Redő Ferenc: Katonák Sallában (Zalalövő öröksége 1. Zalalövő, 2003)
traces and deeper post holes can be found. Several features were attributed to this layer in the former publications of the excavations as the NW-SE directed ditch of a "V"-shaped cross-section of the road in site F, which was about 4 m broad and 1.5 m deep (Fig. 4) and the building with timber foundation NW of the ditch, which was at least 26 m long and 10 m broad. In some of the 20-35 cm broad timbers, postholes were dug at even distances. We found dividing walls in the NW side of the building, which used to be divided along the longitudinal axis and had a central corridor. The dividing walls separated 2-2.5 m broad rooms. The building was interpreted as a military barrack after its analogues (Fig. 6). North of it, three parallel timber traces were found in the length of about 18 m strengthened by a row of postholes, arranged in the same direction as the barrack. This construction was certainly the fragment of a defence work, a fortified wall or fence (Fig. 5). The size of these constructions raised in a single structural order surpasses that of the timber structure buildings unearthed before or after at the same site. Features contemporary to the above ones were unearthed SE of the road as well. They were also timber traces and postholes running parallel to the ditch, and, as far as it could be observed on the small surface, they did not belong to a building. It is also important to mention that the band SE of the ditch, the track of the later road, was not paved with pebbles at the time of the existence of the ditch and did not contain any find or feature, it was practically empty. Most probably, it was the foot of the rampart that was pulled down later when the road was built. These features have two major groups the orientation of which is identical within the group but different between the two groups. The orientation of the barrack building and the fortified fence wall diverged by a few degrees from that of the ditch and the timber construction SE of it (Fig. 7). This difference is reinforced by stratigraphie observations, which registered overlapping settlement features within this layer without having separate floor levels. The proximity of the S corner of the barrack building and the ditch makes it obvious that the ditch was dug out later. There were 31 coins in the entire territory; of which the exact place of occurrence could be determined and that can be dated from the start of coin circulation in the settlement to the middle of the 1 st century AD. Twenty-six other coins were minted between 50 and 85.
