Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Material description of a unique relief fibula from Poland

Material description of a unique relief fibula from Poland A unique relief fibula dated to the Migration Period (first half of the sixth century) was found in Radziejów, Poland. This stray find changes previous opinions on the lack of settlement in central Poland at that time. As the find is the only one of such type in Poland, a special attention was paid to possible analogies, mainly finds from Scandinavia and Western Europe. The fibula underwent technological analyses in order to reveal its technology of manufacture. For this purpose, the chemical and elemental composition of the alloy was studied. Several physico-chemical complementary techniques such as optical microscopy (OM), scanning electron microscopy with energy dispersive X-ray (SEM–EDX) spectroscopy, energy dispersive X-ray fluorescence (ED-XRF) analysis, X-ray diffraction (XRD) analysis and micro-hardness testing (HV0.2) were used to study the technology of the find. The investigations revealed that this artefact was made from brass (4–17 wt% Zn) with an admixture of Sn (2–12 wt%). Two technologies were used: casting for the bow and forming for the part with the axle of the spring. The artefact’s surface was tinned in the hot-dipping process. The physical structural analyses demonstrated that the artefact was cast and ornamented by surface stamping under a relatively low temperature (about 500 °C). . . . . . . Keywords Relief fibula Migration Period Scandinavian metal handicraft Ancient metallurgy SEM–EDX ED-XRF XRD Introduction every respect a unique find. Due to this, apart from a ‘tradition- al’ archaeological discussion on the artefact, we also decided to Apart from traditional typochronological or stylistic examina- carry out a series of technological examinations of the find with tions of archaeological artefacts, the role of technological anal- the use of different analytical techniques. Another reason is that yses of finds has been more and more appreciated by archaeol- the number of technological analyses of analogous artefacts is ogists. Such analyses can often provide crucially important data still far from satisfactory. Therefore, each analysed find may on the technology of manufacture or provenance of raw mate- provide valuable data for future comparative examinations. rials. They can also supply otherwise unavailable pieces of in- formation on the chronology of artefacts. A Migration Period Circumstances of discovery relief fibula which was discovered in Radziejów, Poland, is in The relief fibula was found in 1990 in the town of Radziejów, located in the region of Cuiavia. The discovery was made during a historical battle re-enactment. The artefact was sim- * Ewelina A. Miśta Ewelina.Mista@ncbj.gov.pl ply kicked from the ground by one of the participants in the event. The fibula, being a stray find with no archaeological context, was located in the topsoil layer. The discovery was National Centre for Nuclear Research (NCNR), Świerk, Otwock, Poland made at the border of a wood in the area enclosed between Sportowa and Armii Krajowej Streets (Fig. 1). Institute of Electronic Materials, Warsaw, Poland Institute of Archaeology, University of Warsaw, Warsaw, Poland Archaeological background Faculty of Materials Science and Engineering, Warsaw University of Technology, Warsaw, Poland Settlement in the vicinity of Radziejów has recently been Institute of History, Jan Długosz University in Częstochowa, Częstochowa, Poland discussed by Kontny (2016). The entire Cuiavian settlement 974 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 1 a The town of Radziejów in the map of Poland. b Contour map of the town with the place of discovery cluster which survived unexpectedly in the Migration Period decoration in Animal Style I after Salin (1904). Highly at least until the mid seventh century is a matter of intensive stylised flat images of parts of animal and human bodies are studies by a team of scholars—participants in a multidisciplin- filled with parallel convex strokes. It is possible to recognise a ary research project ‘Migration Period between the Odra and bird’s head, an animal paw and multiple elements of bodies, Vistula’ (http://www.mpov.uw.edu.pl/pl/). Its unique character possibly also heads (double ovals symbolising eyes with up- is related to the fact that according to previous opinions, a right hair) and head fragments with eyes. Unfortunately, a lot major part of the territory of today’s Poland (except for its of motifs seem to be unidentifiable. Generally, the composi- north-east part) was almost totally abandoned in the beginning tion is asymmetrical but some motifs are repeated symmetri- of the Migration Period. New discoveries, including the relief cally, i.e. in similar places of both images (humanoid heads, fibula from Radziejów, change this image drastically. eyes, arrangement of body fragments; see Fig. 3). The bow of The relief fibula weighs 51 g with metal alloy density over the fibula is distinguished at the edges and along the central 7.5 g/cm . The artefact survived in a good condition. Due to axis with elongated laths; the surfaces confined with the laths this, we were able to study the shapes of the relief ornamen- are filled with short horizontal convex lines. In the lower part tation (see Fig. 3). The dimensions of the surviving fragments of the bow, there is a representation of a pair of eyes. The are as follows: length—7cm, widthofthe head-plate— above motif, i.e. a centrally incised beak, is continued in the 5.5 cm, length of the head-plate—2.8 cm, width of the upper part of the foot-plate. Only the upper part of the foot- bow—1.8 cm, height—2.5 cm, height of the catch-plate— plate has survived. It bears an image of a pair of animal heads 1 cm and length of the catch-plate—1.7 cm. with the necks outlined as a contour, filled with parallel The relief brooch survived only in parts (Fig. 2): two frag- strokes. Their eyes are oval and are underlined with grooves. ments have been glued together, but still a large part of the A simplified image of ears is also visible. There are tongues foot-plate is missing. The head-plate is rectangular with a which slide out of the muzzles and are outcurved backwards. distinct ornamental area of two parts bordered from the top The heads of the beasts are preserved partially. Between the and sides with a line of arches. Their arms are directed inwards heads, there are short convex horizontal lines and a centrally and there are triangles inside them. Both decorative zones are situated fragment of an unrecognisable motif (a zigzag bor- separated with double grooves. Inside the zones, there is a dering the decorated surface, the image of an eye?). The Fig. 2 Relief fibula from the wood in Radziejów: top view (obverse) and bottom view (reverse) (photo, M. Jakubowski) Archaeol Anthropol Sci (2019) 11:973–983 975 Fig. 3 Fibula in different projections with reliefs decorating the obverse. Interpretation is offered below (drawing, T. Jakubowski) original shape of the foot-plate is impossible to reconstruct. A good analogy is offered by a fibula from Darmstadt- An axle was fixed in the arched plate protruding perpendicu- Windmühle, Ldkr. Darmstand-Dieburg in Hessen. The arte- larly from the lower side of the head-plate. Only vestigial fact is provided with the same so-called Zangenmuster bor- fragments of the axle and a spring remained stuck in the fixing dering and an almost identical composition of the head-plate plate. The catch-plate is solid; it was cast together with the decoration. Even the eye pattern on the bow is placed in the foot-plate and the entire brooch and was then formed (small same position. After the devastation that had occurred during traces of plastic forming are noticeable on the bottom). This World War II, the artefact was reconstructed on the basis of the was confirmed by technological examinations (see below). gypsum mould that survived in the Römisch-Germanisches Originally, the discussed artefact must have been almost Zentralmuseum in Mainz (Salin 1904: p.295, Fig. 636; twice as long as the remaining part, which situates it in a group Haselhoff 1981:p.474–477, 479, Figs. 306, 307). Originally, of medium-sized fibulae, i.e. 10–13 cm, according to the idea it was 11.6 cm long, so being very much like the find from of Høilund Nielsen (2009: p. 61). The fibula is not worn out, Radziejów. It is believed to be a continental copy of south so probably it was not used for long. Scandinavian origin, made by an artisan in the Middle Rhine The fibula represents Type A-2 after Sjøvold (1993), area, partly without understanding the concept of the image, characterised by a rectangular head-plate and an elongated flat which is especially noticeable in the head-plate animal orna- foot-plate (non-profiled longitudinally). Theoretically, it could mentation (Haselhoff 1981: p.474, 476–477, 479; Høilund be attributed to variant f (ended with an animal mask) or d Nielsen 2009:p.70–71). There are minor differences: instead (without a mask). However, due to its incomplete state of of heads with bristled hair, there are animal paws, the motif in preservation, its precise attribution is not possible (see the upper right corner of the right ornamental zone seems to be Sjøvold 1993: p. 38, pl. 17). As regards the motif of the eyes simpler and some fragments of bodies are replaced by limbs. on the bow and the beak on the upper part of the foot-plate, Additionally, there is no representation of a bird’shead. one may recall the brooch from Vik (II), Fjære, Agder as an However, despite these small discrepancies, it seems obvious analogy, although not very adequate (Sjøvold 1993: p. 62, pl. that both fibulae were made in the same workshop (despite the 17:N19). It is considered a simplification of patterns which lack of metallographic analyses of the missing fibula). were identified on the fibula from Gyland, Bakke, Agder In addition to metallurgical examinations, a quite important (Sjøvold 1993: p. 62, pl. 17:N20, with further literature); how- field to study is the stylistic analysis of the ornamentation of ever, the latter lacks the mentioned motif. Moreover, a simi- the Radziejów fibula. Undoubtedly, the decoration of the larly shaped image of eyes (but doubled, i.e. with a pair of discussed find was arranged according to principles of eyes similarly placed in the upper part of the bow) can be Animal Style I after Salin (1904), more accurately to its found on the brooches with oversimplified ornament (al- Phase B after Haseloff (1981: p.180–196). Phase B finds are though still in Style I, Phase B) which can be seen on characterised by low relief, clear accentuation of contours and Thuringian finds from Mühlhausen, Görmarche Landstraβe, covering the major surfaces with animal decoration. The bod- Ldkr. Unstrut-Hainich-Kreis, grave 2 (Schmidt 1976:pl. ies depicted in outlines are filled with raised stripes (Haselhoff 114:c) and Beuchte, Ldkr. Wolfenbüttel, grave 1 (Schmidt 1981: p.180). This is clearly visible also on the fibula from 1961:Fig. 10:f; Høilund Nielsen 2003: p.213,Fig. 8:26e). Radziejów. Geographically, artefacts representing Phase B It should be noted here that finding an ideal analogy is gener- with its striped animal bodies are mostly found in the Baltic ally impossible. This is due to the fact that relief brooches, Sea zone (Haselhoff 1981: p. 707). However, they also ap- being an element of female elite clothing, had individualised peared as imports in the Gepidic areas or the Lombard territory decorations and morphology. These traits, however, were in in Pannonia, coming there from the north across the Odra and agreement with general rules of a given style (see Høilund Vistula watersheds (Haselhoff 1981: p. 673–674). If we at- Nielsen 2009:p.57). tempt at identifying the fibula from Radziejów more precisely, 976 Archaeol Anthropol Sci (2019) 11:973–983 it should be assigned to Continental fibulae, Group Kirchheim relief fibulae. Schmidt attributed them to his Group IIIb, (besides the general composition, the bristled hair is typical of which falls to AD 560–600 (Schmidt 1961: Fig. 49); however, the group in question). Such fibulae appeared mainly in me- such dating seems dubious in particular cases and should be dium (like the one from Radziejów) but also in small sizes (see shifted for earlier stages (Høilund Nielsen 2009:table 7–8). Høilund Nielsen 2009:p.70–71, Fig. 13). The same identifi- The ‘Lombard Phase’ finds from Bohemia are attributed to the cation refers to a fibula from Darmstadt-Windmühle (Høilund period of AD 530–560/70 (Droberjar 2009:Fig 4). The Italian Nielsen 2009: p.70, table 7). As to the bordering, we do not brooch from Nocera Umbra is dated to the earliest Stage 1 know any exact analogies to the Radziejów fibula, apart from (AD 572–590) of the Lombard stay in Italy. Together with the find from Darmstadt-Windmühle. It looks like a simplified single costume elements made in Style I, they are considered version of the claw-frieze (Germ. Zangenfries or products of Pannonian workshops from the period before the Zangenmuster), Type B2 (Haselhoff 1981: p. 314–318), which end of the Lombards’ migration (see Bierbrauer 2005: Fig. 3; suggests a certain, but not very strong, simplification or 2008: 113, Fig. 2). Taking into consideration the chronology stylisation. Another example of such stylisation of the above of Group Kirchheim, it should be remembered that there are pattern, although appearing in a different motif, was studied by no exact dates for its production (the Kirchheim grave itself Hillberg (2009:p.182–186) who discussed it at against the must be a very late deposition from AD 580–600, see Høilund central-, north- and west-European background. Nielsen 2009: 86, table 7). However, on the basis of stylistic Relief fibulae served as an element of a female dress. It is traits, one is tempted to link it with Stage A after Høilund believed that they were status symbols of female sorceresses Nielsen, i.e. 460/80–510/25 AD. Copies (made in the or priestesses. Such fibulae were worn across the upper part of Middle Rhine area) come from Stage B, i.e. AD 510/25– the bosom which can be seen, e.g. at the image on a pendant 565, and further imitations stem from a not datable context from Aska in Sweden (Olsen 2006:Fig. 7). It could also be (Høilund Nielsen 2009:p.82–84). Therefore, the dating of the concluded from the position of similar fibulae in graves, e.g. fibula from Radziejów should be specified as the first half of Sejlflod, Aalborg Komm., Region Nordjylland, grave DI the sixth century, but having in mind the fact that it is not worn (Jørgensen,Nørgard Jørgensen 1997: Fig. 46:g). Such fibulae out, an earlier stage of this period is probable. were not only artefacts of practical function but were also We should underline that the artefact in question most ornamental pieces of jewellery of symbolic value. As a prob- probably belongs to imported fibulae, perhaps from able attribute of goddess Freya, they may have been worn Scandinavia or maybe the Middle Rhine area. This is due to during sacrifices carried out by noble ladies from the top of the fact that it is not as sophisticated as many (but not all) finds the society, in the course of erotic rituals or while interpreting from the north. It has some stylistic analogies in Thuringia, the omens (Olsen 2006:p.496–497, 503). e.g. the eyes pattern on the bow and the general shape of biting As regards their chronology, relief fibulae have been in- beasts on the foot-plate. On the other hand, the former trait is cluded in Type E1 after Høilund Nielsen, 1997,76),charac- also proved for Scandinavia, while the latter seems to be con- teristic of Phase 1A for Bornholm materials. In most cases, nected with the process of overall simplification of the original they appear in pairs in rich female graves (Høilund Nielsen motif. Copy-level fibulae are also known from Scandinavia. 1987: p. 70–72, Fig. 10). Their absolute chronology generally The bordering looks like a simplified version of a claw-frieze. covers the first half of the sixth century, though a precise There is one very close analogy for this in the Middle Rhine distinction between Phases 1A and 1B (lasting until ca. AD area and it is considered a copy of Scandinavian origin. 600) seems to be problematic (Høilund Nielsen, 1997:p.69, However, it is not entirely clear whether we are really dealing Fig. 18). Bitner-Wróblewska places them in Phases 3 and 4 with a Merovingian copy or a copy-level fibula made in south according to her idea, i.e. a late stage of Phase D and Phase E. Scandinavia. The latter identification is supported by the fact This generally means the second half of the fifth century and that fibulae of that and larger size (over 14 cm) are mostly the first half of the sixth century. This stands in agreement typical of southern Scandinavia (Høilund Nielsen 2009:p.61, with earlier Scandinavian perspectives (Bitner-Wróblewska Fig. 5). It seems slightly more probable that the discussed 2001: p.120, pl. II, LIX). For instance, they are attributed to artefact came here from Scandinavia rather than from a more Gotlandic Phase VI:2 after Nerman (Nerman 1935: p.64, pl. distant place, i.e. the Middle Rhine region. However, the prob- p.32–34), i.e. ca. AD 475/500–550 (Nerman 1935: p.119– lem cannot be solved at the present stage of research. 121; 1969; 1975; see Bitner-Wróblewska 2001:p.16). Relief So far, it is the first relief fibula found in the territory of cast square-headed fibulae from England were also produced Poland. We know only one fragment of gilded foot of a relief between AD 500 and 575, with a decline in numbers after AD fibula from Brzostowo, Piła district. This artefact is either a 550 (Olsen 2006:p.479; Hines 1997:p.301;Høilund Nielsen Scandinavian original or an import of simplified form which 1997:p.79–80, Fig. 28). However, we have to remember that was manufactured in the Merovingian cultural circle the latest ones were embellished in Animal Style II. The finds (Rudnicki 2014,p.287–288, Fig. 6:1). It seems interesting that from central Germany could be later than the Scandinavian no finds of relief brooches are known among numerous Archaeol Anthropol Sci (2019) 11:973–983 977 Fig. 4 Samples A and B with their sampling areas in the fibula imported fibulae from the West Balt Circle, including the Scanning electron microscopy with energy dispersive X- Elbląg Group and the Olsztyn Group (Hillberg 2009:p.188). ray (SEM–EDX) analyses (Goldstein et al., 2007,Barbacki This is in spite of the fact that Scandinavian fibulae and belt 2007) of sample cuts were performed using a Carl Zeiss EVO fittings are proved here, namely, in the Elbląg Group but also MA10 scanning electron microscope equipped with a EDAX in the Sambian-Natangian area (Kontny 2012:p.61–65). X-Flash Detector 5010 with 123 eV spectral resolution (Zeiss, Furthermore, Animal Style I is also represented in the Poland; www.zeiss.com). Image analyses were made using a Olsztyn Group and in the Sambian-Natangian area (Hillberg secondary electron (SE) detector with resolution up to 2.0 nm. 2009:p.190–195). Accelerating voltage of 20 keV was applied; other character- Concluding, the fibula is a unique discovery in Poland and istics of current, field magnifying and type of sampling were it comes from a period which is still poorly studied for this adopted to the type of studied surface area. Spectra were reg- country. Due to the discussed characteristics of the artefact, it istered by an energy dispersive spectrometer that collects the was decided to carry out a series of technological examina- entire spectrum of X-ray from 0 to 20 eV, with resolution of tions. In the future, their results will allow for a more accurate 123 eV. Then, average concentrations of elements were ob- provenance study of fibulae of such type. tained. The measurement time was set to 120 s for multi-point analysis and to 5 min for mapping. Measurements were car- ried out for eight areas of a few micron size. Concentrations of copper (Cu), tin (Sn), zinc (Zn), iron (Fe), lead (Pb) and other Methods metallic and light elements were studied using SEM–EDX examinations. Prior to taking samples, the SEM–EDX analy- Instrumentation ses of the entire artefact were also performed. However, due to the surface contamination by soil ingredients like silicon (Si), Preliminary structural observations were carried out with a aluminium (Al), oxygen (O), calcium (Ca) and other light Nikon SMZ 800 optical microscope and with a Nikon elements, as well as due to a low depth range (up to 2 μm) Ephiphot 200 inverted metallographic microscope at ×50–500 of electrons of excitation source, the analyses did not allow to magnifications. Two samples (see below) were taken and their obtain reliable information about the raw alloy composition. surfaces were etched with 1 part of perchloric acid HClO and 2 An X-ray generator designed and built in the NCNR in parts of acetic acid anhydride C H O solution for 10 s. Świerk, Poland, was used for energy dispersive X-ray 4 6 3 Fig. 5 Sample A. OM images of metal alloy structure obtained after etching at a, b mag × 10 and c mag × 20. There is an internal core dendritic microstructure (two-phase multiple components system) with a porous corrosion layer up to 100 μm thick in the surface (a) 978 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 6 Sample B. OM images of metal alloy structure obtained after etching at a mag × 2.5, b mag × 5 and c mag × 20. One-phase system and an external porous corrosion structure (up to 50 μm thick) can be seen fluorescence (ED-XRF) measurements. This compact X- Description of the samples ray generator (weight 0.835 kg, length 430 mm) was con- structed with the use of transmission X-ray tube technol- Complementary instrumental physico-chemical techniques ogy. A focused electron beam bombards a silver (Ag) were used to examine the material structure (for a similar anode target generating X-ray radiation. The silver anode methodology, see, e.g. Miazga 2014,Miśta et al. 2015 & has been sputtered on beryllium. The maximum acceler- Miśta et al., 2016, Meeks 1993, Ashkenazi et al. 2015, ating voltage is 50 kV. The beam current and accelerating Giumlia-Mair 2005) of this unique fibula. The analyses voltage can be smoothly adjusted. were performed on the unchanged artefact and on the cuts The XRF spectrum was induced by X-ray irradiation on a of two samples. Figure 4 shows the cuts with areas from spot with a diameter of 2.1 cm. Such a large analysing area where the samples were taken. The samples were mounted yielded well-averaged values of the elemental composition. in epoxy resin. They were ground using Premium SiC The X-ray generator was operated with an anode voltage abrasive paper (for 5 s with P220, P500, P800 and P120 and current of 40 kV and 30 μA, respectively. Fluorescence type: http://www.metallographic.com/Brochures/SiCpaper. spectra of studied objects were recorded with an Amptek Si- pdf; accessed on 8 November 2017) and polished using PIN detector. The energy resolution of the detector was as- Al O powder for 2 min. The surfaces of the prepared 2 3 sumed to be 149 eV at 5.9 keV photon energy. The measured cuts were etched in 1 HClO /2C H O solution for 10 s. 4 4 6 3 energy resolution was found to be 152.9 ± 0.3 eV. The dis- Elemental and phase components of the prepared samples tance between the X-ray source and measured objects (SO) which correspond to the artefact’s raw alloy without and between objects and the detector (OD) was set to 9 and corrosion, degraded surface and subsurface structure were 11.5 cm, respectively. Energy spectra were analysed with an determined. Furthermore, the cuts correspond to the parts Amptek ADMCA multi-channel analyser. Data on contents of of the fibula which were manufactured separately, i.e. the copper (Cu), tin (Sn), iron (Fe), lead (Pb), zinc (Zn), silver cross-section of the bow (sample A) and part of the axle of (Ag) and gold (Au) was obtained. Moreover, μ-XRF analyses the spring (sample B). of elemental composition were carried out on sample cuts Optical microscopy (OM) images of the cuts’ surfaces (sample A and sample B, see below). show the crystallographic metal alloy system with corrosion XRD method was used for phase identification. Powder X- structures on the surface layer. In Fig. 5, there are results of ray diffraction measurements were performed with Cu Kα microscopic examinations. radiation (λ = 1,5418 Å) using a Siemens D500 diffractometer Scanning electron microscopy (SEM) revealed information equipped with a Si:Li semiconductor detector cooled with about the surface condition (see Fig. 7). Moreover, using EDX liquid nitrogen and a ICDD PDF4 2014 database. The powder spectrometry allows to determine elemental composition in diffraction pattern was measured in θ/2θ scanning mode with the cuts. XRD and HV0.2 analyses were also performed for a step of 0.02 and integration time of 10 s/step. sample A and sample B. XRF analysis was carried out in Vickers micro-hardness (HV0.2) was measured on micro scale. polished surfaces of the cuts, using a 200-g load applied X-ray fluorescence spectroscopy (XRF) does not require for 12 s. At least 10 measurements per data point were sample preparation (see Gójska & Miśta 2016,Miśta & made. A Zwick Roell ZHU0.2/Z2.5 machine was used. Gójska 2016). The ED-XRF measurements were done on The tests were carried out according to the ASTM E384 the entire artefact in order to determine the elemental compo- standard with a constant distance of 2.5 diameter of a sition of the external alloy layer (up to ca. 20 μm: Henke et al. stamp between every stamp. 1993) by recording signals from components. Archaeol Anthropol Sci (2019) 11:973–983 979 Fig. 7 SEM–SE images of a fragment of the higher disc fragment (obverse side = obv.). Surface relief ornament, soil contamination and corrosive degradation can be seen Results and discussion the methods are based on measurements of X-ray spectra. The characteristic energy of X-ray corresponds to a given A preliminary SEM–EDX study of the fibula shows its sur- element in the sample. The difference between these methods face degradation which was invisible to the naked eye. The consists in another type of X-ray excitation: for EDX, it is results (Fig. 7) demonstrated that it was necessary to work electron beam and for XRF, it is X-ray from an X-ray tube. with pure metal alloy whose structure corresponds to the raw The range of penetration depth for electrons and X-ray is material. Below are examples of SEM images obtained for the different so we can study surface and subsurface layers fibula surface with relief ornaments (obverse side). Irregular (Miśta et al. 2016). This penetration depth is highly dependent cloudy shapes correspond to the corrosion and soil contami- on the material density. Furthermore, the areas of sampling are nation of the structure which is mainly built of light element also different. For ED-XRF, it is a field with a diameter of oxides. The external surface is also rich in iron and copper 2.1 cm, while for EDX and μ-XRF, it is a field within a oxides (up to ca. 50 wt% in total). micrometre area. The obtained numerical data were Table 1 shows the results of the quantitative elemental recalculated and are offered in Table 1 as normalised weight analysis carried out with the use of different methods. All percent taking into account the alloying elements. Table 1 Results of the SEM–EDX, ED-XRF and μ-XRF quantitative elemental composition analyses (wt%) (low detection limit LLD = 0.1 wt%) Technique Sampling No. Cu Sn Zn Pb Ag Fe For the entire artefact: ED-XRF Obv. 1 59.59 ± 5.96 22.46 ± 2.24 10.28 ± 1.03 4.34 ± 0.43 1.58 ± 0.16 1.39 ± 0.14 Rev. 2 52.67 ± 5.23 27.09 ± 2.71 9.36 ± 0.94 5.38 ± 0.54 1.84 ± 0.18 1.63 ± 0.16 For the cuts of samples A and B: μ-XRF A 1 71.23 ± 7.12 10.19 ± 1.01 18.58 ± 1.86 < LLD < LLD < LLD B 2 90.09 ± 9.00 2.85 ± 0.28 4.18 ± 0.42 1.54 ± 0.15 0.60 ± 0.06 0.79 ± 0.08 SEM–EDX A 1 69.95 ± 1.94 12.34 ± 0.41 16.98 ± 0.53 0.60 ± 0.10 0.13 ± 0.04 < LLD 2 66.94 ± 1.85 12.57 ± 0.41 15.96 ± 0.50 4.40 ± 0.27 < LLD 0.12 ± 0.04 3 71.89 ± 1.99 9.83 ± 0.33 18.18 ± 0.56 0.11 ± 0.06 < LLD < LLD 4 67.25 ± 1.86 17.80 ± 0.56 14.62 ± 0.46 0.33 ± 0.08 < LLD < LLD 5 71.75 ± 1.96 9.43 ± 0.31 18.58 ± 0.56 0.24 ± 0.06 < LLD < LLD Av. 69.95 ± 1.06 12.34 ± 1.49 16.98 ± 0.73 0.33 ± 0.82 < LLD < LLD B 6 93.72 ± 2.56 2.13 ± 0.10 4.15 ± 0.17 < LLD < LLD < LLD 7 93.45 ± 2.55 2.09 ± 0.10 4.46 ± 0.18 < LLD < LLD < LLD 8 92.61 ± 2.50 2.03 ± 0.09 4.59 ± 0.16 0.63 ± 0.07 < LLD 0.15 ± 0.03 Av. 93.45 ± 0.34 2.09 ± 0.03 4.46 ± 0.13 < LLD < LLD < LLD 980 Archaeol Anthropol Sci (2019) 11:973–983 Due to a considerable degree of the artefact surface con- tamination and degradation, it was necessary to perform EDX and (for the sake of comparison) μ-XRF analyses on cross- section cuts (samples A and B). Figure 8 shows a juxtaposi- tion of the SEM–EDX mapping analyses performed for sam- ple A. The maps demonstrate the homogeneity of distribution of the main determined elements: copper (Cu), zinc (Zn) and tin (Sn) in the alloy core of sample A. The results of the quantitative elemental composition anal- yses of samples A and B are given in Table 1 while Fig. 9 offers an example of registered EDX spectra. The elemental composition of sample A corresponds to the alloy composition of the raw material of the fibula bow, as shown in Table 1. The main elements determined by SEM– Fig. 9 EDX spectra obtained for the cross-section cuts of the fibula: EDX are (in descending order) copper (av. 69.95 ± 1.06 wt%), sample A (blue, no. 2 EDX in Table 1) and sample B (red, no. 6 EDX zinc (av. 16.98 ± 0.73 wt%) and tin (av. 12.34 ± 1.49 wt%), in Table 1) while lead is present as trace element (av.0.33 ± 0.82 wt%). The results are confirmed by the μ-XRF analysis of sample (0.60 ± 0.06 wt%) and iron (0.79 ± 0.08 wt%) in the alloy. A(Table 1). This elemental composition corresponds to the bi- Due to the fact that these elements were determined by ED- XRF measurements at a significant weight percent level on the phase of casting brass with a significant addition of tin. Structures of this type can be seen in the OM images in artefact surface (see Table 1: ED-XRF results and Fig. 12), Fig. 5 and in the diffractogram in Fig. 10 (see below) their trace amounts in the inner alloy can be linked with partial (Czermiński 1978: p.375, Dobrzański 2002: p. 719). melting from the surface tinning coat (this issue is discussed Addition of tin (at the level of 1–2%) prevents zinc escape below). Therefore, with regard to the contents of the main from the alloy due to its easy evaporation above 907 °C elements, the composition of sample B corresponds to the (Chodkowski 1976: p. 340), while the addition of lead (1.5– composition of brass dedicated for a technique of plastic met- 4%) improves the machinability of the brass alloy (Dobrzański alworking called die forming (Czermiński 1978: p.375, 2002: p. 719). However, in this case, the addition of tin is Dobrzański 2002: p. 719). This type of brass is generally made significant (at about 12 wt%). This leads to the inference that of two components with small admixtures with a mono-phase it was introduced to the alloy in order to provide it with a gold crystallographic system (see Fig. 10). X-ray diffraction analysis (XRD) of metal alloy determined colour. This is also the case with the so-called Mannheim gold used in jewellery production. In its basic form, the alloy con- the crystallographic system of samples A and B (Fig. 10). It sists of 80% Cu and 20% Zn but also occurs with addition of corresponds to the elemental composition of the matrix alloy tin at the level of 9% (http://www.gutenberg.czyz.org/ of both parts of the artefact (see Table 1) which represents the word,44746, accessed on 8 November 2017). technological elements (see Fig. 4). Generally, in both In sample B, which is part of the axle of the spring, the diffractograms, there are visible reflections at approximately main elements are copper (av. 93.45 ± 0.34 wt%), zinc (av. 42° and 50°. These can be identified as fcc (face-centred cubic 4.46 ± 0.13) and tin (av. 2.09 ± 0.03 wt%) with trace amounts cell) solid solution, with a (Cu, Zn) and/or (Cu, Sn) structure, of other elements being below 0.1 wt% (in average value of respectively. It follows from the elemental analysis (see the EDX measurement done for n = 3 micro-sampling points). Table 1) and from the increased value of the lattice parameter However, results of μ-XRF analyses (see Table 1) indicate the with respect to pure copper. For sample A, the obtained value was a = 3.655 Å and for sample B, it was a = 3.642 Å, while presence of trace amounts of lead (1.54 ± 0.15 wt%), silver Fig. 8 a SEM-SE image (mag × 379) of the surface of the cross-section cut (sample A) and b–d (mag × 379) EDX maps with a regular distribution of the main constituting elements: b copper—Cu, c tin—Sn and d zinc—Zn in the alloy Archaeol Anthropol Sci (2019) 11:973–983 981 the relief ornament on the surface and forming the crumple zone like the curvature in the fibula catch-plate (samplings 4 and 5 in sample A) (Ashkenazi & Iddan 2012, Moriera and Tschiptschin, 2016 in Fig. 5). Summing up, brass with addition of tin and trace amount of lead (sample A) is used as casting alloy, while brass for forming (sample B) is more robust and has better malleable properties (Wesołowski, 1972: p.396–430, Doomke 1982). The selection of alloys demonstrates an intentional use of technological and functional criteria. The bow was made of soft alloy to make it easier to shape the artefact’s form and the relief ornament. On the other hand, moving parts like the spring and pin, which had to be durable, were made of harder metal. Such features are confirmed by the average values of Fig. 10 XRD diffractograms of samples A and B micro-hardness which are shown precisely for each measure- ment point in Fig. 11. The micro-hardness is 160 ± 2 HV0.2 standard lattice constant for metallic copper is a =3.615 Å. for sample A and 166 ± 2 HV0.2 for sample B. These differ- This is due to larger radii of tin (Sn) and zinc (Zn) than copper ences in micro-hardness are the results of variations of propor- (Cu), so their participation in the structure expands the crystal tions of copper, zinc, tin and lead (see. Table 1). lattice of pure copper. Additionally, in sample A, there are A quite homogenous crystallographic structure and uni- several crystallographic phases. This is due to the fact that form values of HV0.2 parameter for sample A and sample B the fcc structure contains an intermetallic Cu5.6Sn phase (or demonstrate that each part, that is the bow and moving parts a similar one, having a ‘deformed’ fcc crystal lattice), which is used for attaching the fibula to the garment was made sepa- observed in the dendrite structure in Fig. 5.In sample B,its rately. The bow was made by casting in a mould after melting, structure is quite uniform and more fine crystalline in compar- and the axle of the spring was made by die forming in a ison with sample A, as it can be seen in optical images in temperature range of 200–800 °C. In the case of Cu–Sn, this Fig. 6. The ZnO reflections visible in smaller sample B as corresponds to dendritic structure (see phase diagram of Cu– bigger intensity originate from zinc corrosion. Sn by Johnson in Dobrzański 2002: p.729,Fig. 7.15).The The low HV0.2 values indicate that the melting process Cu–Zn system is homogeneous in the range up to 40% Zn in a was carried at a not very high temperature. temperature range up to the melting point of pure copper No layer dislocation can be seen in OM photos (Figs. 5 and 6), (phase diagram of Cu–Zn by Howkins in Dobrzański 2002: which suggests the use of a method of mould casting without p. 719, Fig. 7.13, Wesołowski 1974: p. 404). treatment at a higher temperature. Below, Fig. 12 offers ED-XRF qualitative analysis results Generally, the HV0.2 values decrease in the subsurface obtained for the artefact’s surface, for the obverse and reverse area. This corresponds to weakening of the alloy structure side of the fibula, respectively. Table 1 offers the results of due to the annealing process. This process is necessary to form quantitative analysis. Fig. 11 Micro-hardness test results for sample A and sample B, measurement error ± 2HV0.2. Stamp places in the samples can be seen on the left 982 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 12 ED-XRF spectra obtained for the fibula, for the obverse side with the relief ornament (blue) and for the reverse side (red), respectively In comparison with the base alloys identified in samples A determined as a brass with different contents of tin, zinc and and B, at both sides of the surface (to the depth 20 μm: Henke trace amounts of lead. The elements were mixed depending on et al., 1993), there is a noticeable tin (Sn) enrichment. Tin con- the functional requirements of each part of the artefact. Casting centration in the surface is above 20 wt%, while in the base brass was used in the manufacture of the bow. On the other alloy in samples A and B, it is at the level of 12 and 2 wt%, hand, forming brass as a base alloy was applied for the axle respectively (see Table 1). Additionally, in the surface, there are of the spring, which was a moving part of the fibula. The HV0.2 higher contents of silver (Ag) and lead (Pb). Furthermore, traces value is higher for mounting parts, that is for the pin and other of iron (Fe) are detected. A higher content of tin can indicate the parts of the attachment mechanism (sample B). The relatively use of hot-dipping tinning in order to decorate the surface of the low HV0.2 value and the presence of fcc network for Cu/Zn fibula (e.g. Miazga 2014, Šmitetal. 2008, Meeks 1993, and Cu/Sn crystallographic system (which corresponds to the Giumlia-Mair 2005). The addition of silver and lead can orig- identification of the elemental composition) indicates that the inate from silver ores (galena PbS) which could be added to the casting process was performed in a not very high temperature. tin solution used for coating the artefact by dipping it into the The presence of the zones with lower HV0.2 values proves the molten tin solution. On the other hand, the presence of iron use of the annealing process for surface ornament forming. should rather be related to soil contamination of the surface. Acknowledgments We are indebted to Mr. Tadeusz Sworobowicz and Mr. Janusz Bojarczuk for technical support in the course of the analyses. Thanks must also go to Mr. Leszek Kalicki who found the artefact and handed it over to the State Archaeological Museum in Warsaw. Conclusions Open Access This article is distributed under the terms of the Creative This archaeological examination and the technological analy- Commons Attribution 4.0 International License (http:// ses of the unique fibula found in Radziejów in Poland can be a creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appro- background for similar studies of other artefacts mentioned in priate credit to the original author(s) and the source, provide a link to the the introduction. In the future, such studies can shed more Creative Commons license, and indicate if changes were made. light on the presence and transformation of fibulae of such types in the Baltic Sea region. Furthermore, the study of the References elemental composition of the artefact, combined with physical micro-hardness tests and crystallographic examinations, Ashkenazi D, Taxel I, Tal O (2015) Archaeometallurgical characteriza- allowed for an identification of the technological process tion of Late Roman- and Byzantine- period Samaritan magical ob- jects and jewelry made of copper alloys. Mater Charact 102:195– which was used in the manufacture of artefacts of this type. 208. https://doi.org/10.1016/j.matchar.2015.01.019 The most important conclusion is that the fibula from Ashkenazi D, Iddan N (2012) Archaeometallurgical characterization of Radziejów was tinned, as evidenced by the XRF analysis of Hellenistic metal objects: the contribution of the bronze objects from the obverse and reverse of the find. Usually, fibulae of this type Rishon Le-Zion (Israel). Archaeometry 54(3):528–548. https://doi. org/10.1111/j.1475-4754.2011.00631.x were gilded in the discussed period. The raw alloy was Archaeol Anthropol Sci (2019) 11:973–983 983 Barbacki A (2007) Mikroskopia elektronowa. Wydawnictwo Politechniki Kontny B (2012) Trade, salt and amber. The formation of late Migration Poznańskiej, Poznań Period elites in the ‘Balti-Culti’ area of northern Poland (the Elbląg Bierbrauer V (2005) Archäologie der Langobarden in Italien: etnische group). In Bliujienė A (ed) People at the Crossroads of Space and Intepretation und Stand der Forschung. In Pohl W & Erhart P (ed.) Time (Footmarks of Societies in Ancient Europe). Archaeologia Die Langobarden. Herrschaft und Identität, Forschungen zur Baltica 17 (1), pp 60–76 Geschichte des Mittelalters 9, Wien, pp 21–66 Kontny B (2016) Unexpected relic—the unique relief brooch from Bitner-Wróblewska A (2001) From Samland to Rogaland. East-West Radziejów, decorated in Salin’s Animal Style I, BŚwiatowit N.S.^ connections in the Baltic basin during the Early Migration Period, IX (2015)/B Warszawa Meeks N (1993) Surface characterization of tinned bronze, high-tin Chodkowski J (1976) Małysłownik chemiczny. 5th edn. Wiedza bronze, tinned iron and arsenical bronze. In: Craddock P, La-Niece Powszechna, Warszawa S (eds) Metal plating and patination: cultural, technical and histori- Czermiński J (1978) Metalurgia. Encyklopedia Techniki. Wyd. Śląsk, cal developments. Butterworth-Heinemann, UK, pp 247–275. Kraków https://doi.org/10.1016/B978-0-7506-1611-9.50025-X Dobrzański LA (2002) Podstawy nauki o materiałach i metaloznawstwo. Miazga B (2014) Tin and tinned dress accessories from medieval Materiałyinżynierskie z podstawami projektowania materiałowego. Wrocław (SW Poland). X-ray fluorescence investigations. WNT, Warszawa Estonian J Archaeology 18(1):57–79. https://doi.org/10.3176/arch. Doomke W (in translation by) Jędrzejewski W & Jędrzejewski K (1982) 2014.1.03 Vademecum Materiałoznastwa,. 2nd edn., WNT Warszawa Miśta EA, Gójska A (2016) A metallographic analysis of copper alloy Droberjar E (2009) Některé problémy mladšidoby stěhováni národů v artefacts from the lake in Lubanowo. In: Nowakiewicz T (ed) Čechách. In Niezabitowska B, Juściński M, Łuczkiewicz P & Ancient sucrificial place in the lake in Lubanowo (former Herrn – Sadowski S (ed.) The Turbulent Epoch. New Materials from the See) in west Pomerania. Institute of Archaeology Warsaw Late Roman Period and the Migration Period, vol. I, Lublin, pp University, Warsaw, pp 213–225 133–148 Miśta EA, Gójska A, Kalbarczyk P, Bojarczuk J, Rzeszotarska- Giumlia-Mair A (2005) Tin rich layers on ancient copper based objects. Nowakiewicz A & Nowakiewicz T (2016) An analysis of the ele- Surface Engineering 21 no. 5-6:359–367 mental composition and a study of precious metal artefacts. In Gójska AM, Miśta EA (2016) Analysis of the elemental composition of Nowakiewicz T. & Rzeszotarska-Nowakiewicz A. (ed) Ancient the artefacts from the Kosewo archaeological site. Acta Phys Pol A sucryficial place in former Lake Nidajno in Masuria. Results of 130(6):1415–1419. 10.12693/APhysPolA.130.1415 laboratory analyses of selected finds, Institute of Archaeology Goldstein J, Newbury ED, Joy CD, Lyman EC, Echlin P, Lifshin E., Warsaw University, Warsaw, pp 17–66 Sawyer L & Michael, RJ (2007) Scanning electron microscopy Miśta EA, Stonert A, Korman A, Milczarek JJ, Fijał-Kirejczyk I, and X-ray microanalysis, 3rd edn. Springer, U.S.A. Kalbarczyk P, Wiśniewska W (2015) Material research on ar- Haselhof GH (1981) Die germanische Tierornamentik der chaeological objects using PIXE and other non-invasive tech- Völkerwanderungszeit. Studien zu Salin's Stil I, Berlin-New York. niques. Acta Physica Polonica A 128(5):815–817. 10.12693/ https://doi.org/10.1515/9783110884111 APhysPolA.128.815 Henke LB, Gullikson EM, Davis JC (1993) X-ray interactions: Moriera CV, Tschiptschin PA (2016) A dilatometric study of the influence photoabsorption, scattering, transmission and reflection at E=50- of residual iron content on the annealing behavior of cartridge brass. 30000 keV, Z=1-92. Atomic Data and Nuclear \Data Tables 54/2: Mater Res 19(2):483–489. https://doi.org/10.1590/1980-5373-MR- 181–342 2015-0597 Hillberg V (2009) Masurische Bügelfibeln. Daumen und Kellaren – Nerman B (1935) Die Völkerwanderungszeit Gotlands, Stockholm Tumiany i Kielary, Band 2, Schriften des Archäologischen Olsen VS (2006) The development of (proto) disc-on-bow brooches in Landesmuseums 9, Neumünster England, Frisia and Scandinavia. Palaeohistoria 47/48, 2005/2006: Hines J (1997) A New Corpus of Anglo-Saxon Great Square-Headed 479–528. http://rjh.ub.rug.nl/Palaeohistoria/search/search Brooches. Reports of the Research Committee of the Society of Rudnicki M (2014) Dwa znaleziska skandynawskich zapinek płytkowych Antiquaries of London 51, Woodbridge z terenów północnej Polski. Wiadomości Archeologiczne LXV: Høilund Nielsen K (1997) The schism of Anglo-Saxon chronology. In: 283–290. https://pbn.nauka.gov.pl/sednowebapp/works/594230 Jensen KC, Høilund Nielsen K (eds) Burial & society. The Salin B (1904) Die altgermanische Thierornamentik. Typologische Chronological and Social Analysis of Archaeological Burial Data, Studie über germanische Gegenstände aus dem IV. bis IX. Aarhus, pp 71–99 Jahrhundert, nebst einer Studie über irische Ornamentik, Stockholm Høilund Nielsen K (2003) Saxon art between interpretation and imitation: Schmidt B (1961) Die späte Völkerwanderungszeit in Mitteldeutschland. the influence of Roman, Scandinavian, Frankish, and Christian art Veröffentlichungen des Landesmuseums für Vorgeschichte in Halle on the material culture of the continental Saxons AD 400–1000. In 18, Halle (Saale) Green HD & Sigmund F (ed) The Continental Saxons from the Schmidt B (1976) Die späte Völkerwanderungszeit in Mitteldeutschland. Migration Period to the tenth century: an ethnographic perspective, Katalog (Nord- und Ostteil). Veröffentlichungen des San Marino, pp 193–233 Landesmuseums für Vorgeschichte in Halle 29, Berlin Høilund Nielsen K (2009) The real thing or just wannabes? Sjøvold T (1993) The Scandinavian relief brooches of the Migration Scandinavian-style brooches in the fifth and sixth centuries. In: Period. An Attempt at a new Classification. Norske oldfunn Quast D (ed) Foreigners in early medieval Europe. Thirteen interna- XV, Oslo tional studies on early medieval mobility, Monographien des Šmit Ž, Istenič J, Knific T (2008) Plating of archaeological metallic ob- Römisch-Germanischen Zentralmuseums 78, Mainz, pp 51–111 jects—studies by differential PIXE. Nuc. Inst. And Methods Physic Jørgensen L & Nørgård Jørgensen A (1997) Nørre Sandegård vest. A Res B 226:2329–2333 th th cemetery from the 6 –8 centuries on Bornholm, Nordiske Wesołowski K (1972) Metaloznawstwo i obróbka cieplna. WNT, Fortidsminder B/14, København Warszawa http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archaeological and Anthropological Sciences Springer Journals

Material description of a unique relief fibula from Poland

Download PDF
11 pages

Loading next page...
 
/lp/springer-journals/material-description-of-a-unique-relief-fibula-from-poland-9XVnOpJFha

References (42)

Publisher
Springer Journals
Copyright
Copyright © 2017 by The Author(s)
Subject
Earth Sciences; Earth Sciences, general; Archaeology; Chemistry/Food Science, general; Geography, general; Life Sciences, general; Anthropology
ISSN
1866-9557
eISSN
1866-9565
DOI
10.1007/s12520-017-0576-4
Publisher site
See Article on Publisher Site

Abstract

A unique relief fibula dated to the Migration Period (first half of the sixth century) was found in Radziejów, Poland. This stray find changes previous opinions on the lack of settlement in central Poland at that time. As the find is the only one of such type in Poland, a special attention was paid to possible analogies, mainly finds from Scandinavia and Western Europe. The fibula underwent technological analyses in order to reveal its technology of manufacture. For this purpose, the chemical and elemental composition of the alloy was studied. Several physico-chemical complementary techniques such as optical microscopy (OM), scanning electron microscopy with energy dispersive X-ray (SEM–EDX) spectroscopy, energy dispersive X-ray fluorescence (ED-XRF) analysis, X-ray diffraction (XRD) analysis and micro-hardness testing (HV0.2) were used to study the technology of the find. The investigations revealed that this artefact was made from brass (4–17 wt% Zn) with an admixture of Sn (2–12 wt%). Two technologies were used: casting for the bow and forming for the part with the axle of the spring. The artefact’s surface was tinned in the hot-dipping process. The physical structural analyses demonstrated that the artefact was cast and ornamented by surface stamping under a relatively low temperature (about 500 °C). . . . . . . Keywords Relief fibula Migration Period Scandinavian metal handicraft Ancient metallurgy SEM–EDX ED-XRF XRD Introduction every respect a unique find. Due to this, apart from a ‘tradition- al’ archaeological discussion on the artefact, we also decided to Apart from traditional typochronological or stylistic examina- carry out a series of technological examinations of the find with tions of archaeological artefacts, the role of technological anal- the use of different analytical techniques. Another reason is that yses of finds has been more and more appreciated by archaeol- the number of technological analyses of analogous artefacts is ogists. Such analyses can often provide crucially important data still far from satisfactory. Therefore, each analysed find may on the technology of manufacture or provenance of raw mate- provide valuable data for future comparative examinations. rials. They can also supply otherwise unavailable pieces of in- formation on the chronology of artefacts. A Migration Period Circumstances of discovery relief fibula which was discovered in Radziejów, Poland, is in The relief fibula was found in 1990 in the town of Radziejów, located in the region of Cuiavia. The discovery was made during a historical battle re-enactment. The artefact was sim- * Ewelina A. Miśta Ewelina.Mista@ncbj.gov.pl ply kicked from the ground by one of the participants in the event. The fibula, being a stray find with no archaeological context, was located in the topsoil layer. The discovery was National Centre for Nuclear Research (NCNR), Świerk, Otwock, Poland made at the border of a wood in the area enclosed between Sportowa and Armii Krajowej Streets (Fig. 1). Institute of Electronic Materials, Warsaw, Poland Institute of Archaeology, University of Warsaw, Warsaw, Poland Archaeological background Faculty of Materials Science and Engineering, Warsaw University of Technology, Warsaw, Poland Settlement in the vicinity of Radziejów has recently been Institute of History, Jan Długosz University in Częstochowa, Częstochowa, Poland discussed by Kontny (2016). The entire Cuiavian settlement 974 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 1 a The town of Radziejów in the map of Poland. b Contour map of the town with the place of discovery cluster which survived unexpectedly in the Migration Period decoration in Animal Style I after Salin (1904). Highly at least until the mid seventh century is a matter of intensive stylised flat images of parts of animal and human bodies are studies by a team of scholars—participants in a multidisciplin- filled with parallel convex strokes. It is possible to recognise a ary research project ‘Migration Period between the Odra and bird’s head, an animal paw and multiple elements of bodies, Vistula’ (http://www.mpov.uw.edu.pl/pl/). Its unique character possibly also heads (double ovals symbolising eyes with up- is related to the fact that according to previous opinions, a right hair) and head fragments with eyes. Unfortunately, a lot major part of the territory of today’s Poland (except for its of motifs seem to be unidentifiable. Generally, the composi- north-east part) was almost totally abandoned in the beginning tion is asymmetrical but some motifs are repeated symmetri- of the Migration Period. New discoveries, including the relief cally, i.e. in similar places of both images (humanoid heads, fibula from Radziejów, change this image drastically. eyes, arrangement of body fragments; see Fig. 3). The bow of The relief fibula weighs 51 g with metal alloy density over the fibula is distinguished at the edges and along the central 7.5 g/cm . The artefact survived in a good condition. Due to axis with elongated laths; the surfaces confined with the laths this, we were able to study the shapes of the relief ornamen- are filled with short horizontal convex lines. In the lower part tation (see Fig. 3). The dimensions of the surviving fragments of the bow, there is a representation of a pair of eyes. The are as follows: length—7cm, widthofthe head-plate— above motif, i.e. a centrally incised beak, is continued in the 5.5 cm, length of the head-plate—2.8 cm, width of the upper part of the foot-plate. Only the upper part of the foot- bow—1.8 cm, height—2.5 cm, height of the catch-plate— plate has survived. It bears an image of a pair of animal heads 1 cm and length of the catch-plate—1.7 cm. with the necks outlined as a contour, filled with parallel The relief brooch survived only in parts (Fig. 2): two frag- strokes. Their eyes are oval and are underlined with grooves. ments have been glued together, but still a large part of the A simplified image of ears is also visible. There are tongues foot-plate is missing. The head-plate is rectangular with a which slide out of the muzzles and are outcurved backwards. distinct ornamental area of two parts bordered from the top The heads of the beasts are preserved partially. Between the and sides with a line of arches. Their arms are directed inwards heads, there are short convex horizontal lines and a centrally and there are triangles inside them. Both decorative zones are situated fragment of an unrecognisable motif (a zigzag bor- separated with double grooves. Inside the zones, there is a dering the decorated surface, the image of an eye?). The Fig. 2 Relief fibula from the wood in Radziejów: top view (obverse) and bottom view (reverse) (photo, M. Jakubowski) Archaeol Anthropol Sci (2019) 11:973–983 975 Fig. 3 Fibula in different projections with reliefs decorating the obverse. Interpretation is offered below (drawing, T. Jakubowski) original shape of the foot-plate is impossible to reconstruct. A good analogy is offered by a fibula from Darmstadt- An axle was fixed in the arched plate protruding perpendicu- Windmühle, Ldkr. Darmstand-Dieburg in Hessen. The arte- larly from the lower side of the head-plate. Only vestigial fact is provided with the same so-called Zangenmuster bor- fragments of the axle and a spring remained stuck in the fixing dering and an almost identical composition of the head-plate plate. The catch-plate is solid; it was cast together with the decoration. Even the eye pattern on the bow is placed in the foot-plate and the entire brooch and was then formed (small same position. After the devastation that had occurred during traces of plastic forming are noticeable on the bottom). This World War II, the artefact was reconstructed on the basis of the was confirmed by technological examinations (see below). gypsum mould that survived in the Römisch-Germanisches Originally, the discussed artefact must have been almost Zentralmuseum in Mainz (Salin 1904: p.295, Fig. 636; twice as long as the remaining part, which situates it in a group Haselhoff 1981:p.474–477, 479, Figs. 306, 307). Originally, of medium-sized fibulae, i.e. 10–13 cm, according to the idea it was 11.6 cm long, so being very much like the find from of Høilund Nielsen (2009: p. 61). The fibula is not worn out, Radziejów. It is believed to be a continental copy of south so probably it was not used for long. Scandinavian origin, made by an artisan in the Middle Rhine The fibula represents Type A-2 after Sjøvold (1993), area, partly without understanding the concept of the image, characterised by a rectangular head-plate and an elongated flat which is especially noticeable in the head-plate animal orna- foot-plate (non-profiled longitudinally). Theoretically, it could mentation (Haselhoff 1981: p.474, 476–477, 479; Høilund be attributed to variant f (ended with an animal mask) or d Nielsen 2009:p.70–71). There are minor differences: instead (without a mask). However, due to its incomplete state of of heads with bristled hair, there are animal paws, the motif in preservation, its precise attribution is not possible (see the upper right corner of the right ornamental zone seems to be Sjøvold 1993: p. 38, pl. 17). As regards the motif of the eyes simpler and some fragments of bodies are replaced by limbs. on the bow and the beak on the upper part of the foot-plate, Additionally, there is no representation of a bird’shead. one may recall the brooch from Vik (II), Fjære, Agder as an However, despite these small discrepancies, it seems obvious analogy, although not very adequate (Sjøvold 1993: p. 62, pl. that both fibulae were made in the same workshop (despite the 17:N19). It is considered a simplification of patterns which lack of metallographic analyses of the missing fibula). were identified on the fibula from Gyland, Bakke, Agder In addition to metallurgical examinations, a quite important (Sjøvold 1993: p. 62, pl. 17:N20, with further literature); how- field to study is the stylistic analysis of the ornamentation of ever, the latter lacks the mentioned motif. Moreover, a simi- the Radziejów fibula. Undoubtedly, the decoration of the larly shaped image of eyes (but doubled, i.e. with a pair of discussed find was arranged according to principles of eyes similarly placed in the upper part of the bow) can be Animal Style I after Salin (1904), more accurately to its found on the brooches with oversimplified ornament (al- Phase B after Haseloff (1981: p.180–196). Phase B finds are though still in Style I, Phase B) which can be seen on characterised by low relief, clear accentuation of contours and Thuringian finds from Mühlhausen, Görmarche Landstraβe, covering the major surfaces with animal decoration. The bod- Ldkr. Unstrut-Hainich-Kreis, grave 2 (Schmidt 1976:pl. ies depicted in outlines are filled with raised stripes (Haselhoff 114:c) and Beuchte, Ldkr. Wolfenbüttel, grave 1 (Schmidt 1981: p.180). This is clearly visible also on the fibula from 1961:Fig. 10:f; Høilund Nielsen 2003: p.213,Fig. 8:26e). Radziejów. Geographically, artefacts representing Phase B It should be noted here that finding an ideal analogy is gener- with its striped animal bodies are mostly found in the Baltic ally impossible. This is due to the fact that relief brooches, Sea zone (Haselhoff 1981: p. 707). However, they also ap- being an element of female elite clothing, had individualised peared as imports in the Gepidic areas or the Lombard territory decorations and morphology. These traits, however, were in in Pannonia, coming there from the north across the Odra and agreement with general rules of a given style (see Høilund Vistula watersheds (Haselhoff 1981: p. 673–674). If we at- Nielsen 2009:p.57). tempt at identifying the fibula from Radziejów more precisely, 976 Archaeol Anthropol Sci (2019) 11:973–983 it should be assigned to Continental fibulae, Group Kirchheim relief fibulae. Schmidt attributed them to his Group IIIb, (besides the general composition, the bristled hair is typical of which falls to AD 560–600 (Schmidt 1961: Fig. 49); however, the group in question). Such fibulae appeared mainly in me- such dating seems dubious in particular cases and should be dium (like the one from Radziejów) but also in small sizes (see shifted for earlier stages (Høilund Nielsen 2009:table 7–8). Høilund Nielsen 2009:p.70–71, Fig. 13). The same identifi- The ‘Lombard Phase’ finds from Bohemia are attributed to the cation refers to a fibula from Darmstadt-Windmühle (Høilund period of AD 530–560/70 (Droberjar 2009:Fig 4). The Italian Nielsen 2009: p.70, table 7). As to the bordering, we do not brooch from Nocera Umbra is dated to the earliest Stage 1 know any exact analogies to the Radziejów fibula, apart from (AD 572–590) of the Lombard stay in Italy. Together with the find from Darmstadt-Windmühle. It looks like a simplified single costume elements made in Style I, they are considered version of the claw-frieze (Germ. Zangenfries or products of Pannonian workshops from the period before the Zangenmuster), Type B2 (Haselhoff 1981: p. 314–318), which end of the Lombards’ migration (see Bierbrauer 2005: Fig. 3; suggests a certain, but not very strong, simplification or 2008: 113, Fig. 2). Taking into consideration the chronology stylisation. Another example of such stylisation of the above of Group Kirchheim, it should be remembered that there are pattern, although appearing in a different motif, was studied by no exact dates for its production (the Kirchheim grave itself Hillberg (2009:p.182–186) who discussed it at against the must be a very late deposition from AD 580–600, see Høilund central-, north- and west-European background. Nielsen 2009: 86, table 7). However, on the basis of stylistic Relief fibulae served as an element of a female dress. It is traits, one is tempted to link it with Stage A after Høilund believed that they were status symbols of female sorceresses Nielsen, i.e. 460/80–510/25 AD. Copies (made in the or priestesses. Such fibulae were worn across the upper part of Middle Rhine area) come from Stage B, i.e. AD 510/25– the bosom which can be seen, e.g. at the image on a pendant 565, and further imitations stem from a not datable context from Aska in Sweden (Olsen 2006:Fig. 7). It could also be (Høilund Nielsen 2009:p.82–84). Therefore, the dating of the concluded from the position of similar fibulae in graves, e.g. fibula from Radziejów should be specified as the first half of Sejlflod, Aalborg Komm., Region Nordjylland, grave DI the sixth century, but having in mind the fact that it is not worn (Jørgensen,Nørgard Jørgensen 1997: Fig. 46:g). Such fibulae out, an earlier stage of this period is probable. were not only artefacts of practical function but were also We should underline that the artefact in question most ornamental pieces of jewellery of symbolic value. As a prob- probably belongs to imported fibulae, perhaps from able attribute of goddess Freya, they may have been worn Scandinavia or maybe the Middle Rhine area. This is due to during sacrifices carried out by noble ladies from the top of the fact that it is not as sophisticated as many (but not all) finds the society, in the course of erotic rituals or while interpreting from the north. It has some stylistic analogies in Thuringia, the omens (Olsen 2006:p.496–497, 503). e.g. the eyes pattern on the bow and the general shape of biting As regards their chronology, relief fibulae have been in- beasts on the foot-plate. On the other hand, the former trait is cluded in Type E1 after Høilund Nielsen, 1997,76),charac- also proved for Scandinavia, while the latter seems to be con- teristic of Phase 1A for Bornholm materials. In most cases, nected with the process of overall simplification of the original they appear in pairs in rich female graves (Høilund Nielsen motif. Copy-level fibulae are also known from Scandinavia. 1987: p. 70–72, Fig. 10). Their absolute chronology generally The bordering looks like a simplified version of a claw-frieze. covers the first half of the sixth century, though a precise There is one very close analogy for this in the Middle Rhine distinction between Phases 1A and 1B (lasting until ca. AD area and it is considered a copy of Scandinavian origin. 600) seems to be problematic (Høilund Nielsen, 1997:p.69, However, it is not entirely clear whether we are really dealing Fig. 18). Bitner-Wróblewska places them in Phases 3 and 4 with a Merovingian copy or a copy-level fibula made in south according to her idea, i.e. a late stage of Phase D and Phase E. Scandinavia. The latter identification is supported by the fact This generally means the second half of the fifth century and that fibulae of that and larger size (over 14 cm) are mostly the first half of the sixth century. This stands in agreement typical of southern Scandinavia (Høilund Nielsen 2009:p.61, with earlier Scandinavian perspectives (Bitner-Wróblewska Fig. 5). It seems slightly more probable that the discussed 2001: p.120, pl. II, LIX). For instance, they are attributed to artefact came here from Scandinavia rather than from a more Gotlandic Phase VI:2 after Nerman (Nerman 1935: p.64, pl. distant place, i.e. the Middle Rhine region. However, the prob- p.32–34), i.e. ca. AD 475/500–550 (Nerman 1935: p.119– lem cannot be solved at the present stage of research. 121; 1969; 1975; see Bitner-Wróblewska 2001:p.16). Relief So far, it is the first relief fibula found in the territory of cast square-headed fibulae from England were also produced Poland. We know only one fragment of gilded foot of a relief between AD 500 and 575, with a decline in numbers after AD fibula from Brzostowo, Piła district. This artefact is either a 550 (Olsen 2006:p.479; Hines 1997:p.301;Høilund Nielsen Scandinavian original or an import of simplified form which 1997:p.79–80, Fig. 28). However, we have to remember that was manufactured in the Merovingian cultural circle the latest ones were embellished in Animal Style II. The finds (Rudnicki 2014,p.287–288, Fig. 6:1). It seems interesting that from central Germany could be later than the Scandinavian no finds of relief brooches are known among numerous Archaeol Anthropol Sci (2019) 11:973–983 977 Fig. 4 Samples A and B with their sampling areas in the fibula imported fibulae from the West Balt Circle, including the Scanning electron microscopy with energy dispersive X- Elbląg Group and the Olsztyn Group (Hillberg 2009:p.188). ray (SEM–EDX) analyses (Goldstein et al., 2007,Barbacki This is in spite of the fact that Scandinavian fibulae and belt 2007) of sample cuts were performed using a Carl Zeiss EVO fittings are proved here, namely, in the Elbląg Group but also MA10 scanning electron microscope equipped with a EDAX in the Sambian-Natangian area (Kontny 2012:p.61–65). X-Flash Detector 5010 with 123 eV spectral resolution (Zeiss, Furthermore, Animal Style I is also represented in the Poland; www.zeiss.com). Image analyses were made using a Olsztyn Group and in the Sambian-Natangian area (Hillberg secondary electron (SE) detector with resolution up to 2.0 nm. 2009:p.190–195). Accelerating voltage of 20 keV was applied; other character- Concluding, the fibula is a unique discovery in Poland and istics of current, field magnifying and type of sampling were it comes from a period which is still poorly studied for this adopted to the type of studied surface area. Spectra were reg- country. Due to the discussed characteristics of the artefact, it istered by an energy dispersive spectrometer that collects the was decided to carry out a series of technological examina- entire spectrum of X-ray from 0 to 20 eV, with resolution of tions. In the future, their results will allow for a more accurate 123 eV. Then, average concentrations of elements were ob- provenance study of fibulae of such type. tained. The measurement time was set to 120 s for multi-point analysis and to 5 min for mapping. Measurements were car- ried out for eight areas of a few micron size. Concentrations of copper (Cu), tin (Sn), zinc (Zn), iron (Fe), lead (Pb) and other Methods metallic and light elements were studied using SEM–EDX examinations. Prior to taking samples, the SEM–EDX analy- Instrumentation ses of the entire artefact were also performed. However, due to the surface contamination by soil ingredients like silicon (Si), Preliminary structural observations were carried out with a aluminium (Al), oxygen (O), calcium (Ca) and other light Nikon SMZ 800 optical microscope and with a Nikon elements, as well as due to a low depth range (up to 2 μm) Ephiphot 200 inverted metallographic microscope at ×50–500 of electrons of excitation source, the analyses did not allow to magnifications. Two samples (see below) were taken and their obtain reliable information about the raw alloy composition. surfaces were etched with 1 part of perchloric acid HClO and 2 An X-ray generator designed and built in the NCNR in parts of acetic acid anhydride C H O solution for 10 s. Świerk, Poland, was used for energy dispersive X-ray 4 6 3 Fig. 5 Sample A. OM images of metal alloy structure obtained after etching at a, b mag × 10 and c mag × 20. There is an internal core dendritic microstructure (two-phase multiple components system) with a porous corrosion layer up to 100 μm thick in the surface (a) 978 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 6 Sample B. OM images of metal alloy structure obtained after etching at a mag × 2.5, b mag × 5 and c mag × 20. One-phase system and an external porous corrosion structure (up to 50 μm thick) can be seen fluorescence (ED-XRF) measurements. This compact X- Description of the samples ray generator (weight 0.835 kg, length 430 mm) was con- structed with the use of transmission X-ray tube technol- Complementary instrumental physico-chemical techniques ogy. A focused electron beam bombards a silver (Ag) were used to examine the material structure (for a similar anode target generating X-ray radiation. The silver anode methodology, see, e.g. Miazga 2014,Miśta et al. 2015 & has been sputtered on beryllium. The maximum acceler- Miśta et al., 2016, Meeks 1993, Ashkenazi et al. 2015, ating voltage is 50 kV. The beam current and accelerating Giumlia-Mair 2005) of this unique fibula. The analyses voltage can be smoothly adjusted. were performed on the unchanged artefact and on the cuts The XRF spectrum was induced by X-ray irradiation on a of two samples. Figure 4 shows the cuts with areas from spot with a diameter of 2.1 cm. Such a large analysing area where the samples were taken. The samples were mounted yielded well-averaged values of the elemental composition. in epoxy resin. They were ground using Premium SiC The X-ray generator was operated with an anode voltage abrasive paper (for 5 s with P220, P500, P800 and P120 and current of 40 kV and 30 μA, respectively. Fluorescence type: http://www.metallographic.com/Brochures/SiCpaper. spectra of studied objects were recorded with an Amptek Si- pdf; accessed on 8 November 2017) and polished using PIN detector. The energy resolution of the detector was as- Al O powder for 2 min. The surfaces of the prepared 2 3 sumed to be 149 eV at 5.9 keV photon energy. The measured cuts were etched in 1 HClO /2C H O solution for 10 s. 4 4 6 3 energy resolution was found to be 152.9 ± 0.3 eV. The dis- Elemental and phase components of the prepared samples tance between the X-ray source and measured objects (SO) which correspond to the artefact’s raw alloy without and between objects and the detector (OD) was set to 9 and corrosion, degraded surface and subsurface structure were 11.5 cm, respectively. Energy spectra were analysed with an determined. Furthermore, the cuts correspond to the parts Amptek ADMCA multi-channel analyser. Data on contents of of the fibula which were manufactured separately, i.e. the copper (Cu), tin (Sn), iron (Fe), lead (Pb), zinc (Zn), silver cross-section of the bow (sample A) and part of the axle of (Ag) and gold (Au) was obtained. Moreover, μ-XRF analyses the spring (sample B). of elemental composition were carried out on sample cuts Optical microscopy (OM) images of the cuts’ surfaces (sample A and sample B, see below). show the crystallographic metal alloy system with corrosion XRD method was used for phase identification. Powder X- structures on the surface layer. In Fig. 5, there are results of ray diffraction measurements were performed with Cu Kα microscopic examinations. radiation (λ = 1,5418 Å) using a Siemens D500 diffractometer Scanning electron microscopy (SEM) revealed information equipped with a Si:Li semiconductor detector cooled with about the surface condition (see Fig. 7). Moreover, using EDX liquid nitrogen and a ICDD PDF4 2014 database. The powder spectrometry allows to determine elemental composition in diffraction pattern was measured in θ/2θ scanning mode with the cuts. XRD and HV0.2 analyses were also performed for a step of 0.02 and integration time of 10 s/step. sample A and sample B. XRF analysis was carried out in Vickers micro-hardness (HV0.2) was measured on micro scale. polished surfaces of the cuts, using a 200-g load applied X-ray fluorescence spectroscopy (XRF) does not require for 12 s. At least 10 measurements per data point were sample preparation (see Gójska & Miśta 2016,Miśta & made. A Zwick Roell ZHU0.2/Z2.5 machine was used. Gójska 2016). The ED-XRF measurements were done on The tests were carried out according to the ASTM E384 the entire artefact in order to determine the elemental compo- standard with a constant distance of 2.5 diameter of a sition of the external alloy layer (up to ca. 20 μm: Henke et al. stamp between every stamp. 1993) by recording signals from components. Archaeol Anthropol Sci (2019) 11:973–983 979 Fig. 7 SEM–SE images of a fragment of the higher disc fragment (obverse side = obv.). Surface relief ornament, soil contamination and corrosive degradation can be seen Results and discussion the methods are based on measurements of X-ray spectra. The characteristic energy of X-ray corresponds to a given A preliminary SEM–EDX study of the fibula shows its sur- element in the sample. The difference between these methods face degradation which was invisible to the naked eye. The consists in another type of X-ray excitation: for EDX, it is results (Fig. 7) demonstrated that it was necessary to work electron beam and for XRF, it is X-ray from an X-ray tube. with pure metal alloy whose structure corresponds to the raw The range of penetration depth for electrons and X-ray is material. Below are examples of SEM images obtained for the different so we can study surface and subsurface layers fibula surface with relief ornaments (obverse side). Irregular (Miśta et al. 2016). This penetration depth is highly dependent cloudy shapes correspond to the corrosion and soil contami- on the material density. Furthermore, the areas of sampling are nation of the structure which is mainly built of light element also different. For ED-XRF, it is a field with a diameter of oxides. The external surface is also rich in iron and copper 2.1 cm, while for EDX and μ-XRF, it is a field within a oxides (up to ca. 50 wt% in total). micrometre area. The obtained numerical data were Table 1 shows the results of the quantitative elemental recalculated and are offered in Table 1 as normalised weight analysis carried out with the use of different methods. All percent taking into account the alloying elements. Table 1 Results of the SEM–EDX, ED-XRF and μ-XRF quantitative elemental composition analyses (wt%) (low detection limit LLD = 0.1 wt%) Technique Sampling No. Cu Sn Zn Pb Ag Fe For the entire artefact: ED-XRF Obv. 1 59.59 ± 5.96 22.46 ± 2.24 10.28 ± 1.03 4.34 ± 0.43 1.58 ± 0.16 1.39 ± 0.14 Rev. 2 52.67 ± 5.23 27.09 ± 2.71 9.36 ± 0.94 5.38 ± 0.54 1.84 ± 0.18 1.63 ± 0.16 For the cuts of samples A and B: μ-XRF A 1 71.23 ± 7.12 10.19 ± 1.01 18.58 ± 1.86 < LLD < LLD < LLD B 2 90.09 ± 9.00 2.85 ± 0.28 4.18 ± 0.42 1.54 ± 0.15 0.60 ± 0.06 0.79 ± 0.08 SEM–EDX A 1 69.95 ± 1.94 12.34 ± 0.41 16.98 ± 0.53 0.60 ± 0.10 0.13 ± 0.04 < LLD 2 66.94 ± 1.85 12.57 ± 0.41 15.96 ± 0.50 4.40 ± 0.27 < LLD 0.12 ± 0.04 3 71.89 ± 1.99 9.83 ± 0.33 18.18 ± 0.56 0.11 ± 0.06 < LLD < LLD 4 67.25 ± 1.86 17.80 ± 0.56 14.62 ± 0.46 0.33 ± 0.08 < LLD < LLD 5 71.75 ± 1.96 9.43 ± 0.31 18.58 ± 0.56 0.24 ± 0.06 < LLD < LLD Av. 69.95 ± 1.06 12.34 ± 1.49 16.98 ± 0.73 0.33 ± 0.82 < LLD < LLD B 6 93.72 ± 2.56 2.13 ± 0.10 4.15 ± 0.17 < LLD < LLD < LLD 7 93.45 ± 2.55 2.09 ± 0.10 4.46 ± 0.18 < LLD < LLD < LLD 8 92.61 ± 2.50 2.03 ± 0.09 4.59 ± 0.16 0.63 ± 0.07 < LLD 0.15 ± 0.03 Av. 93.45 ± 0.34 2.09 ± 0.03 4.46 ± 0.13 < LLD < LLD < LLD 980 Archaeol Anthropol Sci (2019) 11:973–983 Due to a considerable degree of the artefact surface con- tamination and degradation, it was necessary to perform EDX and (for the sake of comparison) μ-XRF analyses on cross- section cuts (samples A and B). Figure 8 shows a juxtaposi- tion of the SEM–EDX mapping analyses performed for sam- ple A. The maps demonstrate the homogeneity of distribution of the main determined elements: copper (Cu), zinc (Zn) and tin (Sn) in the alloy core of sample A. The results of the quantitative elemental composition anal- yses of samples A and B are given in Table 1 while Fig. 9 offers an example of registered EDX spectra. The elemental composition of sample A corresponds to the alloy composition of the raw material of the fibula bow, as shown in Table 1. The main elements determined by SEM– Fig. 9 EDX spectra obtained for the cross-section cuts of the fibula: EDX are (in descending order) copper (av. 69.95 ± 1.06 wt%), sample A (blue, no. 2 EDX in Table 1) and sample B (red, no. 6 EDX zinc (av. 16.98 ± 0.73 wt%) and tin (av. 12.34 ± 1.49 wt%), in Table 1) while lead is present as trace element (av.0.33 ± 0.82 wt%). The results are confirmed by the μ-XRF analysis of sample (0.60 ± 0.06 wt%) and iron (0.79 ± 0.08 wt%) in the alloy. A(Table 1). This elemental composition corresponds to the bi- Due to the fact that these elements were determined by ED- XRF measurements at a significant weight percent level on the phase of casting brass with a significant addition of tin. Structures of this type can be seen in the OM images in artefact surface (see Table 1: ED-XRF results and Fig. 12), Fig. 5 and in the diffractogram in Fig. 10 (see below) their trace amounts in the inner alloy can be linked with partial (Czermiński 1978: p.375, Dobrzański 2002: p. 719). melting from the surface tinning coat (this issue is discussed Addition of tin (at the level of 1–2%) prevents zinc escape below). Therefore, with regard to the contents of the main from the alloy due to its easy evaporation above 907 °C elements, the composition of sample B corresponds to the (Chodkowski 1976: p. 340), while the addition of lead (1.5– composition of brass dedicated for a technique of plastic met- 4%) improves the machinability of the brass alloy (Dobrzański alworking called die forming (Czermiński 1978: p.375, 2002: p. 719). However, in this case, the addition of tin is Dobrzański 2002: p. 719). This type of brass is generally made significant (at about 12 wt%). This leads to the inference that of two components with small admixtures with a mono-phase it was introduced to the alloy in order to provide it with a gold crystallographic system (see Fig. 10). X-ray diffraction analysis (XRD) of metal alloy determined colour. This is also the case with the so-called Mannheim gold used in jewellery production. In its basic form, the alloy con- the crystallographic system of samples A and B (Fig. 10). It sists of 80% Cu and 20% Zn but also occurs with addition of corresponds to the elemental composition of the matrix alloy tin at the level of 9% (http://www.gutenberg.czyz.org/ of both parts of the artefact (see Table 1) which represents the word,44746, accessed on 8 November 2017). technological elements (see Fig. 4). Generally, in both In sample B, which is part of the axle of the spring, the diffractograms, there are visible reflections at approximately main elements are copper (av. 93.45 ± 0.34 wt%), zinc (av. 42° and 50°. These can be identified as fcc (face-centred cubic 4.46 ± 0.13) and tin (av. 2.09 ± 0.03 wt%) with trace amounts cell) solid solution, with a (Cu, Zn) and/or (Cu, Sn) structure, of other elements being below 0.1 wt% (in average value of respectively. It follows from the elemental analysis (see the EDX measurement done for n = 3 micro-sampling points). Table 1) and from the increased value of the lattice parameter However, results of μ-XRF analyses (see Table 1) indicate the with respect to pure copper. For sample A, the obtained value was a = 3.655 Å and for sample B, it was a = 3.642 Å, while presence of trace amounts of lead (1.54 ± 0.15 wt%), silver Fig. 8 a SEM-SE image (mag × 379) of the surface of the cross-section cut (sample A) and b–d (mag × 379) EDX maps with a regular distribution of the main constituting elements: b copper—Cu, c tin—Sn and d zinc—Zn in the alloy Archaeol Anthropol Sci (2019) 11:973–983 981 the relief ornament on the surface and forming the crumple zone like the curvature in the fibula catch-plate (samplings 4 and 5 in sample A) (Ashkenazi & Iddan 2012, Moriera and Tschiptschin, 2016 in Fig. 5). Summing up, brass with addition of tin and trace amount of lead (sample A) is used as casting alloy, while brass for forming (sample B) is more robust and has better malleable properties (Wesołowski, 1972: p.396–430, Doomke 1982). The selection of alloys demonstrates an intentional use of technological and functional criteria. The bow was made of soft alloy to make it easier to shape the artefact’s form and the relief ornament. On the other hand, moving parts like the spring and pin, which had to be durable, were made of harder metal. Such features are confirmed by the average values of Fig. 10 XRD diffractograms of samples A and B micro-hardness which are shown precisely for each measure- ment point in Fig. 11. The micro-hardness is 160 ± 2 HV0.2 standard lattice constant for metallic copper is a =3.615 Å. for sample A and 166 ± 2 HV0.2 for sample B. These differ- This is due to larger radii of tin (Sn) and zinc (Zn) than copper ences in micro-hardness are the results of variations of propor- (Cu), so their participation in the structure expands the crystal tions of copper, zinc, tin and lead (see. Table 1). lattice of pure copper. Additionally, in sample A, there are A quite homogenous crystallographic structure and uni- several crystallographic phases. This is due to the fact that form values of HV0.2 parameter for sample A and sample B the fcc structure contains an intermetallic Cu5.6Sn phase (or demonstrate that each part, that is the bow and moving parts a similar one, having a ‘deformed’ fcc crystal lattice), which is used for attaching the fibula to the garment was made sepa- observed in the dendrite structure in Fig. 5.In sample B,its rately. The bow was made by casting in a mould after melting, structure is quite uniform and more fine crystalline in compar- and the axle of the spring was made by die forming in a ison with sample A, as it can be seen in optical images in temperature range of 200–800 °C. In the case of Cu–Sn, this Fig. 6. The ZnO reflections visible in smaller sample B as corresponds to dendritic structure (see phase diagram of Cu– bigger intensity originate from zinc corrosion. Sn by Johnson in Dobrzański 2002: p.729,Fig. 7.15).The The low HV0.2 values indicate that the melting process Cu–Zn system is homogeneous in the range up to 40% Zn in a was carried at a not very high temperature. temperature range up to the melting point of pure copper No layer dislocation can be seen in OM photos (Figs. 5 and 6), (phase diagram of Cu–Zn by Howkins in Dobrzański 2002: which suggests the use of a method of mould casting without p. 719, Fig. 7.13, Wesołowski 1974: p. 404). treatment at a higher temperature. Below, Fig. 12 offers ED-XRF qualitative analysis results Generally, the HV0.2 values decrease in the subsurface obtained for the artefact’s surface, for the obverse and reverse area. This corresponds to weakening of the alloy structure side of the fibula, respectively. Table 1 offers the results of due to the annealing process. This process is necessary to form quantitative analysis. Fig. 11 Micro-hardness test results for sample A and sample B, measurement error ± 2HV0.2. Stamp places in the samples can be seen on the left 982 Archaeol Anthropol Sci (2019) 11:973–983 Fig. 12 ED-XRF spectra obtained for the fibula, for the obverse side with the relief ornament (blue) and for the reverse side (red), respectively In comparison with the base alloys identified in samples A determined as a brass with different contents of tin, zinc and and B, at both sides of the surface (to the depth 20 μm: Henke trace amounts of lead. The elements were mixed depending on et al., 1993), there is a noticeable tin (Sn) enrichment. Tin con- the functional requirements of each part of the artefact. Casting centration in the surface is above 20 wt%, while in the base brass was used in the manufacture of the bow. On the other alloy in samples A and B, it is at the level of 12 and 2 wt%, hand, forming brass as a base alloy was applied for the axle respectively (see Table 1). Additionally, in the surface, there are of the spring, which was a moving part of the fibula. The HV0.2 higher contents of silver (Ag) and lead (Pb). Furthermore, traces value is higher for mounting parts, that is for the pin and other of iron (Fe) are detected. A higher content of tin can indicate the parts of the attachment mechanism (sample B). The relatively use of hot-dipping tinning in order to decorate the surface of the low HV0.2 value and the presence of fcc network for Cu/Zn fibula (e.g. Miazga 2014, Šmitetal. 2008, Meeks 1993, and Cu/Sn crystallographic system (which corresponds to the Giumlia-Mair 2005). The addition of silver and lead can orig- identification of the elemental composition) indicates that the inate from silver ores (galena PbS) which could be added to the casting process was performed in a not very high temperature. tin solution used for coating the artefact by dipping it into the The presence of the zones with lower HV0.2 values proves the molten tin solution. On the other hand, the presence of iron use of the annealing process for surface ornament forming. should rather be related to soil contamination of the surface. Acknowledgments We are indebted to Mr. Tadeusz Sworobowicz and Mr. Janusz Bojarczuk for technical support in the course of the analyses. Thanks must also go to Mr. Leszek Kalicki who found the artefact and handed it over to the State Archaeological Museum in Warsaw. Conclusions Open Access This article is distributed under the terms of the Creative This archaeological examination and the technological analy- Commons Attribution 4.0 International License (http:// ses of the unique fibula found in Radziejów in Poland can be a creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appro- background for similar studies of other artefacts mentioned in priate credit to the original author(s) and the source, provide a link to the the introduction. In the future, such studies can shed more Creative Commons license, and indicate if changes were made. light on the presence and transformation of fibulae of such types in the Baltic Sea region. Furthermore, the study of the References elemental composition of the artefact, combined with physical micro-hardness tests and crystallographic examinations, Ashkenazi D, Taxel I, Tal O (2015) Archaeometallurgical characteriza- allowed for an identification of the technological process tion of Late Roman- and Byzantine- period Samaritan magical ob- jects and jewelry made of copper alloys. Mater Charact 102:195– which was used in the manufacture of artefacts of this type. 208. https://doi.org/10.1016/j.matchar.2015.01.019 The most important conclusion is that the fibula from Ashkenazi D, Iddan N (2012) Archaeometallurgical characterization of Radziejów was tinned, as evidenced by the XRF analysis of Hellenistic metal objects: the contribution of the bronze objects from the obverse and reverse of the find. Usually, fibulae of this type Rishon Le-Zion (Israel). Archaeometry 54(3):528–548. https://doi. org/10.1111/j.1475-4754.2011.00631.x were gilded in the discussed period. The raw alloy was Archaeol Anthropol Sci (2019) 11:973–983 983 Barbacki A (2007) Mikroskopia elektronowa. Wydawnictwo Politechniki Kontny B (2012) Trade, salt and amber. The formation of late Migration Poznańskiej, Poznań Period elites in the ‘Balti-Culti’ area of northern Poland (the Elbląg Bierbrauer V (2005) Archäologie der Langobarden in Italien: etnische group). In Bliujienė A (ed) People at the Crossroads of Space and Intepretation und Stand der Forschung. In Pohl W & Erhart P (ed.) Time (Footmarks of Societies in Ancient Europe). Archaeologia Die Langobarden. Herrschaft und Identität, Forschungen zur Baltica 17 (1), pp 60–76 Geschichte des Mittelalters 9, Wien, pp 21–66 Kontny B (2016) Unexpected relic—the unique relief brooch from Bitner-Wróblewska A (2001) From Samland to Rogaland. East-West Radziejów, decorated in Salin’s Animal Style I, BŚwiatowit N.S.^ connections in the Baltic basin during the Early Migration Period, IX (2015)/B Warszawa Meeks N (1993) Surface characterization of tinned bronze, high-tin Chodkowski J (1976) Małysłownik chemiczny. 5th edn. Wiedza bronze, tinned iron and arsenical bronze. In: Craddock P, La-Niece Powszechna, Warszawa S (eds) Metal plating and patination: cultural, technical and histori- Czermiński J (1978) Metalurgia. Encyklopedia Techniki. Wyd. Śląsk, cal developments. Butterworth-Heinemann, UK, pp 247–275. Kraków https://doi.org/10.1016/B978-0-7506-1611-9.50025-X Dobrzański LA (2002) Podstawy nauki o materiałach i metaloznawstwo. Miazga B (2014) Tin and tinned dress accessories from medieval Materiałyinżynierskie z podstawami projektowania materiałowego. Wrocław (SW Poland). X-ray fluorescence investigations. WNT, Warszawa Estonian J Archaeology 18(1):57–79. https://doi.org/10.3176/arch. Doomke W (in translation by) Jędrzejewski W & Jędrzejewski K (1982) 2014.1.03 Vademecum Materiałoznastwa,. 2nd edn., WNT Warszawa Miśta EA, Gójska A (2016) A metallographic analysis of copper alloy Droberjar E (2009) Některé problémy mladšidoby stěhováni národů v artefacts from the lake in Lubanowo. In: Nowakiewicz T (ed) Čechách. In Niezabitowska B, Juściński M, Łuczkiewicz P & Ancient sucrificial place in the lake in Lubanowo (former Herrn – Sadowski S (ed.) The Turbulent Epoch. New Materials from the See) in west Pomerania. Institute of Archaeology Warsaw Late Roman Period and the Migration Period, vol. I, Lublin, pp University, Warsaw, pp 213–225 133–148 Miśta EA, Gójska A, Kalbarczyk P, Bojarczuk J, Rzeszotarska- Giumlia-Mair A (2005) Tin rich layers on ancient copper based objects. Nowakiewicz A & Nowakiewicz T (2016) An analysis of the ele- Surface Engineering 21 no. 5-6:359–367 mental composition and a study of precious metal artefacts. In Gójska AM, Miśta EA (2016) Analysis of the elemental composition of Nowakiewicz T. & Rzeszotarska-Nowakiewicz A. (ed) Ancient the artefacts from the Kosewo archaeological site. Acta Phys Pol A sucryficial place in former Lake Nidajno in Masuria. Results of 130(6):1415–1419. 10.12693/APhysPolA.130.1415 laboratory analyses of selected finds, Institute of Archaeology Goldstein J, Newbury ED, Joy CD, Lyman EC, Echlin P, Lifshin E., Warsaw University, Warsaw, pp 17–66 Sawyer L & Michael, RJ (2007) Scanning electron microscopy Miśta EA, Stonert A, Korman A, Milczarek JJ, Fijał-Kirejczyk I, and X-ray microanalysis, 3rd edn. Springer, U.S.A. Kalbarczyk P, Wiśniewska W (2015) Material research on ar- Haselhof GH (1981) Die germanische Tierornamentik der chaeological objects using PIXE and other non-invasive tech- Völkerwanderungszeit. Studien zu Salin's Stil I, Berlin-New York. niques. Acta Physica Polonica A 128(5):815–817. 10.12693/ https://doi.org/10.1515/9783110884111 APhysPolA.128.815 Henke LB, Gullikson EM, Davis JC (1993) X-ray interactions: Moriera CV, Tschiptschin PA (2016) A dilatometric study of the influence photoabsorption, scattering, transmission and reflection at E=50- of residual iron content on the annealing behavior of cartridge brass. 30000 keV, Z=1-92. Atomic Data and Nuclear \Data Tables 54/2: Mater Res 19(2):483–489. https://doi.org/10.1590/1980-5373-MR- 181–342 2015-0597 Hillberg V (2009) Masurische Bügelfibeln. Daumen und Kellaren – Nerman B (1935) Die Völkerwanderungszeit Gotlands, Stockholm Tumiany i Kielary, Band 2, Schriften des Archäologischen Olsen VS (2006) The development of (proto) disc-on-bow brooches in Landesmuseums 9, Neumünster England, Frisia and Scandinavia. Palaeohistoria 47/48, 2005/2006: Hines J (1997) A New Corpus of Anglo-Saxon Great Square-Headed 479–528. http://rjh.ub.rug.nl/Palaeohistoria/search/search Brooches. Reports of the Research Committee of the Society of Rudnicki M (2014) Dwa znaleziska skandynawskich zapinek płytkowych Antiquaries of London 51, Woodbridge z terenów północnej Polski. Wiadomości Archeologiczne LXV: Høilund Nielsen K (1997) The schism of Anglo-Saxon chronology. In: 283–290. https://pbn.nauka.gov.pl/sednowebapp/works/594230 Jensen KC, Høilund Nielsen K (eds) Burial & society. The Salin B (1904) Die altgermanische Thierornamentik. Typologische Chronological and Social Analysis of Archaeological Burial Data, Studie über germanische Gegenstände aus dem IV. bis IX. Aarhus, pp 71–99 Jahrhundert, nebst einer Studie über irische Ornamentik, Stockholm Høilund Nielsen K (2003) Saxon art between interpretation and imitation: Schmidt B (1961) Die späte Völkerwanderungszeit in Mitteldeutschland. the influence of Roman, Scandinavian, Frankish, and Christian art Veröffentlichungen des Landesmuseums für Vorgeschichte in Halle on the material culture of the continental Saxons AD 400–1000. In 18, Halle (Saale) Green HD & Sigmund F (ed) The Continental Saxons from the Schmidt B (1976) Die späte Völkerwanderungszeit in Mitteldeutschland. Migration Period to the tenth century: an ethnographic perspective, Katalog (Nord- und Ostteil). Veröffentlichungen des San Marino, pp 193–233 Landesmuseums für Vorgeschichte in Halle 29, Berlin Høilund Nielsen K (2009) The real thing or just wannabes? Sjøvold T (1993) The Scandinavian relief brooches of the Migration Scandinavian-style brooches in the fifth and sixth centuries. In: Period. An Attempt at a new Classification. Norske oldfunn Quast D (ed) Foreigners in early medieval Europe. Thirteen interna- XV, Oslo tional studies on early medieval mobility, Monographien des Šmit Ž, Istenič J, Knific T (2008) Plating of archaeological metallic ob- Römisch-Germanischen Zentralmuseums 78, Mainz, pp 51–111 jects—studies by differential PIXE. Nuc. Inst. And Methods Physic Jørgensen L & Nørgård Jørgensen A (1997) Nørre Sandegård vest. A Res B 226:2329–2333 th th cemetery from the 6 –8 centuries on Bornholm, Nordiske Wesołowski K (1972) Metaloznawstwo i obróbka cieplna. WNT, Fortidsminder B/14, København Warszawa

Journal

Archaeological and Anthropological Sciences – Springer Journals

Published: Dec 16, 2017

There are no references for this article.