Savaria - A Vas Megyei Múzeumok értesítője 30. (2006) (Szombathely, 2007)

Régészet - Szilasi Attila Botond: Kelta település részletre Sárvár határában

Szilasi Attila Botond: Kelta település részlete Sárvát határában PARTS OF A CELTIC SETTLEMENT ON THE EDGE OF SÁRVÁR Preliminary investigation of a site along the new Route 84 bypassing Sárvár to the S was made on May 9—fune 11, 2002. Excavation of an area of almost 6400 sq. m was directed by archaeologist Ildikó Katalin Pap. The finds were made in a well-defined band on the Wside of the part of Sárvár known as Móka-dűlő, on a small river terrace on WandEsides of an old natural course of the River Gyöngyös (Table 1). Finds allow only 18 of the 150 objects to be dated firmly to any stage in the Celtic period. Only a narrow E—W band was examined, not the whole settlement, whose N—S extent is still unknown, but the terrain data and excavation picture suggest it extended towards the brow of the hill. The finds at Object 92 (Table 24) from the earliest stage include classic marks of early LT Bl: vessels hand-formed and thrown at low speed on a wheel, with an S profile, retracted rim and straight neck (Tables 25, 26), and a group of pots of individual patterns, with incised decoration on the shoulders (Table 28: 2). The finds from the other early La Tine stage (Objects 63, 82 and 93 — Tables 15, 19, 29) point to a strong LT B2. The articles that can only be dated more generally are S-profile bowls with typically decorated rims. Also important are various undecorated and stamp-decorated situlae (Table 22: 5), which spread widely in that period (thanks to southern, Italian connections). There is further evidence of this early period in a large number of hand formed vessels. The one pointer to a kind of transition between early and late La Tene is Object 105 (Table 32). Here the finds include S-profile bowls of the classic LT B2 and a so-called Hunyady 3-type deep-welled bowl, whose real stage can be better dated to the beginning of LT Cl, but whose appearance has been observed at the turn ofLTB2 and LT Cl (Table 34: 5—6). Object 2 (a pottery — Table 2) has yielded classic LT C2 finds. This is supported by a short-bodied fibula (Table 4: 1) and individual stamping on the pots in the collapsed kiln (Tables 7, 8). The bowls and urn forms also point to a transition to the later period, which signifies the end of mid-La Tene. Evidence of the change from IT Сto D is embodied in Object 70 (Table 17), and in Object 84 (Table 23), where ceramic materials show both LT С and D forms. Various areas of the excavation yielded independently fragments of classic LT D vessels in secondary or disturbed layers, showing that life was extensive in the 1st century. But it is important to note, as a possible subject of later investigation, that E of the vanished bed of the Gyöngyös lies a vast 2nd-century indigenous Celtic settlement that may have had a similar structure in La Tene. Lt can be assumed that the local underlying inhabitants remained for centuries even after the Roman 'conquest', taking over the new customs to some extent. All the buildings are aligned E—W except Object 84 (NW—SE). They are sunk into the ground and oblong in shape with rounded corners and their sides slightly curved, some with a surrounding sill or steps. The base is flat. The N side of Object 70 shows signs of two steps, probably up to an entrance. It can be surmised that the other houses were entered at a similar place. Holes for posts have been found only in Objects 2, 84 and 82, on the E and W sides. This suggests a pavilion roof structure held up by purlins, i.e. the roof structure supported from the ground. In the other cases, only the pit structure can be seen, without pestholes. The post-frame houses resting on the surface (Objects 111—

16 and 120—21) are also aligned E—W. The walls and the posts and middle purlin supporting the roof show a classic form. The posts supporting a traditional gabled roof had sizeable diameters of 65—70 cm. Earlier writers regularly called the buildings sunk into the ground dwelling houses, but 252