In 2007 the Trust for Thanet Archaeology carried out an evaluation. Two Early Roman -2nd century features containing pottery were identified within the evaluation trench which may in fact form part of the same feature. Two other features may be of the same date but of a later stratigraphic phase were exposed within the trench. Neither feature contained datable finds. (1)

Further investigations were carried out in 2009 when The Trust for Thanet Archaeology excavated the site.The earliest feature encountered was a steep sided cut with a flat base, which appeared to be an irregular rectangular cellar with a rounded ramp on its southern edge. The cut was regular enough in plan on the north and eastern sides to suggest that the feature had originally formed a cellar measuring approximately 5m east to west and 4m north to south, possibly with a ramp at the south corner to allow access. The feature was probably the sunken element of a building that originally had a timber superstructure. Above chalky primary fills produced by weathering of the edges, were deposits of mixed clay with chalk fragments, interspersed with thin chalk bands marking a series of tips. A shallow deposit above the chalk indicated that later fills had been cut away by features of later phases.

Later a pit was cut through the filled early cellar, also in the Roman period. A foundation trench, cut to receive the stones of a right angled wall foundation, truncated the upper chalk capping of the pit. The right angled wall foundation set into the cut was constructed from a base course of massive natural flint nodules, some up to 0.5m in height, standing with their longest axis upright and butted together closely as their irregular shape allowed. The foundation course was overlaid with more regular rounded water-worn cobbles laid flat across width of the foundation stones in two irregular faces, measuring between 0.5 and 0.6m in width. The east west axis of the structure measured at least 2.2m and extended under the limit of excavation on the western side. The north south stretch measured 1.8m from the corner. The northern limit of the north south wall terminated against the steep face of the earlier cellar cut indicating that the foundation trench had been cut down into the fills of that feature. It is possible that the structure encountered was built as a deep foundation to some later building whose floor levels had been raised above the upper edge of the cellar cut. It is also probable that this could have been a formed as part of flint cobble lined cellar associated with a building founded well above the present artificially reduced level of the site.

In the north east area of the excavation another deep feature was excavated. This feature appears to have originated as an element of another sunken cellar, cut into the chalk geology below a superstructure built at the much higher elevation of Roman ground level now entirely lost through later truncation of the deposits on the site. The central void of the cellar was lined with a wall that was founded on the ledges located on the eastern and southern sides of the cut. Courses of rounded flint cobble were widely spaced between a matrix of chalk silt and clay packing. The structure should be interpreted as one where the flint cobbles acted as tempering within a clay mass, rather than a stone wall where the contact between the stones created the structural framework.

A third deep cutting into the chalk was located close to the eastern edge of Cellar 1. Measuring approximately 7.5m long on its east west axis and 5m wide on its north south axis the pit was a regular rounded rectangle in plan. This pit had steeply sloping sides breaking sharply at the bottom to form a reasonably flat base which was made slightly irregular where interfaces with the bedding planes of the hard chalk had been reached when the cellar was originally formed. The western edge of the feature had removed part of the edge and backfills Cellar 1 demonstrating that it was of a later structural phase. It may have been contemporary with the building complex represented by Structure 1 and Cellar 2.

An irregular rectangular pit had been cut through the demolition deposits filling Cellar 2. The cut had symmetrical, steeply angled sides, which broke sharply toward the base to form the very steep sides of an irregular slot. This pit was probably cut after the earlier building had been demolished, robbed of its stone and levelled so that it could not be seen from the surface. The base of the pit had been excavated to a depth where a dump of building stones laid in the robber cuts along the western edge of the cellar was exposed, Where the stones had been cleared, a thick layer of ashy material containing oyster, whelk and mussel shells as well as animal bones, Iron nails and other debris had been dumped in the bottom of the pit. Interspersed with this material were large sherds of many pottery vessels, many of which must have been nearly complete when they were deposited. The pottery ranged from a very large comb-decorated storage jar, a well preserved mortarium, dishes, bowls, flagons and cups. At least one decorated Samian bowl was present along with a small undecorated cup, together with dishes in Black-Burnished ware fabrics. Rather than a deposit of rubbish it appears that a group of pots from a kitchen or storeroom had been tipped into the hole along with burnt debris and midden material.

The features were found to belong to Roman cellared buildings. Lengths of flint cobble walling were built within the cellar cuts, possibly to support a building over the cellar. The cellars were backfilled during the Roman period and a pit was found cutting this deposit. Stones were dumped into this pit along with domestic refuse. Amongst this was a large deposit of Roman pottery. The vessels must have been largely complete at the time of deposition. Vessels included a very large comb-decorated storage jar, a well preserved mortaria, dishes, bowls and cups, including at least one Samian bowl.

At least three phases of occupation are indicated by the stratigraphic relationships recorded on the site. There are clear discontinuities in the sequence that show major changes in the pattern of occupation, these are indicated by the pit cut through the fills of Cellar 1, the truncation of Cellar 1 by Cellar 3 and the late pit cut through the fills of Cellar 2. Intensive construction activity and restructuring of the layout of the site took place within a short period time within the Roman period from the late 2nd century AD to the middle of the 3rd century. It is likely that this level of activity does not reflect an unusually dynamic society, but intensive replacement and remodelling within one spatial context. (information summarised from sources)(2-3)

<1> Trust For Thanet Archaeology, 2007, Land adjacent to Bleak House, Fort Road, Broadstairs, Kent: Archaeological Evaluation Report (Unpublished document). SKE12984.

<2> Trust for Thanet Archaeology, 2010, Land adjacent to Bleak House, Fort Road, Broadstairs: Interim archaeological report (Unpublished document). SKE18058.

<3> Trust for Thanet Archaeology, 2010, Land adjacent to Bleak House, Fort Road, Broadstairs: Archaeological Assessment Report (Unpublished document). SKE24843.