THE METALLURGY OF TWO VILLANOVAN SITES IN THE REGION EMILIA- ROMAGNA (ITALY

THE METALLURGY OF TWO VILLANOVAN SITES IN THE REGION EMILIAROMAGNA (ITALY) C.Chiavari1, M.Degli Esposti2, A. Giumlia-Mair3, G.L.Garagnani4, C.Martini1, M. Pacciarelli5, D.Prandstraller1, T.Trocchi6, P.von Eles6 1 Dipartimento di Scienza dei Metalli, Elettrochimica e Tecniche Chimiche, Università di Bologna 2 Scuola di Specializzazione in Archeologia, Università di Bologna 3 AGM Archeoanalisi, via E.Toti 8, I - 39012 Merano (BZ) 4 Dipartimento di Ingegneria, Università di Ferrara 5 Dipartimento di Discipline Storiche, Università di Napoli Federico II 6 Soprintendenza per i Beni Archeologici dell’Emilia Romagna ABSTRACT The finds from two Villanovan cemeteries located near Castel San Pietro Terme (Bologna) and Pontesanto (Imola), in the region Emilia Romagna (Italy) have been recently studied and analysed in the frame of different projects. In particular two classes of objects – fibulae and weapons – were examined and on them several kinds of analyses (XRF, AAS, SEM/EDS) were performed. They show interesting elements for the reconstruction of the metallurgical tradition of the Iron Age culture in the sites in the South-Eastern area of the Po Plain. The items in both groups are particularly suitable for the evaluation of the skills of the artisans who produced them and for the determination of the copper-based alloys in use in this area and period for the production of decorative objects of personal use and for different kinds of weapons. This paper discusses the results of the analyses and the metallurgical characteristics of the objects. KEYWORDS Villanova, necropolis, Iron Age, copper-based finds, XRF, AAS, SEM/EDS, fibulae, weapons INTRODUCTION In the region Emilia-Romagna several cemeteries dated to the Villanovan period, i.e. to the Early Iron Age of the South Eastern Po Plain, have been unearthed in different periods and particularly in recent years. Two of them have been studied in the frame of different projects and the results of the chemical analyses of part of the copper-based finds are presented here. The artefacts considered here come from the Villanovan cemeteries of Imola – Pontesanto and Orto Granara, both discovered and investigated between 1998 and 1999, and located in the Po Plain at the east from Bologna, not far from the watercourses of the rivers Santerno and Sillaro. The two sites consisted of two small groups of graves dating to the Early Iron Age. At Pontesanto ten graves connected to incineration rituals have been unearthed, nearly all burials were originally inside wide wooden cases. The graves contained various offerings and the characteristic vases commonly named “biconici” (biconical) which held the ashes of the dead.,; Among the offerings are to be mentioned prestigious ceramic and bronze vessels sets, personal ornaments – like fibulas and pins -, horse fittings and, in male burials, tools or weapons. The weapons often had just a strong symbolic meaning, as it is the case with the axes with thin and expanded blade. These must be seen more as a rank display than as usable objects. The characteristics of these burial sets seems to indicate that the inhabitants of this small site, (some interesting dwellings remains have also been brought to light), enjoyed a great economic wealth. The site can be dated to a period between the middle of the 8th century B.C. and the first years of the following century. In the necropolis at Orto Granara 14 incineration burials have been explored, all in a wooden case and furnished with precious goods indicating a high economic potential, nevertheless at a definitely lower level than the necropolis discovered at Pontesanto. In these graves, with the exception of a rich masculine burial, ceramic vessels were more numerous than metallic ones, while personal ornaments, horse fittings, axes and other weapons were widely documented. As in Pontesanto, here too some remains of the settlement have been discovered. They include the remains of a ceramic furnace. The burial offerings allow dating the site between the second half of the 8th and the first 25 years of the 7th century B. C. The two sites can be considered examples of the numerous agricultural settlements spread between the middle of the 8th century B. C. and the beginning of the 7th, in a wide area surrounding the proto-urban settlement of Bologna, enclosed on the western side by the Panaro river, on the east by the Santerno river, and on the south by the Apennines. These are clearly high rank centres, where a ruling social class, directly linked to the rising aristocracy of Bologna, exerted – at different grades of social relevancy and possibly with a definite territorial hierarchy – a systematic and capillary control over the economic resources of the territory, related to agriculture and sheep breeding. Aim of this study was the identification and the comparison of both the alloys and the manufacturing techniques used in the production of the copper-based finds from these two sites. 1. EXPERIMENTAL The samples for wet chemical analysis were taken with a jeweller drill. The upper layer of patina was discarded and only clean metal turnings were collected. Any visible corrosion remain was cleaned under the microscope before weighing and dissolving the samples to prepare the common solution suitable for AAS analysis [1]. XRF analyses were also performed on all pieces and on the clean metal turnings to evaluate the degree of corrosion and compare the results. In some cases metal turnings taken from the samples before weighing were also examined by SEM/EDS. The results of the different analytical methods were compared as a check and to determine the degree of error due in particular to corrosion phenomena. The objects resulted to be in general in excellent conditions. The corrosion present on some parts of the thinner objects could be easily avoided and does not seem to be relevant for the analyses. The data obtained by surface analyses techniques (XRF or SEM/EDS) on spots which resulted to be corroded were not taken into consideration. In the case of objects from Orto Granara, EDS analysis on wide areas of the core metal was used to obtain representative average values of alloy composition when it was not possible to perform AAS analysis. 2. RESULTS AND DISCUSSION Pontesanto (Imola) A large group of items from the necropolis of Pontesanto, near the town of Imola have been studied and analysed. In this study around 20 pieces, decorative items, such as fibulae and pins, weapons, such as axes and spearpoints, horse fittings and the ingots found in the graves will be discussed, as they are easily comparable with the objects found during the excavations of the necropolis of Orto Granara. The group of decorative objects comprises 12 items: 3 spirals, 7 pins with different shapes, a pendant shaped like an axe, and a serpeggiante-type fibula. Among the weapons there is an axe with a wide blade, one of the type “ad alette” and a spearhead with spirals. A phalera, probably used as decoration of horse trappings has been also examined. A small ingot and a larger piece of pick-ingot (or biconical ingot) resulted to be of great interest. Orto Granara (Bologna) The group of finds from the necropolis of Orto Granara consisted of 19 pieces belonging to different classes of objects, which were discussed with more detail in a previous paper [2]: here only a few objects, belonging to the same classes of the finds from Pontesanto, will be discussed for comparison. The decorative objects from Orto Granara consist of 4 fibulae of different types and 3 pins with different shapes. Among the weapons, two axes of the type “ad alette” have been investigated. A couple of bits, made by joining two pairs of bronze rods heavily cold-worked in order to obtain the final shape, belong to the class of objects related to horse fittings. A group of four little ingots turned out to be very interesting as regards the composition and microstructure. 2.1. TIN CONTENT The tin content of the group of decorative objects retrieved at Pontesanto compares well with other analyses of Etruscan material [3, 4], as well as with that of the contemporary finds belonging to the same class of objects, from the Eastern Alps [5-11], with Sn weight percentages between ca. 6 and 12%. As expected, also the finds from Orto Granara show the same characteristics, with a tin content spreading across almost the same range of percentages (6 – 13% - see Table 2). Other comparisons with decorative objects coming from the near Villanovan site of Verucchio can be drawn [12,13]. They show a knowledge of the properties of bronze alloys, shared by different communities settled in the same territory, but possibly occupying different steps of the hierarchical organization. The only exception is represented by the trident shaped pin, with a 4,2% Sn content: this could testify that the production of this rather rare artefact was conducted by a different workshop than the other objects. However, compositional data for this find come from EDS elemental analysis which may be more influenced by local heterogeneities than other bulk analysis methods. In the case of the axes from Pontesanto, the tin content is slightly lower than in the decorative objects, but still correlated with the use of these superbly decorated weapons which had for sure a ceremonial function (Fig.1). In the case of the the axes from Orto Granara, found in grave 6, the Sn content is coherent with that of the other objects, if not slightly higher. It should also be noticed that the tin percentage determined in the axe 6.6 (9,6%), is very close to the value obtained from the analysis of the axe T4R65 from Pontesanto (9,3%). (a) (b) Figure 1: one of the analysed ceremonial axes from Pontesanto (a), and the spearpoint from the same site (b) The spearpoint from Pontesanto (Fig.1b) resulted to be slightly corroded, as are also its spirals. Nevertheless it seems that the copper employed for the production of this piece had been carefully refined before using it. The tin content is relatively high (9,6%) also in this case and the weapon must have had a golden appearance. Also the axes from Orto Granara, which have a tin content comparable to that of the spearpoint from Pontesanto must originally have had the same colour. This decorative aspect of the analysed weapons seems to indicate that they had an almost exclusively ritual function. Nevertheless, in the case of the axe 6.7 from Orto Granara metallographic observations showed a strain-hardened microstructure [2], induced from cold working in the final step of the manufacturing process. The cast phalaera is also slightly corroded, but the tin content seems to be still that of the original alloy, as it correlates well with the function of the object. The composition can also be compared with the results obtained from the analyses of the horse bits from Orto Granara and from other specimens coming from Verucchio [12, 13]. The objects belong to the same class and might be an item of local production or might have been supplied from the same commercial network, since, up to now, no structure that could be related to metallurgical production has been found on the sites. Until now, the existence of bronze workshops can only be hypothesized on the basis of the presence of ingots and finished artefacts in both sites. 2.2. LEAD CONTENT The lead percentages appear to be widely scattered in the items from both sites. The presence of lead in the bronze alloy was supposedly related to the amount of cold working to be performed on the objects after casting. This can clearly explain the absence of this element in the analysed spirals from Pontesanto, subjected to an extensive plastic deformation in order to obtain the final shape, as well as the low amount of lead in the bits from Orto Granara which were heavily cold worked to obtain their convoluted shape. The cast phalaera and the axes from Pontesanto show lead percentages which suggest that the final shape was obtained directly from casting The only object from Orto Granara containing a comparable amount of lead, is the a drago-type fibula 5.30; in this case, the alloy was probably produced with the aim of achieving a good castability, because of the complex shape of the object. The value determined in this sample is in contrast with the low lead content revealed by the thickened arc fibula 11.25, while it compares well with other objects of the same class but different typology) from the National Museum of Ancona [14], probably of Villanovan production, and in general with the data from other analyses of Etruscan materials [2]. In the case of the fibula 11.25 from Orto Granara the reason for its composition is probably connected with the production technique, which involved plastic deformation of the pin-holder and of the pin after casting. The composition of the serpeggiante-type fibula from Pontesanto is quite peculiar and will be discussed below in more detail. The complex bow is obtained by casting and the presence of arsenic and antimony facilitated the casting process. 2.3. TRACE ELEMENTS The trace elements determined in the objects from Pontesanto are rather irregularly distributed. In general it seems that either the copper employed for the production of the objects from the graves of the necropolis was not always carefully purified, as it was for example the case with the metal used for the production of the spearhead and the spirals belonging to this weapon, or that the copper came from different sources. The trace elements distribution in the artefacts from Orto Granara appears quite regular, testifying careful refining of the metal. Ni, Fe, As, Sb and Ag all are present as trace in the finished artefacts, while the content of these elements in the ingots from grave 6 are rather high. In particular, the iron content of the finds from Orto Granara is constantly lower than 0,3%, while at Pontesanto its concentration goes from amounts as low as the 0,08% to others as relevant as the 1,72%, however this is for sure correlated to the different use of the objects which needed different degrees of purification.. The differences in the arsenic content of the two groups are not significant: this element is present in different amounts in all objects and its high volatility is the reason for the wide range of concentrations. Also this fact seems to be related to the differences among various classes of objects and their production techniques. The antimony percentages in the two sites are comparable, as are also the nickel concentrations. The silver content is quite low in both groups of objects . Perhaps lower percentages can be recognized in the group of artefacts from Pontesanto. In some of the objects the traces of this element were too low to be determined. This difference is perhaps due to the different provenance of the metal, as shown by the ingots analyses discussed below. It is quite important to note that the small ingots found in Orto Granara and Pontesanto seem to belong to a broken pick ingot (or biconical ingot). A larger piece of pick ingot (fig.2) was present in grave 4 (R46) of the necropolis of Pontesanto and it shows the unmistakable composition , characteristic for this kind of ingots which circulated in particular in the regions around the Eastern Alps [15]. The general composition of the ingots is similar. The differences in the Sb and As contents are due to the fact that these ingots are produced by smelting mixed ores which had of course an irregular composition. Sb and As are always present in noticeable amounts. The ingot 6.26d, is the only piece from Orto Granara for which Sb was determined, and gave the value of 4,67% Sb. The irregular composition is evident also by comparing the Ni and Fe contents in the ingots from Orto Granara and from Pontesanto. The same can be said about the silver contents: while in the pieces from Pontesanto this element resulted rather low, in the ingots from Orto Granara, the silver content is always higher than 1% . Pick ingots can show relatively high lead contents, while the presence of Sn in the ingot 6.26d from Orto Granara might indicate remelting of metal scrap. However the high percentages of volatile elements like As and Sb seems to be inconsistent with repeated heating and remelting of the metal. (a) (b) Figure 2: Large fragment of pick ingot from Pontesanto, grave 4 (a), and serpeggiante –type fibula from the same grave (b) Surprisingly the fibula of serpeggiante-type, found in grave 4 (R76) shows the same characteristic high percentages of As, Sb, Ni, and Co, albeit with the addition of some Sn. The silvery colour of this fibula was for sure the reason for the choice of this particular material (fig. 2b). The composition is very similar to that of the much later Almgren-type fibula [16] recovered from the excavations in the ancient Roman town Emona. This suggests that this kind of material had been in use for a long time in this particular area. The artisan added some tin also in this case, however there is an important difference: the fibula from Emona contained over 40% of Ag which is not present in the fibula from the necropolis of Pontesanto. The addition of Sn improved further the casting properties of the serpeggiante-type fibula which has a complex shape. The addition of lead would have darkened the alloy and was of course avoided. The pin must have been slightly hammered, taking care not to break the fragile object. The fibula, together with the pieces of biconical ingots demonstrates quite clearly the relationships of the site with the regions around the Eastern Alps and perhaps even with Istria and the Dalmatian islands. The seemingly irregular distribution of trace elements in the objects does not mean that we are dealing with different workshops. The number of observations on objects belonging to the same classes is too low to be statistically relevant. A larger number of analyses might give a completely different picture. On the contrary, the similarities in the composition and type of alloys choosen for the objects belonging to the same class seems to indicate that in both sites the metallurgical tradition is the same. However there are also clear differences in the composition of the ingots and of some of the objects, which in this case can be explained with the use of the metal of pick ingots imported from transalpine sites for the production of some of the objects. The different refining degree is of course due to the fact that the raw copper had to be purified in the workshops. In many cases, in particular in the case of castings or large objects, the smiths did not bother to completely refine the metal, because it was not necessary for the production of the items they were working on. Table 1. XRF analyses of the finds from Pontesanto Object Inv. Nr. spiral spiral spearpoint pin pin pin pin pin spiral fib. serp. phalaera phalaera axe axe pin axe pend. pin pin pin pin axe ingot (small) ingot (large) pin T4R14 T4R15 T4R16 T4R3 T4R3 T4R3 T4R3 T4R3 T4R13 T4R76 T4R100 T4R100 T4R83 T4R41 T4R75 T4R17 T4R44 T4R35 T4R66 T4R39 T4R65 T4R46 T4R45 Part Cu Sn 89,5 8,87 - 0,07 0,05 0,79 - 89,8 7,57 - shaft point protection shaft disc rim 94,24 5,62 0,52 1,02 0,08 1,33 0,08 0,02 0,18 - head top 85,2 4,89 0,02 0,06 0,02 3,2 - - - 0,01 disc prot. 90,2 10,3 0,66 0,12 0,08 0,08 - - - - 88,9 9,23 - 0,06 0,05 0,09 - - - - pin holder 82,1 6,65 - 4,82 3,94 0,97 0,76 0,26 0,01 - rim ring side 90,5 89,2 7,63 2,26 0,12 0,11 0,12 8,12 2,71 0,09 0,08 0,09 - 90,2 6,42 2,11 0,08 0,04 0,43 0,09 0,17 0,11 - side 84,3 5,98 2,34 1,03 0,71 0,57 0,13 - shaft 85,1 13,9 0,83 0,5 ring 83,3 12,3 1,22 0,25 0,08 0,33 0,08 0,02 0,12 - shaft 89,1 10,2 - 0,08 0,02 0,31 0,08 - 0,09 - shaft 86,8 11,1 0,62 0,06 0,04 0,11 0,09 - 0,15 - shaft 83,2 8,9 - - - 1,54 - - - 0,03 shaft 88,2 9,2 0,44 - - 0,98 - - - 0,01 side 89,1 9,28 0,92 0,23 0,35 0,08 0,13 0,02 0,1 90,3 - 0,04 4,02 3,12 0,19 0,05 0,14 0,11 - 88,1 - - 87,8 10,5 - shaft Pb As Sb Co Ag Zn - - - 0,04 0,02 1,62 0,19 - - - 85,1 9,56 0,98 0,08 0,03 0,08 0,13 - - - 84,6 13,1 - 89,3 10,2 0,76 0,16 0,04 0,23 0,13 - 0,13 - - Fe Ni 0,14 0,09 0,03 0,02 - 0,20 - - - 0,09 - 0,27 0,17 0,13 0,08 - - 3,91 4,49 1,72 0,99 0,39 0,41 0,04 - 0,83 - - 0,12 - TABLE 2 – Analysis results of the artefacts from Orto Granara AAS analysis (wt%) Inv. Nr. Fibula “a drago” Axe Horse-bit Horse-bit Horse-bit Horse-bit Goad Fibula with lowered thickened bow Ingot Ingot Ingot Ingot EDS analysis (wt% ) Eyelet Eyelet Trident-shaped pin Pin with globular metallic head Pin with composite head Sn Pb Fe Ni As Zn Ag Sb 5.30 6.6 15.a1 15.a2 15.b1 15.b2 11.23 11.25 Sampling point apophysis bulk eyelet eyelet eyelet eyelet shaft bow 13,0 9,6 7,7 7,2 8,3 7,4 6,8 6,2 2,50 0,61 0,32 0,34 0,32 0,30 0,25 0,77 0,27 0,17 0,09 0,20 0,15 0,09 0,30 0,20 0,20 0,11 0,17 0,18 0,22 0,22 0,25 0,09 0,80 0,75 0,19 0,24 0,46 0,40 0,44 0,44 0,01 0,01 0,01 0,01 0,01 0,01 0,05 0,28 0,16 0,14 0,14 0,17 0,15 0,13 0,10 0,35 0,14 n.d 0,16 0,16 0,17 6.26a 6.26b 6.26c 6.26d bulk bulk bulk bulk 0,1 0,1 0,0 2,4 0,08 0,02 0,03 1,13 0,93 1,47 1,25 0,89 0,98 3,89 ,98 3,12 1,55 6,12 2,69 2,79 0,02 0,02 0,08 0,06 1,84 n.d. 1,69 1,26 n.d. n.d. n.d. 0,67 shaft shaft arm shaft Sn 8,8 7,3 4,2 8,9 Pb 0,8 0,9 Fe - 11.13 shaft 8,0 0,5 0,2 11.33 11.36 5.31 5.47 3.CONCLUSIONS The excavations in the cemeteries discovered at Imola-Pontesanto and at Orto Granara brought to light a group of bronze artefacts which seem to belong to the same metallurgical tradition of previously analysed objects belonging to Villanovan and Etruscan sites. The amounts of the main alloying element Sn appear to have be carefully controlled in order to obtain an alloy with the best mechanical properties (6-13%). The lead contents are rather irregular, with relatively low percentages also in castings. This compares well with other contemporary materials from other sites. The presence of fragments of pick ingot clearly show the trade links with transalpine metal production sites. The serpeggiante-type fibula from Pontesanto was produced with this metal and shows the characteristically high arsenic and antimony percentages. The choice of this alloy for the fibula is for sure due to the shiny silvery colour of these particular ingots, and was in this case enhanced by the addition of some tin. Also here, as in other contemporary sites, the copper came in form of raw ingots of different composition and had to be refined in the local workshops. REFERENCES 1) M. J. HUGHES, M. R. COWELL and P. T. CRADDOCK, Archaeometry 18 (1976), pp.19-37 2) C. CHIAVARI, M. DEGLI ESPOSTI, G. L. GARAGANANI, C. MARTINI, D. PRANDSTRALLER, T. TROCCHI, La Met. Italiana (2007), in press 3) P.T.CRADDOCK, J. of Arch. Sci. 4, (1977), pp.103-123 4) P.T.CRADDOCK, Studi Etruschi 52, (1986), 211-271. 5) A.GIUMLIA-MAIR, in Pozzuolo del Friuli, II, 2, La prima età del ferro nel settore meridionale del castelliere. Le attività produttive ed i resti faunistici. Quasar ed., Roma (1998), pp.181-230. 6)A. GIUMLIA-MAIR , in Metallurgica Antiqua, (=Der Anschnitt, Beiheft 8), Bochum, (1998), pp.45-55. 7)A. GIUMLIA-MAIR , Proc. XIII Int. Cong. of Prehist. and Protohist. Sci., Forlì (1996), 4 Sections (1998) , pp.665 - 672. 8)A. GIUMLIA-MAIR , Aquileia Nostra 69, (1998), pp. 29-136. 9)A. GIUMLIA-MAIR , Proc. of the Workshop on Ancient Metallurgy between Oriental Alps and Pannonian Plain, Trento (1998), (=Quaderni di Aquileia Nostra 8), Trieste (2000), pp.77-91. 10) A. GIUMLIA-MAIR , S. VITRI, S. CORAZZA, Archaeomet. in Europe 2003, Milano (2003), AIM ed., Milano (2003), pp.157-166. 11) A. GIUMLIA-MAIR , La necropoli di Misincinis. La metallurgia nell’Età del Ferro. Arti Grafiche Friulane, Tavagnacco, Udine (2003). 12) L. FOLLO, Guerriero e sacerdote. Autorità e comunità nell’età del ferro a Verucchio. La tomba del Trono. Il Giglio ed., Firenze (2002), pp.308-314 13) E. ANTONACCI SANPAOLO, Il dono delle Eliadi. Ambre e oreficerie dei principi etruschi di Verucchio. Ramberti Arti Grafiche, Rimini (1994), pp.49-53 14) S. SIANO, L. BARTOLI, J. R. SANTISTEBAN, W. KOCKELMANN, M. R. DAYMOND, M. MICCIO, G. DE MARINIS, Archaeometry 48, (2006), pp. 77-96 15) N. TRAMPUŽ OREL, D. J. HEATH, Arheološki Vestnik 52, (2001) PP. 143-171. 16) A.GIUMLIA-MAIR, Berliner Beiträge zur Archäometrie, Band 18, (2001), pp.5-42.