevated gold content of the sediments found beneath this deposition provides reason to reconstruct a temporarily used production area (tab. 1). The covered area once consisted of a deep smelting pit (with an ashy filling on which a crucible fragment was found) and two other shallow pits nearby. But in comparison to house 4 there are no indications of feasting practices and no longer-lasting cult activity at all: it may seem that all the work tools used were carefully put on top of the area before it was left. In comparison to house 4, one should mention finally the close connection of some of the working areas with graves in general. As with the working area and its ritual interplay, we see this close spatial relationship also with the burial grounds (see above the II.7 and II.8 area) – there is a general close intermingling especially in the areas II and III. In one case there is even a tight interconnection with the mining community. At grave 9 a stone tool was deposited on top of the grave similar to other libation vessels, something that points to some commemorative practices posterior to the funeral. It is interesting to note that nearby grave 9 another suspicious complex was discovered. The rounded platform of “house 5” was first identified as a house with the typical hearth in the centre, but at least the hearth was noticeable and seemed to be a very special and large example. Sidewards a male grave was found in a small stone slab cist, while a deposition of animal bones discard came to light in the northeast of the round-shaped platform. As there were nearly no domestic findings and even the house construction (no postholes at all were found) seemed unclear, we thought rather on a ritual place than on a normal domestic area (fig. 12; Gambashidze, Stöllner 2016, pp. 86‑89).

Fig. 12 – Dzedzvebi, area II.7, house 5; 1: map of stone pebble layer with central hearth and grave 7 sidewards; 2: central hearth, section and drawing; 3: view from north to the 2015 excavation of “house 5” and grave 9 in the back part with ceramic/stone deposits (GNM; DBM/RUB, F. Klein [map], T. Sachvadze [drawings], T. Stöllner [photos of feature]).

Fig. 12 – Dzedzvebi, area II.7, house 5; 1: map of stone pebble layer with central hearth and grave 7 sidewards; 2: central hearth, section and drawing; 3: view from north to the 2015 excavation of “house 5” and grave 9 in the back part with ceramic/stone deposits (GNM; DBM/RUB, F. Klein [map], T. Sachvadze [drawings], T. Stöllner [photos of feature]).

24Let us shortly sum up what already had been argued on the basis of the archaeological observations: nearly all of the stages so far discovered in the gold-production complex of Sakdrisi and Dzedzvebi were accompanied by ritual activities. The main components are deposition rituals that left parts of the tools, or even important ones, in the production areas (inside the mines, the craft zones). Most of these depositions had been made subsequent to the actual work processes, but this temporal emphasis has its reason also in the character of the archaeological data. There are also indications of contemporaneous but also prior rituals: the results from house 4 (area II.8), and also from the workshop of house 2 (area II.2), show the deposition of a skull and two vessels (grave 7) beneath a metallurgical activity zone before the actual grinding and smelting process had started. The deposition and usage of red pigments was connected likely with the gold-panning process (the smelting is unclear yet) contemporarily. Although it is difficult to prove evidence of contemporaneity with the work processes, it is likely that at least some social interaction took place during the work. Feasting most likely took place in the context of all the processes at the end, but who knows whether an assemblage like the deposition of vessels and an old milling rubber in pit 37006 was not assembled before the work had started?

Conclusions

25As for a conclusion on the economic activities of the Sakdrisi mining operation it has to be asked if the activities have been carried out all through the year or intermittently. The easier a question is raised, the greater is the difficulty to answer it. If we look to the workshop area and its houses in Dzedzvebi it is not completely sure if they were used the whole year round: it is a striking aspect that kitchen refuse or other domestic debris as well as grain mills have not been discovered in large quantities if at all. This makes one wonder if the houses have not been used only for some periods in the year, especially when groups were processing the ores from the Sakdrisi deposit. According to our surveys at Dzedzvebi it is likely that the settlement plateau was not densely occupied during the earlier Kura-Araxes periods: it looks rather like a loosely populated site with scattered house groups especially at the slope area of Dzedzvebi II. Such a settlement structure would undoubtedly allow a continuous resettling and shift of occupied areas over the whole plateau thus indicating a more flexible pattern of people coming and going over the year. An area that confirms such an impression is the grinding and smelting site III.5 that was installed isolated in the open countryside of the plateau, obviously used only for a rather short operation period.

26As it has been outlined for the gold operations in general, there is no need to assume permanent mining, neither by the yield argument nor by technical reason. As far as the yield is concerned smaller mining groups could have entered the Kachagiani hillock for some days in order to produce a smaller amount of prestigious items. But how would this explain the workshop houses and the settlements and burial grounds in Dzedzvebi? There was at least a permanency in continuous entering and using the settlement area either on multiple occasions or for a longer-lasting seasonal activity needing the workshop for repeatable activities. However we interpret the find situation, it certainly needs more investigations of Kura-Araxes sites at Dzedzvebi and in the surrounding to understand subsistence structures and population density. R. Berthon’s results on the domestic animal pattern at least indicate a more permanent and specialized husbandry (Berthon et al., this volume). This would slightly indicate also stable settling conditions but the archaeobotanical results are still too unclear. As no Kura-Araxes storage pits (despite three completely clean ones at site III.2; Stöllner et al. 2014, p. 100, fig. 26) have been discovered and investigated at all, we are lacking simply basic data to further understand the subsistence structure.

27However we interpret the durability of the mining operation, it is clear that the work practice itself had a major effect on the ideological sphere of the involved groups. Nearly all the gold-working and also the other metallurgical activities were accompanied by ritual activities, something that might have bound the Kura-Araxes crafts activities to a religious sphere in general. The recent excavations of Chobareti (Javakheti region in south-eastern Georgia: e.g. Kakhiani et al. 2013) brought another example of such connections to light. At the burial 11 the excavators discovered a saddle quern and an ingot mould, something that is possibly related also to crafts activities and the ritual interplay with it. The late Antonio Sagona wrote:

  • 8 Email from A. Sagona, Melbourne, 15/07/2016. Thanks to Tony, we are able to mention this important (...)

At first we thought it might have been pit (hence the sectioning), but it turned out to be a pit burial covered by flat long stones. On top of the burial were several items, including what I assume is an ingot mould, and a grinding stone and rubber [fig. 13].8

Fig. 13 – Chobareti, Javaketi region, burial 11; 1: upper level with saddle quern and grinder; 2: lower level with burial; 3: ingot found on top nearby the saddle quern (Chobareti expedition, by courtesy of the late Prof. Dr A. Sagona, Melbourne).

Fig. 13 – Chobareti, Javaketi region, burial 11; 1: upper level with saddle quern and grinder; 2: lower level with burial; 3: ingot found on top nearby the saddle quern (Chobareti expedition, by courtesy of the late Prof. Dr A. Sagona, Melbourne).

28At the moment we do not know how large the (semi-) permanent mining group of Sakdrisi/Dzedzvebi has been. According to the size of the settlement, we can estimate there were living some dozens of people permanently while others might have joined the dwellers particularly during the time of the mining operations. This model is based on the scattered structure of the settlement, the highly specialized character of some of the settlement parts but also on the small evidence of agricultural activities. Such a model would neither mean that agricultural practices were not followed nor that the dwellers altogether were mobile. However, such a model would fit best to the evidence of rituals carried out at graves in a commemorative manner as well as at areas that were used to assemble people for special feasting occasions. The complex of graves 8 and 9 as well as the rounded platform of “house 5” come together with a special and large hearth complex in the centre such a meeting site (fig. 12; Gambashidze, Stöllner 2016, pp. 86‑89). It is likely that the complex social reality of the gold-mining societies was constantly re-arranged by the manifold working processes and by a volatile coming and leaving of people. This stimulated practices of mediation such as feasting, commemorative practises and others that were related to the “magic” gold production itself. But if this also stimulated the evolvement of social hierarchies that later became apparent with the kurgan burials of the Martkopi-Bedeni is another question that cannot be followed here. At the gold-mining communities of Sakdrisi-Dzedzvebi, there is nothing recognizable that has evolved beyond a short-lived social institution, very much as in a “big-manship”.

Appendix: the Dzedzvebi and Sakdrisi complex in the light of some archaeobotanical data (N. Boenke, W.H. Schoch)

  • 9 Although the data and detailed data processing will be published by N. Boenke and W.H. Schoch else (...)

29The plateau around the site of Dzedzvebi is nowadays characterized by a dry grassland vegetation with some shrubs like Christ’s thorn (Paliurus spina-christi Mill.), common hawthorn (Crataegus monogyna Jacq.) and other Rosaceae (Rosa spec.), which are remnants of the land-use for pastoralization.9 Just little areas mainly in the lowland are cultivated. In the deeply incised small river valleys a speciose forest with oak trees (Quercus spec.), elm (Ulmus spec.), maple (Acer spec.), common hornbeam (Carpinus betulus L.), oriental hornbeam (Carpinus orientalis Mill.), flowering ash (Fraxinus ornus L.) and single walnut trees (Juglans regia L.) is growing. The surrounding lower eastern slopes of the Lesser Caucasus are dominated by oak forests. In the dry soil at Dzedzvebi only charred botanical material have survived; this was the same inside the Sakdrisi mine where just charcoal was detected (Schoch 2013). The preservation of the charred material is very brittle. The preservation situation inside the mine is better. Nevertheless, botanical material is available for all phases of the settlements at Dzedzvebi (tab. 3). The preservation of the material is not correlated with the age but with the character of the features. While the botanical evidence for the early phases and the Iron Age is quite equally distributed to several pits and some Iron Age ovens, the botanical evidence of Kura-Araxes times just derives from one single feature, the floor and loam construction layer 36082 from Dzedzvebi IV.3 (South plateau) – despite charcoal remains from various features.

Tab. 3 – The botanical record at Dzedzvebi; seeds, fruits and vegetative remains documented per period (after N. Boenke, Bochum).

    Late Neolithic Chalcolithic Kura‐Araxes Iron Age
Cultivated plants          
Triticum aestivo‐compactum bread wheat, naked wheat        
Triticum aestivum L. – spikelets bread wheat        
Triticum monococcum L. einkorn wheat        
Triticum monococcum L. – two grained einkorn wheat – two grained        
Triticum monococcum L. – spikelets einkorn wheat        
Triticum dicoccum Schrank emmer wheat        
Triticum spec. wheat        
Panicum miliaceum L. broomcorn millet        
Avena spec. oat        
Cerealia cereals indet.        
Cerealia – spikelets cereals indet.        
Bread/Cereal/Fruit fragments          
Weeds          
Chenopodium album L. allgood        
Buglossoides arvensis L. corn groomwell        
Fallopia convolvulus (L.) Á. Löve black bindweed        
Fumaria spec. fumitory        
Galium cf. aparine L. catchweed        
Galium spec. bedstraw        
Neslia cf. paniculata (L.) Desv. ball mustard        
Silene spec. campion        
Trifolium spec. clover        
Seeds/fruits indet.          
Fruits          
Cornus sanguinea L. common dogwood        

30Overall, the archaeobotanical record is limited on cultivated cereals and a small range of weeds and one wild fruit species. Legumes or oil plants had not been detected yet. The use of cereals starts as typical with einkorn wheat in the Late Neolithic and einkorn and emmer wheat in Late Chalcolithic features. Broomcorn millet is also surprisingly frequent from the beginning. The Kura-Araxes feature shows a shift to the cultivation of bread wheat (e.g. also Longford, this volume; Longford, Drinnan, Sagona 2009). No other cereal species is documented, but the presence of bread wheat chaff points to the cultivation and processing of cereals and not only on consumption. The documented weed seeds are different to the other periods. In Iron Age times again bread wheat, einkorn wheat and broomcorn millet were used. One mineralized fruit of common dogwood is just evidenced from the Chalcolithic.

31From Sakdrisi mine 2041 stratigraphically determined charcoal pieces were collected from a total of eight mine pits, that were exploited on the surface, and mine-shafts that were driven down to a 28 meters level (mine pits A and B, mine 1 and 2 underground) (fig. 14). It was difficult, but in most cases possible, to identify special anatomical forms, such as reaction wood and curl-wood, which is mechanically more resistant because of its thick-walled tracheids. Quercus sp. L., subgenus Quercus Oersted, (most probably Quercus petraea) dominates, making up over 90% of the particles, followed by 12 other genera or species. The analysis aims to establish the use of the wood as a fuel for “fire-setting”, the selection of stem-wood and branch-wood and the quality of the wood, as well as to reconstruct the vegetation at the time of the ancient mining. The present material particularly illustrates the difficulty of determining wood right down to the species level, for instance for cases of the pimpenut (Staphylea sp.), in which cases we are not able to distinguish between different species or individual variations.

Fig. 14 – Sakdrisi mine, wood species of charcoal samples (after W.H. Schoch, Birmensdorf).

Fig. 14 – Sakdrisi mine, wood species of charcoal samples (after W.H. Schoch, Birmensdorf).

32The anthracological examination of Dzedzvebi and Sakdrisi reflects the big variety of the surrounding Caucasian forests (tab. 4); from the closed, oak dominated mixed forests to more open condition and the forest edge, from the drier, opened and maybe pastoralized landscape to the fresh or moist habitats near the rivers. With an average of 88% in numbers oak is the most important wood from the economical and absolute perspective. This applies for all periods just with little variation to the layers of Dzedzvebi and Sakdrisi, so oak must have been a valuable and available source for firing and timbering all the time. To gather more information other accounts are necessary. Therefore, the depiction of the continuity is chosen (tab. 4) because of the different state of the archaeological record, e.g. some samples contain just single pieces of charcoal, others lot of pieces or just tiny fragments. This is another reason why a first evaluation by continuity seems reasonable; 45 features had been evaluated, 13 of them belonging to Dzedzvebi settlement, the others to the Sakdrisi gold mine. If we go more in detail, it can be added that just 30% of the features in the mine contain only one species (half of them oak, the other different taxa). At Dzedzvebi – even in situations with a little number of fragments per feature – all results are mixed ensembles in average with at least two to five taxa. Overall again oak is the dominating species, but it is visible that other species are also used frequently and are not single exceptions. It is possible to argue that the people visited and used all parts of their environment.

Tab. 4 – Overview of the recorded wood taxa at Sakdrisi and Dzedzvebi from the Late Neolithic, the Late Chalcolithic and the Kura-Araxes period according to their vegetation zones (N. Boenke, RUB).

  character of the vegetation
Paliurus spina-christi Miller, Christ’s thorn open forests
Rhamnus spec., buckthorn
Rhus coriaria L., sicilian sumac
Cornus spec., dogwood clear forest, forest edge
Maloideae, pomaceous fruit
Prunoidea, stone fruit
Acer campestre L., field maple oak dominated mixed forest
Acer spec., maple
Carpinus betulus L., hornbeam
Ostrya carpinifolia Scop., hop-hornbeam
Fagus spec., beech
Quercus spec., oak
Fraxinus spec., ash riparian forest, moist grounds
Staphylea spec., bladdernut
Ulmus spec., elm
Salix spec., willow
decidious wood  
wood, taxa unknown

33Although we can see quite an undifferentiated picture, some results can be highlighted for the Sakdrisi gold mine. The first is the presence of taxa from the open habitats with Christ’s thorn, buckthorn and Sicilian sumach. Those are found predominately at the mine. Perhaps these taxa with a low caloric value had been used for lighting the fire-setting at the mine. In this context probably also the occurrence of branch-wood can be interpreted.

34The second aspect is the wood composition at feature 24011 (mine 1/2), where nine different species had been recorded. Species of the open habitats are lacking, but therefore trees like ash tree, bladdernut, elm and even willow occur as well as species of the mixed oak forest and its edges. This might either be a result of a mixture of different disposal origins or a special selection of firewood by accident or on purpose.

Bibliographie

Agricola 1556: G. Agricola, Zwölf Bücher vom Berg- und Hüttenwesen, Übersetzt und bearbeitet von C. Schiffner, Munich, Beck, 1556.

Akhundov 2004: T. Akhundov, “South Caucasus in the Neolithic and Early Bronze Age: the question of epochs and periods”, in A. Sagona (dir.), A view from the highlands. Archaeological studies in honour of Charles Burney, Leuven, Peeters, 2004, pp. 421‑435.

Akhundov, Makhmudova 2008: T.I. Akhundov, B.A. Makhmudova, Juschnij Kavkaz v kavkazsko-peredneaziatskikh ethnokulturnykh prozessakh. IV. tys. do n.e., Baku, 2008.

Bachmann 1999: H.‑G. Bachmann, “On the early metallurgy of gold. Some answers and more questions”, in A. Hauptmann, E. Pernicka, T. Rehren, Ü. Yalcin (dir.), The beginnings of metallurgy, published in Der Anschnitt. Beiheft 9, 1999, pp. 267‑275.

Batiuk 2013: S. Batiuk, “The fruits of migration: understanding the ‘longue durée’ and the socio-economic relations of the Early Transcaucasian culture”, Journal of Anthropological Archaeology 32/4, 2013, pp. 449‑477.

Chataigner, Palumbi 2014: C. Chataigner, G. Palumbi (dir.), “The Kura Araxes culture from the Caucasus to Iran, Anatolia and the Levant: between unity and diversity”, Paléorient 40/2, 2014, pp. 7‑260.

Courcier 2015: A. Courcier, “Archaeometallurgical studies on crucibles, slags, ores and metal objects discovered at Dzedzvebi (Georgia)”, Bochum, 2015 (unpublished).

Gambashidze et al. 2001: I. Gambashidze, A. Hauptmann, R. Slotta, Ü. Yalçin, Georgien. Schätze aus dem Land des Goldenen Vlies, Katalog der Ausstellung (Bochum, Veröffentlichungen aus dem Deutschen Bergbau-Museum Bochum 100), Bochum, Deutsches Bergbau-Museum, 2001.

Gambashidze et al. 2010: I. Gambashidze, G. Mindiashvili, G. Gogotshuri, K. Kakhiani, I. Dshaparidze, Ancient metallurgy and mining in Georgia in the 6th to 3rd millennium BC, Tbilisi, 2010 (in Georgian).

Gambashidze, Stöllner 2016: I. Gambashidze, T. Stöllner, The gold of Sakdrisi. Man’s first gold mining enterprise, Rahden, Leidorf, 2016.

Gilman 1996: A. Gilman, “Craft specialization in late prehistoric Mediterranean Europe”, in B. Wailes (dir.), Craft specialization and cultural evolution. In memory of V. Gordon Childe, University of Pennsylvania Museum Monograph 93, Philadelphia, The University Museum of Archaeology and Anthropology, 1996, pp. 67‑71.

Govedarica 2002: B. Govedarica, “Die Maikop-Kultur zwischen Europa und Asien: Zur Entstehung einer Hochkultur im Nordkaukasus während des 4. Jts. v. Chr.”, in R. Aslan, S. Blum, G. Kastl, F. Schweizer, D. Thumm (dir.), Mauerschau. Festschrift für Manfred Korfmann 2, Remshalden/Grunbach, Greiner, 2002, pp. 781‑799.

Hansen 2013: S. Hansen, “Innovative metals: copper, gold and silver in the Black Sea Region and the Carpathian Basin during the 5th and 4th millennium BC”, in S. Burmeister, S. Hansen, M. Kunst, N. Müller‑Scheessel (dir.), Metal matters. Innovative technologies and social change in Prehistory and Antiquity, Rahden, Leidorf, 2013, pp. 137‑167.

Hauptmann et al. 2010: A. Hauptmann, C. Bendall, G. Brey, I. Japaridze, I. Gambashidze, S. Klein, M. Prange, T. Stöllner, “Gold in Georgien. Analytische Untersuchungen an Goldartefakten und an Naturgold aus dem Kaukasus und dem Transkaukasus”, in S. Hansen, A. Hauptmann, I. Motzenbäcker, E. Pernicka (dir.), Von Majkop bis Trialeti. Gewinnung und Verbreitung von Metallen und Obsidian in Kaukasien im 4.‑2. Jt. v. Chr., Bonn, Rudolf Habelt, 2010, pp. 139‑160.

Japaridze 2013: O. Japaridze, Brinjaos industriis istoriisatvis udzveles Sakartveloshi (On the History of Bronze industry in Old Georgia), Tbilisi, 2013.

Javakhishvili 1998: A. Javakhishvili, “Ausgrabungen in Berikldeebi (Šida Kartli)”, Georgica 21, 1998, pp. 7‑20.

Kakhiani et al. 2013: K. Kakhiani, A. Sagona, C. Sagona, E. Kvavadze, G. Bedianashvili, E. Messager, L. Martin, E. Herrscher, I. Martkoplishvili, J. Birkett-Rees, C. Longford, “Archaeological investigations at Chobareti in southern Georgia, the Caucasus”, Ancient Near Eastern Studies 50, 2013, pp. 1‑138.

Kohl 2009: P.L. Kohl, “Origins, homelands and migrations: situating the Kura-Araxes Early Transcaucasian ‘culture’ within the History of Bronze Age Eurasia”, Tel Aviv 36, 2009, pp. 241‑265.

Korenevskij 2004: S.N. Korenevskij, Drevnejshie zemledelzy I skotovody Predkavkazia, Moscow, 2004.

Kushnareva 1997: K. Kushnareva, The southern Caucasus in Prehistory. Stages of cultural and socioeconomic development from the eight to the second millennium BC, Pennsylvania University Museum Monograph 99, Philadelphia, 1997.

Kvavadze et al. 2007: E. Kvavadze, K. Kakhiani, N. Pataridze, S. Konnor, “The results of palynological investigation of Paravani Kurgan”, Proc. Georgian Acad. Science Biol. SerB 5/2, 2007, pp. 97‑107.

Longford, Drinnan, Sagona 2009: C. Longford, A. Drinnan, A. Sagona, “Archaeobotany of Sos Höyük, northeast Turkey”, in A. Fairbairn, S. O’Connor, B. Marwick (dir.), New directions in archaeological Science, Terra Australis 28, Canberra, Australian National University E Press, 2009, pp. 121‑136.

Lyonnet et al. 2008: B. Lyonnet, T. Akhundov, K. Almamedov, L. Bouquet, A. Courcier, B. Jellilov, F. Huseynov, S. Loute, Z. Makhazadze, S. Reynard, “Late Chalcolithic kurgans in Transcaucasia. The cemetery of Soyuq Bulaq (Azerbaijan)”, Archäologische Mitteilungen aus Iran und Turan 40, 2008, pp. 27‑44.

Makharadze 2007: Z. Makharadze, “Nouvelles données sur le Chalcolithique en Géorgie orientale”, in B. Lyonnet (dir.), Les cultures du Caucase (VIe‑IIIe millénaires avant notre ère). Leurs relations avec le Proche-Orient, Paris, CNRS Éditions, 2007, pp. 123‑132.

Munchaev 1975: R.M. Munchaev, Kavkaz na zare bronzovogo veka: Neolit, Eneolit, Rannaja Bronza, Moscow, Nauka, 1975.

Omiadze 2007: A. Omiadze, “Mineralogical and petrological analysis of Sakdrissi Deposit”, MA thesis, Tbilisi State University, 2007 (unpublished).

Otchvani et al. 2015: N. Otchvani, I. Gambashidze, T. Stöllner, M. Jansen, “Early Bronze Age pottery from Sakdrisi mine and Dzedzvebi settlement”, Dmanisi. Bull. Georgian National Mus. 6 (51‑B), 2015, pp. 6‑29 (in Georgian).

Palumbi 2007: G. Palumbi, “From collective burials to symbols of power. The translation of role and meanings of the stone-lined cist burial tradition from southern Caucasus to the Euphrates valley”, Scienze dell’Antichità 14, 2007, pp. 17‑44.

Ramsl 1998: P.C. Ramsl, Inzersdorf-Walpersdorf. Studien zur späthallstatt-/latènezeitlichen Besiedlung im Traisental, Niederösterreich, Materialheft A6, Vienna, Fundberichte Österreich, 1998.

Rehren, Temme 1994: T. Rehren, M. Temme, “Pre-Columbian gold processing at Putushio, South Ecuador. The archaeometallurgical evidence”, in D. Scott, P. Meyers (dir.), Archaeometry of Pre-Columbian sites and artefacts, Los Angeles, Getty Conservation Institute, 1994, pp. 267‑284.

Sagona 1984: A. Sagona, The Caucasian Region in the Early Bronze Age, BAR International Series I.3, Oxford, BAR Publishing, 1984.

Schoch 2013: W.H. Schoch, “Charcoals from possibly the world’s oldest gold mine in Sakdrissi (Georgia)”, in 4th International meeting of anthracology Brussels (8‑13 September 2008), BAR International Series 2486, Oxford, Archaeopress, 2013, p. 131.

Stöllner 2016: T. Stöllner, “The beginnings of social inequality. Consumer and producer perspectives from Transcaucasia in the 4th and the 3rd millennia BC”, in M. Bartelheim, B. Horeijs, R. Krauss (dir.), Von Baden bis Troia. Ressourcennutzung, Metallurgie und Wissenstransfer. Eine Jubiläumsschrift für Ernst Pernicka, Oriental and European Archaeology 3, Rahden, Leidorf, 2016, pp. 209‑234.

Stöllner 2017: T. Stöllner, “The ritual interplay: gold mining practices in the late 4th and early 3rd millennia BC”, in V. Eriksen, A. Abegg-Wigg, R. Bleile (dir.), Interaction without borders. Exemplary archaeological research at the beginning of the 21st century. Festschrift für Claus von Carnap-Bornheim zum 60. Geburtstag, Schleswig, Stiftung Schleswig-Holsteinische Landesmuseen, 2017, pp. 119‑136.

Stöllner et al. 2010: T. Stöllner, I. Ġambašiże, A. Hauptmann, G. Mindiašvili, G. Gogočuri, G. Steffens, “Goldbergbau in Südostgeorgien. Neue Forschungen zum frühbronzezeitlichen Bergbau in Georgien”, in S. Hansen, A. Hauptmann, I. Motzenbäcker, E. Pernicka (dir.), Von Majkop bis Trialeti. Gewinnung und Verbreitung von Metallen und Obsidian in Kaukasien im 4.‑2. Jt. v. Chr., Bonn, Rudolf Habelt, 2010, pp. 103‑138.

Stöllner et al. 2012: T. Stöllner, B. Craddock, S. Timberlake, I. Gambashidze, “Feuersetzen im frühesten Metallerzbergbau und ein Experiment im frühbronzezeitlichen Goldbergbau von Sakdrisi, Georgien”, in K. Oeggl, V. Schaffer (dir.), Die Geschichte des Bergbaues in Tirol und seinen angrenzenden Gebieten. Proceedings 6. Milestone-Meeting Klausen 2011, Innsbruck, University Press, 2012, pp. 65‑77.

Stöllner et al. 2014: T. Stöllner, B. Craddock, I. Gambaschidze, G. Gogotchuri, A. Hauptmann, A. Hornschuch, F. Klein, I. Löffler, G. Mindiashwili, B. Murwanidze, S. Senczek, M. Schaich, G. Steffens, K. Tamasashvili, S. Timberlake (with an appendix of M. Jansen, T. Stöllner, A. Courcier), “Gold in the Caucasus: new research on gold extraction in the Kura-Araxes Culture of the 4th millennium BC and early 3rd millennium BC”, in H. Meller, E. Pernicka, R. Risch (dir.), Metalle der Macht. Frühes Gold und Silber, published in Tagungen des Landesmuseums für Vorgeschichte Halle 11, 2014, pp. 71‑110.

Stöllner, Gambashidze 2011: T. Stöllner, I. Gambashidze, “Gold in Georgia II. The oldest gold mine in the world”, in Ü. Yalçın (dir.), Anatolian Metal V, published in Der Anschnitt. Beiheft 24, 2011, pp. 187‑199.

Stöllner, Gambashidze 2014: T. Stöllner, I. Gambashidze, “The gold mine of Sakdrisi and the earliest mining and metallurgy in the Transcaucasia and the Kura-Valley system”, in G. Narimanishvili, M. Kvachadze, M. Puturidze, N. Shanshashvili (dir.), Problems of Early Metal Age archaeology of Caucasus and Anatolia, Tbilisi, 2014, pp. 102‑124.

Tylecote 1987: R.F. Tylecote, The early history of metallurgy in Europe, London, Longman Archaeology Series, 1987.

Notes

1 Grain mills are still missing as a category of artefact at the crafts-shop part of Dzedzvebi settlement (area II.2-3, II.7-8). If this picture may be confirmed by the archaeobotanical data, we need the results from the final examination of soil samples from that area: charred wheat for instance has not been discovered yet.

2 The single operation steps are currently studied by various colleagues: While F. Klein (MA) is investigating the mining and refilling processes at the Sakdrisi mine, K. Tamazashvili (MA) studies the tool sets in Sakdrisi and Dzedzvebi by including use-wear analyses. The metallurgical evidence at Dzedzvebi had been studied by Dr A. Courcier, M. Jansen (MA), T. Skowronek (BA), and the author. Gratefully remembered is Prof. Dr A. Hauptmann (Bochum), who provided us with various help and advice.

3 Pers. comm. from Dr I. Gambashidze and Dr G. Mindiashvili.

4 The ceramic vessel from the Kura-Araxes period found in Sakdrisi comprises only fragments of big storage vessels and handled beakers and mugs but no bowls and other vessel types common in the settlement of Dzedzvebi, although shapes and forms from both sites are clearly interrelated (Otchvani et al. 2015). It is likely that the Sakdrisi pottery was used basically as a drinking set that was practically interrelated with partly decorated storage vessels. While we can assume that normal drinking water was carried to the mines in leather sleeves, such drinking mugs might have been used for a more special purpose and beverage (see also the mapping in Otchvani et al., this volume, fig. 8).

5 All the collective burials investigated so far contained male and female adults but also one juvenile. This also points to a peculiar ritual status of juveniles, at least of some of them. Thanks to Assistant Professor Dr Lars Fehren-Schmitz (UC Santa Cruz, California, USA), who carried out the physical anthropology and provided us with information and the data (Stöllner 2017, tab. 2).

6 As the pit was heavily destroyed and overlaid by the ceramic deposition and its cultural layer, no original sediment was detected. So we were not able to take original soil samples.

7 No crucible fragments were found and a fired hearth near the panning installation was heavily destroyed by the den.

8 Email from A. Sagona, Melbourne, 15/07/2016. Thanks to Tony, we are able to mention this important find that is related to grave 7 in workshop house 2 in Dzedzvebi II.2, where two saddle querns were discovered on top of the juvenile burial 7 nearby the smelting hearth of the workshop.

9 Although the data and detailed data processing will be published by N. Boenke and W.H. Schoch elsewhere we found it useful to at least present some of the archaeobotanical data to contribute to the general discussion.

Table des illustrations

Titre Fig. 1 – Rise and decline of the ritual display of social inequality in Transcaucasia (DBM, RUB, T. Stöllner [graphics]; Lyonnet et al. 2010; Gambashidze et al. 2010; Gambashidze et al. 2001 [O. Lordkipanidze]; T. Rabsilber, DBM [photos/drawings]).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-1.jpg
Fichier image/jpeg, 448k
Titre Fig. 2 – The plateau of Balitchi-Dzedzvebi and its Kura-Araxes settlements, workshops and graves in relation to the gold-mining site at the Kachagiani hillock/Sakdrisi (ArcTron/DBM, F. Klein).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-2.jpg
Fichier image/jpeg, 849k
Titre Fig. 3 – Dzedzvebi, III.5, map of the gold-grinding and smelting site, sections through the working and smelting pits (DBM, I. Luther, F. Klein).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-3.jpg
Fichier image/jpeg, 895k
Titre Fig. 4 – Dzedzvebi, area II.8, house 4, map of upper level with ceramic/stone deposits and of lower level with gold-panning installation, central pit and pigment deposit (GNM; DBM/RUB, F. Klein).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-4.jpg
Fichier image/jpeg, 660k
Titre Fig. 5 – Dzedzvebi, area II.8, house 4; left: gold-panning installation; right: ceramic pan from gold-panning installation (GNM; DBM/RUB, H.-J. Laufer, T. Sachvadze, T. Stöllner).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-5.jpg
Fichier image/jpeg, 619k
Titre Fig. 6 – 1: Dzedzvebi III.5, 18185, crucible fragment, photomicrograph of slag, sample no. 4036/15, taken under plane polarized light, magnification×2.5, the slag contains magnetite clusters precipitated into undetermined siliceous matrix and small metal prills in copper and in arsenical copper spread out everywhere in the slag; 2: 18185, crucible fragment, macro-scale cross section views of smelting crucible from Dzedzvebi, sample no. 4036/15, a large black mineral inclusion of anorthite/augite is noticed on the right cross section, the thin grey bullous slag is characterized by cuprite zones, red, and small malachite fragments, green; 3: SEM‑EDS compositional data for the slag and the metal prills of crucible fragment, sample no. 4036/15, results normalized and expressed in weight percentage (DBM, A. Courcier, T. Skowronek).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-6.jpg
Fichier image/jpeg, 656k
Titre Fig. 7 – The “Paravani” calculation shows the time-consumption for producing 1 g of gold in the ore-deposit of Sakdrisi in regard to the technical processes reconstructed by excavations and experiments and the Paravani earring (after Stöllner et al. 2014).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-7.jpg
Fichier image/jpeg, 348k
Titre Fig. 8 – Sakdrisi, deposits of hammer-stones and an antler fragment underground; 1: hammerstones found in the NW niche of mine 2, 25011 (DBM/RUB, T. Rabsilber); 2: find situation of deposit 25011 (DBM/RUB, T. Stöllner); 3‑7: find situations of hammer tool deposits at mine 1/3, feature 23005 (7) in mine 1/2 north extension, features 24069/5, 24022 (3‑4), of an antler (5) and in the northernmost gallery (6) (DBM/RUB, T. Stöllner, P. Thomas).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-8.jpg
Fichier image/jpeg, 1,7M
Titre Fig. 9 – The Sakdrisi-Dzedzvebi complex near Kazreti-Balitchi; 1: Dzedzvebi area II.2, house 2, workshop, vertical photography; 2: Dzedzvebi area II.2, grave 2, 3D SFM model of the skull and vessel deposition; 3: Dzedzvebi area II.2, house 1 and 2 plus workshop, drawing (GNM; DBM/RUB, T. Rabsilber, F. Klein [3-D models], T. Sachvadze [map], T. Stöllner).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-9.jpg
Fichier image/jpeg, 780k
Titre Fig. 10 – Dzedzvebi, area II.8, house 4; 1: map of upper level with ceramic/stone deposits; 2: ritual deposit of vessel with red pigment and two ceramic stands beside; 3: artefacts of the ritual pigment deposit; 4: decorated plaster fragment 25521, found with burned clay debris on top of the vessel (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [maps], T. Sachvadze [drawings], T. Stöllner [photos of features]).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-10.jpg
Fichier image/jpeg, 913k
Titre Fig. 11 – 1: Dzedzvebi area II.7, pit 37006, photograph; 2: Dzedzvebi area II.7, sections through pit 37006; 3: pottery and stone tool from the pit (GNM; DBM/RUB, H.‑J. Laufer [artefacts], F. Klein [drawings], T. Stöllner [photo]).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-11.jpg
Fichier image/jpeg, 679k
Titre Fig. 12 – Dzedzvebi, area II.7, house 5; 1: map of stone pebble layer with central hearth and grave 7 sidewards; 2: central hearth, section and drawing; 3: view from north to the 2015 excavation of “house 5” and grave 9 in the back part with ceramic/stone deposits (GNM; DBM/RUB, F. Klein [map], T. Sachvadze [drawings], T. Stöllner [photos of feature]).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-12.jpg
Fichier image/jpeg, 328k
Titre Fig. 13 – Chobareti, Javaketi region, burial 11; 1: upper level with saddle quern and grinder; 2: lower level with burial; 3: ingot found on top nearby the saddle quern (Chobareti expedition, by courtesy of the late Prof. Dr A. Sagona, Melbourne).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-13.jpg
Fichier image/jpeg, 1,0M
Titre Fig. 14 – Sakdrisi mine, wood species of charcoal samples (after W.H. Schoch, Birmensdorf).
URL http://books.openedition.org/momeditions/docannexe/image/12477/img-14.jpg
Fichier image/jpeg, 255k

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search