Notes The Harappan flint quarries of the Rohri Hills (Sind-Pakistan) PAOLO BIAGI & MAURO CREMASCHI* Flint quarries in the Rohri hills supplied stone to the city of Mohenjo-Daro, out on the silty river-plain and lacking local supply. A new survey has identified workshop sites and an extraordinary scale of production. Introduction In his geology of Western Sind, Blandford (1880) reports the presence of flint cores and flakes on the hills near Sukkur and Rohri. Even though no further description is given, these artefacts ‘scattered about in abundance in some places’ (Blandford 1880: 103) are to be inter- preted as prehistoric sites later re-discovered by Cousens (1929). A more accurate description of both Palaeolithic and Harappan assemblages is presented by De Terra & Paterson (1939), who also suggested that some of the flint tools they saw on the Rohri Hills, which resembled those from Mohenjo-Daro, were to be attributed to the Harappan civilization (Allchin et al. 1978). More recently, investigations were carried out by B. Allchin (1976) in 1975-6. She discovered several Palaeolithic and Harappan sites mainly located at the northern and south- western ends of the hills (Allchin 1979; Allchin & Allchin 1982). Recent and old excavations conducted at Mohenjo-Daro have revealed that the flint employed by the Harappan commu- nities was not available from the silty-clayey alluvial plain of the River Indus, but had been imported from outside. The more probable source for raw material is that of the Rohri Hills that lie close to Kot Uiji and some 50 km to the northeast of iblohenjo-Daro. The Rohri Hills are a dissected limestone plateau, composed of flat mesas, in all some 40 km long by 16 wide which are cut by the River Indus, to the west, and the Nara Canal, an old bed of the Indus itself, to the east (FIGURE 1) (Holmes 1968). Their name derives from the near-by town of Rohri, in front of Sukkur, at the northern edge of the hills, on the left side of the Indus. They are composed of stratified cherty limestones. Their surface is actually covered with an extraordinary amount of flint nodules whose presence is to be related with Late Tertiary weathering. Limestone blocks and fragments, which originated from thermoclastic activities, are visible between the flint nodules (Biagi & Cremaschi 1988). The 1986 survey Many Palaeolithic and Harap pan sites, mainly concentrated in the northern and central parts of the hills, were found during the survey of February 1986 (Biagi & Cremaschi in press). Most of the Harappan sites consisted of work- shops and chipping-floors connected with the exploitation of the raw material and the manu- facture of the flint artefacts. Two characteristic types of structures were recognized: simple chipping-floors surrounded by wide discharges of waste flakes (FIGURE Z), and complex features delimited by walls of limestone blocks (FIGURE 3). The more impressive structures were distri- buted along the edges of a limestone plateau in the neighbourhood of Shadi Shaheed (FIGURE 4). Here, at Hilly Site, the structures were often * Paolo Biagi, Dipartirncnto di Scienzc Storico-Arc:hcologichr e Orientalistichc. 1~nivcrsit;l di Vciiezid. Palazzo Bernardo. S. polo 1977A. 1.30125 Venice, Italy. Mauro Crcmaschi, Dipartimento di Scienzc drlh Trrrd. l1niversit;l di Milano. Via bfangiagalli 34, 1-20133 Milan. Italy. ANTIQUITY 65 (1991): 97-102