Since 1966 excavations at Poggio Civitate (Murlo), a site in central Italy about twenty kilometers south of Siena (Fig. 1), have revealed an Etruscan habitation of the 7th and 6th centuries B.C. Heretofore our knowledge of the Etruscans, the inhabitants of central Italy during the Iron Age, had been based almost exclusively on funerary evidence, the rich and famous Etruscan tombs which have been avidly `excavated’ since the Renaissance. In the last two decades, with a new wave of archaeological interest in the economy and material culture of early Italy, the excavations at Poggio Civitate and at a number of other Etruscan habitation sites have radically altered our view of the Etruscans by providing timely evidence for daily life rather than for funerary ritual.
At Poggio Civitate the results have been rewarding (Nielsen and Phillips 1977: 73100 with previous bibliography). Excavation has produced a large building complex, nearly square and about sixty meters on a side with a large porticoed courtyard, the whole richly embellished with terracotta architectural decoration which includes a number of life-size animal, human, and mythological figures placed along the ridge of the roof. This building was deliberately dismantled in the second half of the 6th century B.C., judging by the latest pottery from the site. The rich architectural decoration was toppled, carted away, and buried. The site thereafter remained desolate. The reasons for this destruction remain a mystery, but the excavators, Kyle Phillips and Erik Nielsen, have suggested that the edifice, because of the nature of the destruction, the richness of its decoration, and certain aspects of planning, may have been a sanctuary.
Underneath this 6th century edifice a smaller but no less impressive building was found, It dates to the end of the 7th century B.C. and was destroyed violently by fire. Because of the suddenness of this destruction, the 7th century finds were much richer than those from the 6th century building. They include a large number of metal objects, imported Etruscan and Greek pottery, furniture, ivories, and so on, clearly the furnishings of an extremely wealthy Etruscan household. This earlier building, although relatively modest in scale, was also decorated with architectural terracottas.
The site of Poggio Civitate is admirably situated. It is a high hill at the eastern edge of a rough mountainous area known as the Colline Metallifere, the metal-bearing hills (Fig. 2). Farther to the east is less rugged terrain which might have been agriculturally productive in antiquity; today this eastern area encompasses the rich agricultural valleys that undulate toward Arezzo and Chiusi as well as the picturesque Chianti valley to the north. From the top of Poggio Civitate it is possible to see all the way to Siena and far off to the east toward Chiusi. The site also controls the natural trade routes. It and the nearby citadel of Montalcino sit astride the valley of the Ombrone, a small river which traverses the Colline Metallifere and thus provides the major route between the Sienese interior and the coastal flatlands of the Maremma, site of the early and rich Etruscan cities of Populonia, Vetulonia, and Roselle.
The wealth of Poggio Civitate, manifest in the rich finds of the 7th century phase, the size of the 6th century complex, and the architectural decoration found on buildings of both periods, is difficult to explain. The strategic location of the site could have given it control of the trade routes, but we know little of this remote section of Etruria and cannot specify, yet, exactly what was being traded. Nor can we assess the locality’s agricultural wealth in the 6th century until faunal and floral studies are completed. However it is suggestive that the site is located at the edge of one of Europe’s richest areas in metal ores.
Near Murlo there are a number of copp deposits. There is copper in the Crevole lode at Vescovado di Murlo (about a kilometer from the site), as well as at Res Monte Pescini, Monte Acuto, and Santo, all in the immediate vicinity. In the nearb Merse river copper is so abundant that it found as a precipitate along the banks of the stream. Traces of copper mining of undetermined date are found today in the hills immediately adjacent to Poggio Civitate. It seems certain that the Etrusca in this area would have been aware of these ores, especially given the evidence for metalworking at Poggio Civitate.
This all suggests another reason for the prosperity of the site. Thus the study of the metal objects found there, as well as the by-products of metalworking and ore refining, is particularly important. The metal finds were abundant and occasionally spectacular. They included objects of iron, lead, bronze, silver and gold. The artifacts range from mundane agricultural implements and household utensils to luxury items such as bronze cauldrons, figured bronze stands (Warden 1982), horse trappings (Fig. 4) and chariot fittings, as well as jewelry. Stylistic analysis of the objects has revealed that most were made locally; only a few pieces of gold jewelry are probable imports. Also, many of the objects are closely paralleled at the coastal site of Populonia, indicating that the metalworkers at Murlo had close ties to this famous Etruscan metallurgical center. From the 6th century period at Muria we also have primary evidence for metalworking. Although no actual workshops have been found, there are slag and metalworking by-products from stratified contexts, including a heap of slag just outside the northwestern corner of the building. Metallographic analysis of the objects themselves is currently being carried out by the Florence Archaeological Museum, while study of certain of the metalworking by-products has been carried out at the Laboratory for Research in the Structure of Matter at the University of Pennsylvania.
Several metallurgical by-products were analyzed to determine the nature of the metals being smelted and of the metalworking activities at the site. The materials studied included slags and crucible or furnace fragments with slag adhering. Because bluish-green flecks were visible in two of the crucible slags, it was assumed that copper was being processed. Elemental analysis of seven specimens of slag revealed that iron was the major constituent of six of them, but a high iron content is to be expected of a copper smelting slag because iron in some form—either as part of the ore itself or added as flux—is needed to slag off the siliceous component of the copper ore. The only material analyzed which had a high copper content was a green speck entrapped in a crucible slag; it must represent a prill caught up in the slag or dross from refining, which has oxidized since the time of deposition.
Copper production at Murlo is not surprising given the great amount of bronze found at the site. Even though Murlo is an Iron Age site most of the artifacts are not of iron but of bronze. Here, at any rate, copper and bronze are of major economic importance at least through the 8th century B.C., a phenomenon which can be documented in other parts of Iron Age Italy as well.
Although iron objects have been found at the site, however, iron ores are not found in the immediate vicinity, and it remains to be explained how the iron at Poggio Civitate was obtained.
It is therefore of considerable significance, both economic and technological, that at least one of the slags analyzed, and probably another, can be connected with iron working. When sectioned, one slag was found to contain a metallic iron bar (Figs. 5, 8), rectangular in section (L. 1.17 cm., W. 0.5 cm., Th. 0.3 cm.). The metallurgical implications of the bar and the entrapping slag are discussed below, but it is worth noting at this point that the small iron bar must indicate that some sort of iron production was taking place at Murlo for the following reasons. The bar is probably a piece of scrap fed into the furnace with other materials which were known to yield metallic iron. Iron rods may have been used in copper production to stir the molten copper, but the Murlo bar is too small to have had any effect on molten metal. It is conceivable that the bar has partially melted; that is, it may have originally had a thickness sufficient to make it a useful stirring rod. The Murlo bar shows little evidence of surface alteration; it retains a regular shape which indicates that it is a man-made product. It would not have contributed to copper production in any way, so it is likely that the slag from which it came is a residue of iron production.
Analysis of the slag itself aids in understanding the process by which iron was produced. Microscopically, the slag shows well developed fayalite dendrites (Fig. 7), whose presence indicates that the slag was formed at a temperature greater than 1177° C. The microstructure of the iron bar supports this inference. The bar is made of partially carburized iron. A flow pattern (Fig. 8) and elongated slag stringers (Fig. 9) indicate that the bar had been forged. It therefore cannot be an accidental production and must have been added as scrap as part of an attempted recycling operation. In addition, there are concentrations of impurities and slag at the grain boundaries; these impurities must have precipitated while the slag was hot.
Several small areas of metallic iron were also visible in the polished cross section of the slag. These iron patches are concentrated at one edge of the slag chunk and consist of sponge iron with entrapped slag (Fig. 10). Therefore, they are representatives of the stage of the iron production process represented by this slag: metallic iron produced during the heat treatment. Although iron does sometimes occur in metallic copper (Cooke and Aschenbrenner 1975), and virtually always is the major constituent of copper slag (although not in metallic form), might it not be that at Poggio Civitate attempts were made to produce metallic iron from iron-rich slags, probably those that resulted from copper smelting?
This suggestion finds support in another piece of slag from Murlo which in contrast to the one which contained the iron bar is non-magnetic, unlike other copper slags we have studied. It also shows signs of having been more strongly reduced, an indication that it may have yielded metallic iron, in a process in which the slag was ‘resmelted.’ This slag sample was taken from a heap of similar slag outside the northwestern corner of the 6th century building complex, an area which also produced the crucible or furnace fragments.
A parallel situation seems to have existed at 13th and 12th century Kition, on the eastern shore of Cyprus. Most of the slags studied appear to be of the type expected to result from copper smelting, while several are of the strongly reduced type which, on analogy with similar slags from Murlo, might have been used to produce metallic iron (for detailed reports see Stech, in press, and Maddin, Muhly, and Stech, forthcoming). Cyprus is a land poor in iron ores, yet we know that production of iron artifacts had begun there by 1220 B.C. Therefore some source of iron was available. It could have been imported as ore, but the possibility that the iron-rich copper slags were processed to obtain iron cannot be dismissed and indeed has been suggested previously by both Muhly and Frank Koucky. Of significance in this connection is documentary evidence for the export of Cypriot copper slags to France during the Ottoman period to be used as sources of iron (Steinberg and Koucky 1974: 178).
One of the slag specimens from Murlo yielded large enough pieces of charcoal for the wood from which it was made to be identified. The wood is an oak (Quercus) of the ring-porous red oak type. In modern Italy the most common oak of that type is Q. Cerris, sometimes called the turkey oak. Even today the hill of Poggio Civitate is covered with scrub oak which is periodically harvested as fuel, and the Colline Metallifere are also covered with scrub oak as well as scrub pine. In 1937, analysis of slag from furnaces at Campiglia, on the coast near Populonia, also revealed the presence of charcoal made from Q. Cerris as well as from another type of oak, Q. Robur (Tongiorgi 1937: 332-3).
Iron ores are not found near Poggio Civitate. The nearest iron ores are at Montieri, more than thirty kilometers away, a site which also has the richest silver ores in Europe. At Montieri the iron ores are found in small quantity and there is no evidence for ancient exploitation; it is doubtful that the inhabitants of Poggio Civitate would have had access to these ores. Apart from Montieri the nearest iron ores in any quantity are on the coast, in the region around Populonia and on the island of Elba (Fig. 1). Here the iron ores are very abundant and were undoubtedly known to the Etruscans. By the end of the 6th century Elban ores were transported to the mainland to be smelted at Populonia where over the next few centuries slag heaps accumulated over the 7th century cemeteries. In some cases the tumuli were used as furnaces, showing, if nothing else, a certain ingenuity an the part of the Etruscan metalworkers but also rather a casual attitude toward the tombs of their ancestors. What is strange is that at Populonia we have no evidence for extensive iron smelting much before the middle of the 8th century B.C., while at nearby Campiglia there is only evidence for the smelting of copper ores (Warden, in press). In view of the close connection between the metal assemblages of Poggio Civitate and Populonia, the evidence for smelting at Murlo at the same date or at an earlier date is interesting. There is clearly a connection between the two sites, but at Murlo, it would seem, iron ore was more difficult to obtain and iron metal had to be extracted from copper slags.
This brings up a number of questions. If, as now seems likely, the wealth of the site at least partially resulted from metal trade and metal exploitation, could it also be possible that the decline or destruction of the site might also have partially resulted from economic factors? Also, is there any connection between the destruction of Poggio Civitate—with the resulting termination of metal production and trade in the eastern part of the Colline Metallifere and the sudden rise of iron production at
The problem is tantalizing but the solution will have to await a careful assessment of all the economic factors. In the end this may have to include study of the entire Colline Metallifere and neighboring sites, but this has already proven itself an area In which metallurgy and archaeology together provide answers to crucial historical problems.
Populonia, production of such great scale that it surely must have affected the economy of all the minor metalworking centers in the interior?
The revised dating for the beginnings of iron smelting at Populonia and the suggestion that iron ore from Elba was not being shipped to the mainland before the end of the 6th century, also calls into question the theory that Greek colonists came to western Italy [Pithecusa and Cyme] in search of new sources of iron ore as early as the mid-8th century. The evidence presented here certainly supports the argument that, at least in its early phases, Greek colonization represented a search for fertile land, not mineral resources (see Graham 1982: 102-3).