Mesopotamian" Haematite" Seals in a new Light

Proceedings of the 6th International Congress on the Archaeology of the Ancient Near East May, 5th-10th 2008, “Sapienza” - Università di Roma Volume 1 Near Eastern Archaeology in the Past, Present and Future. Heritage and Identity Ethnoarchaeological and Interdisciplinary Approach, Results and Perspectives Visual Expression and Craft Production in the Definition of Social Relations and Status Edited by Paolo Matthiae, Frances Pinnock, Lorenzo Nigro and Nicolò Marchetti with the collaboration of Licia Romano 2010 Harrassowitz Verlag · Wiesbaden MESOPOTAMIAN “HAEMATITE” SEALS IN A NEW LIGHT MARTINE M. DE VRIES-MELEIN, DIRK VISSER, JUDITH J. MULDER, LUC MEGENS, SILVIA IMBERTI, WINFRIED KOCKELMANN, HENK KARS ABSTRACT A database of iron oxide stone Mesopotamian seals and weights is under construction to obtain temporal and spatial patterns of use. Linking the occurrence of iron oxide stone deposits in the Near East to ancient transport limitations may narrow down the number of possible mining positions in use for the production of iron stone objects. A non-destructive route for provenancing the iron oxide materials is being developed. In this paper a preliminary report is given on the database, material sourcing survey and the results of XRF and neutron diffraction on 13 Mesopotamian seals from the De Liagre-Böhl collection (NINO, Leiden). It is concluded that the labeling as “haematite” for a Mesopotamian seal is too simple. The iron oxide stones may consist of varying compositions of basically three different iron oxide compounds: haematite, magnetite and goethite, while Fe coated calcite may also pass as an iron oxide stone object. INTRODUCTION Iron oxide stone was a popular material for seals, weights and jewelry during the Old Babylonian Period, c. 2000-1600 BC. Its appearance in the material record is as sudden as its disappearance about 400 years later. For seals ‘haematite’ seems to replace formerly popular materials such as carnelian and lapis lazuli.1 One of the tantalizing questions therefore concerns the origin(s) of the iron oxide stone. It can be speculated that its sudden popularity was caused by changes in the accessibility of deposits.2 To test this hypothesis it is essential to investigate the temporal and spatial patterns of the use of iron oxide stone artifacts, to assess the geographical origin of the raw material, and to identify discerning aspects in mineralogy and chemical composition. In this study various iron oxide artifacts from Mesopotamia were investigated according to the following lines of research: − Investigation of the temporal and spatial patterns of use through the 1 2 Collon 1986: 11. Collon 1986; Moorey 1994; Potts 1997. Martine M. de Vries-Melein et al. 220 composition and analysis of a database of iron oxide stone artifacts from Mesopotamia. − Locating the most probable region(s) of origin by mapping natural occurrences of iron oxide stone in the regions around Mesopotamia, and combining these data with historical, economical, technological and cultural information. − Identifying different material groups and gaining insight in the geogenesis of the stone by determining the mineralogy and chemical composition of artifacts and raw material. This paper presents the preliminary outcome of the database analysis, the results from a literature study of the geological sources of iron oxide stones in the Near East and the non-destructive structural and elemental analysis of 13 Mesopotamian seals. BACKGROUND A Database of Iron Oxide Stone Artifacts At present, the database consists of descriptions of nearly 1000 artifacts. The descriptions of the artifacts are taken from excavation reports, museum catalogues, seal catalogues and books and articles on weights. From the work of D. Collon and others it was evident that the Mesopotamian seal repertoire shows a dominant presence of “haematite” seals during the Old Babylonian Period. In the British Museum Collection the iron oxide stone seals make up almost 70% of all seals from this period, while before c. 2000 BC and after c. 1600 BC less than 5% of the seals were made of iron oxide stone.3 This pattern is confirmed by our database; of the 824 dated seals in the database, 786 can be dated to c. 2000-1600 BC, see Table 1. Period Before 2300 BC 2300 - 2100 BC 2100 – 2000 BC 2000 – 1800 BC 1800 – 1600 BC 1600 – 1200 BC After 700 BC all Total 1 3 9 469 317 42 1 824 Found in situ 1 54 165 7 1 228 Table 1: Iron oxide stone seals sorted by period. 3 Collon 1986. Not found in situ 1 3 8 415 152 35 0 614 Mesopotamian “Haematite” Seals in a New Light 221 Not just the seals show this pattern. Although “haematite” is often thought of as the pre-eminent material for Mesopotamian weights, the weights follow the same temporal spread: iron oxide stone weights occur predominantly in the Old Babylonian Period. Of the 132 weights in the database, 110 can be dated. 87 of them are dated to c. 2000-1600 BC, i.e. 79% (table 2). It should be noted that there are still a lot of descriptions of weights waiting to be added to the database. A more definite pattern is bound to emerge. Period No date 2300 - 2100 BC 2000 – 1800 BC 1800 – 1600 BC 1600 – 1200 BC 1200 – 700 BC After 700 BC all Total 22 1 21 66 20 2 132 Found in situ 13 21 66 20 1 121 Not found in situ 9 1 1 11 Table 2: Iron oxide stone weights sorted by period. On the general spatial and temporal pattern the following can be said: the earliest iron oxide stone artifacts date from the mid-third millennium BC, and do not come from excavations. During the first centuries of the second millennium BC there is a tremendous increase of the use of iron oxide stones. Artifacts are found in all categories of spatial contexts. Many seals were not found in regular excavations, but can be firmly dated to the Old Babylonian Period on the basis of their iconography. During the second half of the second millennium iron oxide stone artifacts disappear from the archaeological record. Excavated finds from these later periods come mainly from the South of Mesopotamia (Ur, Uruk), and from burial contexts. With every addition to the database, the percentages change. For the seals at least, the temporal pattern seems to be consistent. More data from excavations in Northern Mesopotamia will contribute to a more complete picture. The Geological Data A literature survey of geological deposits of iron oxide stones in the Near East has been conducted. Although Anatolia is generally suggested as place of origin of the raw iron oxide stones,4 deposits occur all around Mesopotamia. Deposits of iron stone 4 E.g. Collon 1986. Martine M. de Vries-Melein et al. 222 materials such as haematite, goethite and magnetite, as found in literature, have been geographically mapped, as well as deposits of iron oxides and iron. These deposits have then been reviewed according to their geological context. Rocks are divided into three groups according to their formation process: igneous, metamorphic and sedimentary rocks. Iron oxides occur in all of these three groups (table 3). Mainly iron oxide stones of igneous or metamorphic formation are suitable for making artifacts. Sedimentary material is not compacted enough, contains too many inclusions and does not form in large enough deposits. Using general geological maps, deposits in areas of sedimentary origin were not considered. Haematite Goethite Magnetite Colour Metallic grey to earthy red Metallic grey to dark brown Metallic grey to black Streak Bright red to dark red Brown, brownish yellow to orange yellow Black Chemical formula Fe2O3 FeO(OH) Fe3O4 Hardness (Mohs’ Scale) 6.5 5-5.5 5.5-6 Density (g/cm3) 5.2 4.3 4.9 Formation in magmatic rocks Yes, more in intrusive than in extrusive formations. No Yes, in Precambrian shields. Often titanomagnetite. Formation in metamorphic rocks Yes, in metamorphosed iron- No formations, metabasites with low grade metamorphism, aerobic clay rocks, metamorphosed rocks containing manganese and BIFs. Often Mg, Mn or Ti in the haematite. Yes, in metaperidotites, metabasites, ironformations, gneisses, metapelites and BIFs. Formation in sedimentary rocks Yes, in red beds, BIFs, ironstones, skarns and margins of basins. Yes, in soils, BIFs, skarns and ironstones. Yes, in soils, ironstones in coastal zones and iron oxide veins. Note: BIF: Banded Iron Formation. Table 3: Characteristics and formation types of haematite, goethite and magnetite. Apart from the geological factors mentioned above, logistic factors, such as distance from a source and available ways of transport, should be considered. For this reason any deposits in the Caucasus, Turkmenistan, Northern Pakistan and Eastern India were rejected as being too far from Mesopotamia. The locations of the remaining deposits from the literature study are displayed in Fig. 1. Mesopotamian “Haematite” Seals in a New Light 223 From this map it is clear that iron oxide stones are common in all regions around Mesopotamia. Most sedimentary material is not compact enough, contains too many inclusions and does not form in large enough deposits. An exception to this is iron oxide precipitated into veins, which form in sedimentary deposits. The Chemical and Mineralogical Analyses In order to ensure that these valuable objects are left fully intact, a non-destructive analytical approach for the elemental and structural characterization of the 13 Mesopotamian objects has been chosen XRF X-ray Fluorescence (XRF) measurements on opposite sides of each seal have been carried out at the ICN, Amsterdam (NL) by means of a hand-held XRF apparatus (a Bruker Tracer III-V handheld X-ray fluorescence spectrometer with a Rhodium tube operating at a tube voltage of 40 KeV and a tube current of 2.2 μA). The analysis was performed in air which prevented the detection of elements lighter than potassium (K). As expected, the data, which are only qualitative, show for all but one of the seals that Fe is the major component. However, surprisingly in seal no 55 the major component is Ca. Figure 2 displays three seals from the De Liagre-Böhl collection: no. 55, no. 45 and no. 75. No clear distinct color difference is visible. The minority elements are given in Table 4. Object nr. Main element Minor and trace elements DLB 45 Fe Mn, V DLB 46 Fe Mn, V DLB 47 Fe Mn, V DLB 55 Ca Fe, Sr DLB 57 Fe K DLB 59 Fe Mn,V DLB 61 Fe DLB 62 Fe As, Mo DLB 66 Fe Mn, Au DLB 67 Fe As DLB 75 Fe Cr DLB 84 Fe Mn, As, Sr DLB 118 Fe Cr, Sr Table 4: Elements detected in the Mesopotamian ‘Haematite’ seals from the De Liagre-Böhl collection. One observes the usual first row transition metal elements, often associated with Martine M. de Vries-Melein et al. 224 occurrence in a mineral iron ore: Mn, V, Cr. Sr, As and K are most likely connected to small inclusion minerals in the objects. The occurrence of Au remains unexplained. The minor and trace elements present in seal 55 are different compared to the others, as expected for different mineral composition of this object. The hand-held XRF spectra of the seals no. 55, no. 45 and no. 75 are given in Figure 3 a-c. It should be noted that the obtained information from XRF only provides information on the elements present in the upper few μm of the surface of the seals. In fig. 3a, the spectrum of seal 45 (haematite) is dominated by the Fe XRF spectrum. The two major Fe intensities are indicated as well as the V peak at 4.95 kV and Mn as shoulder to the first large iron peak. The not identified peaks in the spectra of fig. 3 a-c (from left to right with increasing energy) are due to the edge effect, the escape, the 2x Raleigh scattering and 2x sum peaks of Fe. A clean spectrum of Fe is also displayed in the goethite seal 75, with an additional Cr peak at 5.4 kV. In the calcite Seal 55, a comparable spectrum of Ca is observed. The small not indexed peaks in the background are due to similar processes as described above for Fe. A small amount of Fe and Sr is also indicated. Neutron Diffraction The structure and phase composition of the 13 seals of the De Liagre-Böhl collection were determined from time-of-flight neutron diffraction data taken at the ROTAX (nos. 118, 68) and INES diffractometers at the ISIS spallation neutron source of the Rutherford Appleton Laboratory, Chilton Didcot (UK).5 For the analysis the object was placed in the neutron beam. The size of the beam was larger than the sample itself, hence the data provide average bulk information on the whole seal. The averaged neutron diffraction pattern of the bulk material of all seals could be refined by means of the Rietveld method for powder samples.6 The diffraction pattern of object no 55, 45 and 75 are given in Figure 4a-c. It is obvious from the diffraction patterns that the seals no 55, 45 and 75 consist of different mineral phases. The diffraction pattern of seal no. 55 can be fitted to the mineral calcite and a few percent of quartz. The diffraction pattern of seal no. 45 is explained by the mineral haematite: Fe2O3 and seal no 75 by the mineral goethite: FeOOH. It can be mentioned that the increased background of the goethite seal (no. 75) is due to the incoherent scattering of neutrons by the hydrogen atoms in the goethite, a typical observation for hydrogen containing samples. The mineral phase compositions of the 13 Mesopotamian seals are given in Table 5. 5 6 Imberti et al. 2008; Kockelmann et al. 2001. Young 1993. Mesopotamian “Haematite” Seals in a New Light 225 Object no. Composition Unknown phases DLB 62 haematite (92.5wt%)+ magnetite (0.5wt%)+ quartz (7wt%) DLB 84 haematite (100wt%) DLB 75 goethite (100wt%) DLB 67 magnetite (62.7wt%)+ haematite (37.3wt%) DLB 66 haematite (100wt%) DLB 61 goethite (100wt%) DLB 59 haematite (99.1wt%) +quartz (0.9wt%) +X DLB 57 haematite (94.1wt%) +magnetite (5.9wt%) +XX DLB 55 calcite (97.3wt%) + quartz (2.7wt%) +X DLB 47 haematite (100wt%) DLB 46 haematite (100wt%) DLB 45 haematite (100wt%) DLB 118 goethite (93.2 wt%) + haematite (3.3wt%) + bornite (3.5wt%) +X X and XX: a few unknown reflections from not yet identified phases. Table 5: Mineral phases of the seals. The diffraction results show that the iron ore stone seals which are normally labeled in literature on Mesopotamian artifacts as ‘haematite’ is more varied. Small quantities of different minerals are present too. The presence of the inclusions may provide a route to provenance the seals. A summery of the iron ore minerals found in the De Liagre-Böhl collection is given in table 6. Number of seals Composition DLB 5 haematite (100wt%) 45, 46, 47, 66, 84 2 goethite 61, 75 1 goethite + haematite + bornite 118 2 haematite + magnetite 57, 67 1 haematite + quartz 59 1 haematite + magnetite +quartz 62 1 calcite + quartz 55 Martine M. de Vries-Melein et al. 226 Table 6: Distribution of the different types of iron stones. The three major iron ores, haematite, goethite and magnetite are present in different seals. Haematite and magnetite may occur next to each other in the seals while goethite seals are of single phase. In seal no 55 the calcite particles must have a substantial Fe coating in order to have an appearance as black as the haematite objects, an assumption which is supported by the XRF data. It can be mentioned that the hematite and magnetite phases produce magnetic Bragg peaks in the diffraction patterns, which originate from neutron scattering by the antiferromagnetic and ferromagnetic structures of the two iron oxides, respectively. Goethite on the other hand is non-magnetic at room temperature. The observation of these magnetic peaks will be discussed elsewhere. DISCUSSION The currently available information is insufficient to pinpoint the origin(s) of the raw iron oxide stone that was so popular in Mesopotamia in the Old Babylonian Period. The extensive geological research supports the insight that the occurrence of natural deposits of iron oxide stones is not restricted to Anatolia. The available literature will be searched to map the geology on a more local scale, and bring more detail in our understanding of the geogeneses at the various locations. The database is a powerful tool in discerning the spatial and temporal patterns of use of the artifacts, and already yields valuable information. More entries will make it even more representative of all artifact categories. Both the archaeological and the geological information need to be firmly embedded in the historical, political and economical background, which has been extensively described in several recent publications.7 The structural phase analysis of the Mesopotamian seals shows the presence of the three types of iron stone: haematite, magnetite and goethite, while calcite may also have a black appearance. The fact that the neutron diffraction data can be analyzed as a powder averaged sample indicates the presence of fine grains in the seal objects. Some minority phases like quartz are also present. The XRF data only indicate next to Fe, as main constituent of the seals, the presence of some first–row transition elements, especially the occurrence of V in 4 of the 6 haematite seals stands out. The presence of As and Sr indicates the presence of further inclusion material on a level that can not be determined from the ND data. Quantitative XRF data will be required to obtain information on the relevance of the minority elements in the Mesopotamian seals. CONCLUSION A temporal and spatial pattern emerges from the data in the database: whereas earlier 7 E.g. Charpin, Edzard, Stol 2004. Mesopotamian “Haematite” Seals in a New Light 227 finds come from the whole of Mesopotamia, later finds, after the middle of the second millennium BC, come mainly from the South. The material is more generally used in the north and west during its heydays, which would suggest a provenance north or west from Mesopotamia. The geological data show that there is no reason to restrict the quest for the provenance to Anatolia: natural deposits occur all around Mesopotamia. In order to address questions about provenance and to explain the temporary popularity of the iron oxide stones the material analyses will have to be extended to a larger number of artifacts. The preliminary results of the present study already indicate that the materials are well suited for non-destructive bulk analyses techniques. Acknowledgements We would like to thank the Jo Kolk Stichting for the grant which made this research possible, the NINO for the permission to measure the seals, and Theo Krispijn for his never-ending support for the project. Bibliography Collon, D. 1986 Catalogue of the Western Asiatic Seals in the British Museum, vol. III, London. Cornell, R.M., Schwertmann, U. 2003 The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, Weinheim. Charpin, D., Edzard, D. O., Stol, M. 2004 Die altbabylonische Zeit, Göttingen. Imberti, S. et al. 2008 Neutron Diffractometer INES for Quantitative Phase Analysis of Archaeological Objects: in Meas. Sci. Technol. 19 034003, pp. 1-8. Kockelmann, W. et al. 2001 Non-Destructive Phase Analysis of Archaeological Ceramics Using TOF Neutron Diffraction: in Journal of Archaeological Science 28, pp. 213222. van Loon, M. 1968 First Results of the 1967 Excavations at Tell Selenkahiye: in Annales Archéologiques Arabes Syriennes 18, pp. 21-32. van Loon, M. (ed.) 2001 Selenkahiye, Final Report on the University of Chicago and University of Amsterdam Excavations in the Tabqa Reservoir, Northern Syria 1967-1975, Istanbul. 228 Martine M. de Vries-Melein et al. Moorey, P.R.S. 1994 Ancient Mesopotamian Materials and Industries: the Archaeological Evidence, Oxford. Otto, A. 2006 Alltag und Gesellschaft zur Spätbronzezeit: eine Fallstudie aus Tall Bazi (Syrien), Turnhout. Potts, D.T. 1997 Mesopotamian Civilization: the Material Foundations, London. Young, R.A. (ed.) 1993 The Rietveld Method, Oxford. Mesopotamian “Haematite” Seals in a New Light 229 Fig. 1: Deposits of iron oxides, haematite, goethite and magnetite. 230 Fig. 2a-c: DLB seals 45 (a), 55 (b) and 75 (c). Martine M. de Vries-Melein et al. Mesopotamian “Haematite” Seals in a New Light Fig. 3. a-c XRF spectra of the seals n. 45, n. 55 and n. 75. Fig. 4 a-c: Neutron diffraction patterns of the seals n. 45, n. 55 and n. 75. 231