Folia archeologica 54.
12 ANDRÁS MARK O Only the last step of the reduction could have been documented at the complex nr. 28 (Fig. 2.1.), when the damaged distal part of a flat core was regulated by a flake, refitted to the only one opposed-direction scar of this piece. Actually, taking into consideration the concept of the stone working, this artefact is considered as a repaired and exhausted single platform core. The elements of the complexes nr. 30 and 31 and further nineteen other pieces, including pre-cores, cortical flakes and blades of the same raw material are linked to the exploitation of one or two nuramulithic chert pebbles. One of the thirteen pieces of a large siliceous pebble of extremely poor quality was shaped as a unipolar core. From the side of this piece a laterally cortical flake was removed and on the opposite side two cortical pieces were broken off (complex 5a-Fig. 3.3.). After knapping the last short flake, absent from the assemblage, the core was abandoned. Finally two small conjoinable chips show the on-site preparation of a core of mauve-coloured radiolarite (complex 1). This raw material seems to have a distant source, lying probably in the southern part of Transdanubia; the use of the same rock was reported from the Ságvárian site of Madaras' 2 4 too. A completely different method of core exploitation could be traced by the refitting of cortical flakes of angular limnic quartzite pebbles (complexes 3, 4 and 8: Fig. 3.1, 4.) and quartzite cobbles (complex 7: Fig. 3.2.). At both raw material types the flakes with cortical base were removed maily from the shorter edge of the pebble and later the prepared blocks were exported from the excavated territory. A third method of flaking was observed on the oval-shaped (complexes 15 and 17: Fig. 4.), rarely irregular cobbles (complex 20) or slabs (complex 18) of quartzite, which were sliced according to a centrifugal pattern by series of large, partly cortical flakes (pebble slices and segments). As our data show, the next step of the core exploitation was the removing of several thin flakes with cortical base and partly weathered or cortical dorsal side from the relatively flat part of the cobble (complex 14). Another group of the reduction refits document the on-site retouching and the rejuvenation of formal tools. At the complex nr. 27 (Fig. 2.3.) the first phase of the pebble working was represented by the decortications of a natural fragment of a pebble (i.e. the removing of a flake from the dorsal part of the tool) or perhaps the forming of a dihedral burin on the natural cleavage surfaces of the piece. Later this burin edge was rejuvenated and the opposite-lying part of the tool was shaped as an end-scraper. In the other case the local tool-rejuvenation is documented by refitting of the last spall to a multiple burin (complex 26: Fig. 2.2.), made of a rarely used raw material type. The next large group of conjoined pieces links to the natural fragmentation of blanks. In the studied assemblage the flaking accidents with f racture lines, running along the morphological axe of the flake, similar to the 'Siret' pattern 2 5 occur exclusively on coarse-grained quartzite flakes (complexes 7a, 17a, 18a, 20a, 23 and 29). Based on the characteristic breakage pattern another flake of the same raw material was broken during the retouching works (complex 24), finally four blades of nutnmulithic chert, siliceous and reddish radiolarite pebble were fragmented according to a pattern (complexes 2a, 25, 30 and 31 2 6). 2 4 DOBOSI et al. 1989. 2 5 INIZAN et al. 1999, 156. 2 6 e.g: MARKÓ 2007: Fig 5. 2.
