LAUSR

aspects of the extraction, geochemistry, working and transport of gold throughout the Sahara in the Islamic period, and thus one of the most informative chapters in the volume. There is a good deal of discussion of West African gold and reference to specific coinages including that of the Almoravids. His treatment of the pre-Islamic gold trade (163–68) is useful but regrettably does not offer much on the gold coinage of the Byzantine-period mint at Carthage, though Andrew Wilson suggests, as yet, ‘we cannot say whether Sub-Saharan gold formed a significant component of the Late Roman gold supply’ (197). It is remarkable, in view of the impressive fieldwork in the Sahara over the last few decades and all of the archaeological work in North Africa over the last century and a half, to read Andrew Wilson’s declaration in chapter 7 that ‘we have to start by acknowledging that the number of archaeologically identified objects that can be unequivocally shown to have come to the Roman world from the Garamantes is, currently, precisely zero’ (191). He makes a plausible inferential case, drawing on the very limited evidence of ancient texts and archaeology, that Saharan exports as part of an interlocking trade network probably included slaves, gemstones, gold, minerals, natron, alum, sulphur, agricultural goods, cotton, animals and animal byproducts, and invisible services and technologies. Chapters 8–10 by David Mattingly and Franca Cole, Lise Bender Jorgensen, and Stephane Guédon build substantively on Wilson’s contention in demonstrating that there is important new archaeological evidence from the Fazzan Project and Egypt, as well as documentary evidence from the Roman-period Zarai Tarif inscription, and from the Bu Nijim ostraca for the manufacture, circulation and trade in organic commodities, such as fabric, clothing and textiles. Turning to inorganic evidence in the form of pottery, glass, bead working and metallurgy, Victoria Leitch et al. (chapter 11) present a remarkably thorough progress report on the archaeological evidence of Garamantian trade connections which asserts, compellingly, that Mediterranean pottery and glass, and by inference other less visible commodities, were traded in considerable volume and scale to the Garamantes. Significantly, they also contend that ceramics and glass were an abnormal component of ancient sub-Saharan trade, reflective of the particular commerce generated by the Roman Empire. Michel Bonifay (chapter 13) complements these observations by tracking the evolution of the Roman pottery trade with the Garamantes. The quantity of Roman pottery imports was considerable between the first and fourth centuries (though not evidently as far as the sub-Saharan regions), but a shift and possible disruption in pottery sourcing in Roman Tunisia and Libya and along trade routes is evident in the fifth and sixth centuries before the Saharan market Roman pottery declines in very late antiquity. On the other hand, Sonja Magnativa (chapter 14) detects archaeometric evidence of a trans-Saharan trade network in glass beads, copper beads and cowries in the first half of the first millennium BC which had some connections to sub-Saharan West Africa. Laure Dussubieux (chapter 15) contends through chemical analysis that the Middle East (possibly the region of Iraq and Iran) was providing glass beads to sub-Saharan Africa in the eighth to ninth centuries. This was followed by the opening up of a glass bead trade between the Indian Ocean trade network in the eastern and southern sub-Sahara. The concluding discussion by David Mattingly et al. (chapter 16) arrives at a consensus-driven set of provisional statements, which I summarise here. First, a pre-Islamic trans-Saharan network of contacts existed which could have supported the development of trade and exchange links across at least some of the network. And it was upon this network that later Islamic trade may have evolved, at least in part. This network in the pre-Islamic period supported trade in slaves, textiles, garments, pottery, metals, glass and other beads. Second, the Sahara was occupied by interconnected pastoral and oasis communities, inhabiting a range of settlements some of which approximated urban scale and character. Third, there is circumstantial evidence of Garamantian contact (including trade) with the sub-Saharan regions via north–south routes that took hold in the last half of the first millennium BC and reached its maturity in the first half of the first millennium AD. I would add the following provisional conclusion. Modern maps of northern Africa in antiquity used to stop at the border with the Sahara, which apart from the occasional citing of the Garamantes and other tribes mentioned in the ancient sources was largely left blank. Thanks to the work of David Mattingly and the many scholars who have contributed to this volume, we now know that the ancient Sahara was home to a rich and complex set of cultures in regular contact with the Roman African provinces and, no less importantly, with Saharan and sub-Saharan communities across the entire expanse of the desert. I suggest that readers turn the new maps in this volume upside down, that is with south at the top, to see how our knowledge and understanding of that world has changed.