S
The original intention of the project, conceived in the context of the British School at Rome's excavation and survey programme in the Bradano basin (based on Gravina) was to establish the composition patterns of Hellenistic glazed wares from sites in southern Apulia, and then to use these to trace the trading patterns of the area, against a control group of sherds from the Athenian Agora. It proved possible to distinguish Attic from Apulian sherds by their nickel and chromium content, and average‐link‐cluster‐analysis by computer showed that Gnathian ware from Taranto could be distinguished from seven other south Italian groups, suggesting that speciality wares were made of clays from special beds. Because of the geological homogeneity of the area no further distinctions could be made, but some sherds were also analysed petrologically and by heavy mineral analysis, for a methodological comparison. The very homogeneous geology, coupled with the fineness of the wares showed these methods to be at the limits of their usefulness. Only very generalized conclusions could be drawn concerning the origin of the pottery from the petrological evidence; the micro‐mineralogical data showed that interesting differences existed between pottery from the main Apulian sites, but these could equally have reflected natural variations occurring within fairly homogeneous clay deposits.