rummy writes, 'The heart and earliest part of the Gosbecks complex was a large native farmstead which had been remodelled several times. In its original form it consisted of a large trapezoidal enclosure protected by a substantial defensive ditch. Later (presumably Roman) periods of alterations are
represented by straight stretches of narrow ditch forming a series of secondary enclosures and internal subdivisions. In the later periods the perimeter ditch was of much smaller proportions than before, indicating that it was no longer defensive. It enclosed a bigger area. Traces of occupation in the form of pits and a possible hut are visible as cropmarks in the western corner of the trapezoidal enclosure. The system of trackways which surrounded the enclosure focused on what was presumably an entrance at the west corner'.<1>

Two small trenches were dug by Lt Col R.J. Appleby under the supervision of Mr M.R. Hull in 1949. One trench (6.4m long) was dug across the northern ditch of the trapezoidal enclosure and the other (4.5m long) across a smaller ditch c.120m to the north. Hull concluded that 'these ditches are pre-Roman in origin, but remained open late enough for Roman wares to get into them. The large ditch was the earlier and was, it seems, filled in before the smaller.'<1><2>

The trapezoidal enclosure, aligned roughly WNW-SSE, was defined by a large ditch, 5.5m wide x 2.5m deep. Crummy writes, 'At the bottom of the ditch was 0.4 m (1.5 ft) of 'rapid silt' which contained no finds. The fill above, separated from the rapid silt by a spit of yellow sand, was of 'darkish earth' and contained mainly Late Iron Age pottery of Sheepen types. Near the bottom of the main fill, a few sherds were discovered which were described as possibly Roman. Above these, at a depth of five feet, was a Langton Down brooch (CM 151.49, Hull's Corpus no 323) datable from perhaps as early as the mid
1st century BC to 60-70 AD (M R Hull's unpublished corpus of Roman brooches). Higher up in the fill, about half-way up, were some Gallo-Belgic sherds and a few fragments of South Gaulish terra sigillata. From this, there is little doubt that the ditch was silting up in the Late Iron Age and Roman periods and, assuming the 'uniform darkish earth' to be backfill, was levelled some time after the conquest. However, what is not clear is the date when the trapezoidal enclosure, and therefore presumably the farmstead, was set out.'<1>

The smaller ditch to the north was found to be 3.6m wide x 1.4m deep below the modern surface. Crummy records, 'It appears to form the north boundary of a small enclosure abutting the north ditch of the trapezoidal enclosure. It contained no 'rapid silt' and the fill was 'dark sandy loam' which, according to Hull, changed gradually to 'brown earth' higher up the fill. There was a distinct old land surface, above which until recent times the ditch had been visible as a slight earthwork. Compared with the ditch of the trapezoidal enclosure, the lower fill contained proportionally more Roman than native pottery and included pieces of mortaria and flagon indicating a post-conquest date. Like the other sectioned ditch, there is no clear evidence to show its date of construction, but the fact that the earthwork was not levelled in antiquity coupled with its relatively high proportion of Roman pottery suggests that the feature was late in the sequence of ditches.<1>

Geophysical survey on trapezoidal enclosure and area to the south was carried out in 2001/2 (ECC3782). However, the survey failed to clarify whether the southwest corner possessed an entrance to a farmstead. The survey clearly identified the enclosure and many other linear features that suggest ditched enclosures, together with pit-like features. The complexity of the enclosure pattern to the south implies use and re-use of the area over an extended period of time.<3> see also ECC2832.<4>