Fieldwork in the Zemplen Mountain area of north-eastern Hungary showed that there are at least eight geological sources of obsidian here, five of which have obsidian of a workable quality. There are a further three sources in the Slovak Zemplen, all of which provide workable obsidian. Sources in Central Slovakia are highly devitrified and not useable, and reported sources in Rumania had been discounted earlier (Nandris, 1975). Forty-six samples of obsidian from the Zemplen sources, and 293 pieces from 87 archaeological sites in Central and Eastern Europe, were analysed by neutron activation analysis for 15 trace and two major elements. The trace elements used included those which are geochemically likely to show the greatest variation between different obsidian sources, and which are not badly affected by devitrification and hydration of the obsidian, for example the rare earth elements. The analytical data was processed using Cluster Analysis. 242 of the archaeological samples came from Slovak sources, 22 from Hungarian sources, 9 from Lipari and 5 from Melos. In addition, 6 samples were tentatively assigned to Carpathian sources, and 9 could not be assigned to any source. Obsidian from the Zemplen Mountains was distributed up to a distance of approximately 480 km from the sources; it was used extensively in Slovakia and Hungary and reached southern Poland, Austria, Moravia, central Yugoslavia, north-east Italy and central Rumania. Obsidian use in central and eastern Europe began in the Mousterian period. The earliest pieces analysed were Aurignacian and came from Hungarian sources. Later, in the Gravettian, Slovakian sources began to be exploited and remained predominant until obsidian use declined sharply in the Later Neolithic, and Copper and Bronze Ages. The Carpathian obsidian distribution overlaps with the Liparian distribution at one site in north-east Italy. There is no evidence for an overlap with Aegean or Near Eastern sources. The rate of fall off of obsidian away from the sources suggests a down-the-line trading mechanism.

Fieldwork in the Zemplen Mountain area of north-eastern Hungary showed that there are at least eight geological sources of obsidian here, five of which have obsidian of a workable quality. There are a further three sources in the Slovak Zemplen, all of which provide workable obsidian. Sources in Central Slovakia are highly devitrified and not useable, and reported sources in Rumania had been discounted earlier (Nandris, 1975). Forty-six samples of obsidian from the Zemplen sources, and 293 pieces from 87 archaeological sites in Central and Eastern Europe, were analysed by neutron activation analysis for 15 trace and two major elements. The trace elements used included those which are geochemically likely to show the greatest variation between different obsidian sources, and which are not badly affected by devitrification and hydration of the obsidian, for example the rare earth elements. The analytical data was processed using Cluster Analysis. 242 of the archaeological samples came from Slovak sources, 22 from Hungarian sources, 9 from Lipari and 5 from Melos. In addition, 6 samples were tentatively assigned to Carpathian sources, and 9 could not be assigned to any source. Obsidian from the Zemplen Mountains was distributed up to a distance of approximately 480 km from the sources; it was used extensively in Slovakia and Hungary and reached southern Poland, Austria, Moravia, central Yugoslavia, north-east Italy and central Rumania. Obsidian use in central and eastern Europe began in the Mousterian period. The earliest pieces analysed were Aurignacian and came from Hungarian sources. Later, in the Gravettian, Slovakian sources began to be exploited and remained predominant until obsidian use declined sharply in the Later Neolithic, and Copper and Bronze Ages. The Carpathian obsidian distribution overlaps with the Liparian distribution at one site in north-east Italy. There is no evidence for an overlap with Aegean or Near Eastern sources. The rate of fall off of obsidian away from the sources suggests a down-the-line trading mechanism.