In conventional reconstructions of southern African archaeology, the production of iron has been seen as unchanging for the last 2000 years. Significantly, this contrasts with the changes that have been noted in broader society and other classes of material culture of the same period. Despite iron being used as a chronostratigraphic indicator, virtually nothing is known on the patterns of iron production within the Iron Age and whether change in technology and the socio-cultural context of production took place. From a combined archaeological and metallurgical perspective, the historical development of iron working has never been explored. For example, it is not known whether similar types of furnaces were constantly operated throughout the last two millennia. Excavations at two sites in northern Zimbabwe, one Gokomere-Ziwa (800 - 1200 cal AD) and one Zimbabwe tradition (1500 - 1700 cal AD), have shown differences in iron pyrometallurgical debris, tentatively suggesting that they represent separate metal working practices. By comparing the archaeological and metallurgical evidence from the two sites, this paper represents an initial step in delineating patterns of indigenous iron production in one region of Zimbabwe.
Fragmentary glass-working crucibles, drawn glass beads and ritual glass objects (aje ileke) from Ile-Ife, southwestern Nigeria, were analysed using scanning electron microscopy (SEM-EDS), electron probe microanalysis (EPMA) and X-ray fluorescence (XRF). The very unusual high-lime, high-alumina glass lining the crucibles matched the composition of the dark blue drawn beads and some of the blue and green glass fragments in the aje ileke. Similar crucible fragments, glass cullet and drawn glass beads were recovered during Frank Willett's excavations (1956-63) of two sites in Ile-Ife, and Claire Davison's unpublished chemical analyses from 1972 show the same high-lime, highalumina glass from Ita Yemoo, with radiocarbon dates from the eleventh to thirteenth century CE, and Orun Oba Ado, with radiocarbon dates from the eighth to twelfth century. Such high-lime, high-alumina glass has been found only in West Africa, including Igbo-Ukwu in southern Nigeria, and is not known from Europe, the Middle East or Asia, ruling out the possibility that the glass was imported. We interpret these findings to propose the primary manufacture of high-lime, high-alumina glass in sub-Saharan Africa in the early second millennium CE, with production centred in southern Nigeria, and quite possibly in or near Ile-Ife. The results of our study, combined with those of Davison, provide the first strong evidence for early primary glass production in sub-Saharan Africa.
With exception of Maluma (1979) and Musambachime (2016, 2017), there have been no archaeometallurgical publications on the technology and culture of iron production in Zambia. This paper presents archaeological and archaeometallurgical evidence of a technology of iron production in Chongwe in terms of spatial organization, the process of metal production (either a three-stage process involving smelting in relatively tall furnaces, refining in miniature (vintengwe) furnaces, and smithing on a hearth or a two-stage process involving smelting and smithing), furnace air supply mechanisms, liquid slag handling techniques, variation in the geochemistry of ore and clay, and the nature of the final smelting products. Archaeological field data collection techniques included ethnoarchaeological interviews, (furnace) excavation, surface collections, and surface walkover surveys, while laboratory analytical techniques included optical microscopy (OM), scanning electron microscopy (SEM), and x-ray fluorescence (XRF). New field evidence indicates that iron production in Chongwe in the previous two centuries was secluded from respective pre-modern settlements for socio-cultural and technical reasons. There are no settlement remains in and around Chongwe smelting sites. Also, most of the archaeological data in Chongwe are supportive of the two-stage process that did not involve iron refining in vintengwe furnaces. There were no iron refining sites in Chongwe. Archaeological evidence also strongly points to the use of natural air supply mechanism for the smelting furnaces because proximal ends of tuyères inter alia were not trumpeted. All smelting sites were systematically located on termite mounds. There were three to four smelting furnaces located on the western side of a termite mound. The presence of tuyère mould slags and thin and elongated slag microstructures strongly indicates that liquid slag was tapped outside the furnace apparently through tuyères and was left to cool quickly. Presence of primary wüstite and iron particles in the slags strongly suggests the production of iron as the final smelting product in Chongwe. The results are compared with the archaeology, chemistry, and mineralogy of iron production from other parts of sub-Saharan Africa, particularly in the Lake Tanganyika-Nyasa Corridor. The presence of three to four smelting furnaces per termite mound makes iron production in Chongwe a unique technology in the Corridor.