K. Palágyi Sylvia szerk.: Balácai Közlemények 2005/9. (Veszprém, 2005)

T. LÁNG, ORSOLYA: Control Excavations in the Territory of the Civil Town of Aquincum: the so-called „Testvérhegy-villa"

was also discovered in which glass fragments and charcoal were found. Late Roman coins and a cikkada-type brooch from the erosion layer above, speak for the Late Roman use of this area. 3. Following a landslide from the hill above, the building was renovated and enlarged. The courtyard, which had been filled with mud was renewed as a ponded clay floor. Impressions of wooden frames resting on the floor speak to some kind of industrial activity. Two steps lead toward the east, where new rooms with pounded clay and stone paved floors were added at this time. How these rooms were used has not yet been determined although the easternmost room with a stone pavement suggest that some activity using fire took place here. 4. Another period of earth movement resulting from erosion required the next renovation. The floors of the rooms were renewed (pounded clay floor) and an U-shaped, burnt clay bench were created in the eastermost room with the stone pavement (Fig. 13). Based on archaeomagnetic investigations carried out by R Márton of the the geology department of Eötvös Loránd University, it had been repeatedly burnt at a temperature of approximatly 500C. It may have been a bread-baking oven. Building debris comprising charcoal and daub was also found in the neighbouring rooms. The following phases could not be reconstructed and the building slowly disappeared under continuously developing landslip. During the Middle Ages it was not used but the precise datation of the phases will only be possible after evaluation the find material. An internal, approach road was also discovered in this area. It may have connected the buildings on the slope to the main diagonal road running next to the above-mentioned out-building (Fig. 14). The southwest-northeast oriented road was 2 meters wide and had four layers related to erosion of earth from the slopes. The cleaning of the topmost layer yield a „trumpet" brooch. The road led to a gate, opening on the north-south oriented fencing wall which must have bordered the estate from the diagonal road. The wall was 50 cm thick and was made without any binding material, though an imbrex was inserted to carry off water. It was built directly on the ground and its original height could not be determined. The gate itself was later walled-in for some, as yet, unknown reasons and a stone-covered channel was inserted here, which probably carried water away from the buildings on the slope. Due to modem levelling at the site only a small section of this chanel could be observed. Perpendicular to the above-mentioned fence there was another fencing wall to the north. This suggests that this represents the northeast comer of the settlement. This hypothesis is further supported by the fact that no remains were found north of this latter fence.

10 The debris from the fencing wall clearly indicated the way the original area sloped: from this point the area began to slope upwards to the north. In the zone of the entrance, but already outside the settlement, traces of a dirt road and a pit were found. The latter was filled with charcoal and fragments of Roman pottery, including the handle of a glazed casserole. The junction of the main diagonal road and the road leaving the villa was not found because of heavy disturbances in the area A broad Medieval (probably dirt) road ran