INTRODUCTION: BACKGROUND AND METHODOLOGIES A 10-ha site discovered on Langwood Ridge, a spur of Chatteris Island, during the Fenland Survey is amongst the largest open Iron Age settlement scatters known in Britain (Site 26; TL4180 8520). Correlating with a series of linear cropmarks and the possible outline of a stone building on aerial photographs, in the course of the survey a distinct Roman component was also identified. 1 This multi-period settlement, whose 'in-fen' situation is directly comparable to Stonea, was assessed by the Cambridge Archaeological Unit of the University of Cambridge in the winter of 1993 in conjunction with the Fenland Management Project (FMP). 2 Nine other Iron Age/ Roman scatters are known upon the ridge, including two within the immediate vicinity (Sites 24 and 25; FIG. 2). Moreover, cropmark complexes, respectively involving large field boundaries and more minor paddocks, lie just northeast of the main scatter (Sites 12 and 13). The nearest, the Site 12 field-system, was also trenched in conjunction with the Site 26 fieldwork. Given these densities, one theme of the programme was that of site definition; it is difficult to know where each site begins and ends. As a result, in part, of not being able to adequately delineate the allocated scatter, the investigations grew and to provide broader landscape context neighbouring sites were also sampled (Sites 13, 14, 23-25, and 29). 3 The main Chatteris 26 spread clearly relates to a major complex. With dense and complicated archaeology, more than 450 features were recorded in the course of this programme. The finds densities were equally prolific and some 21,500 artefacts were recovered (e.g. 9,290 pottery sherds and 965 pieces of metalwork). In the main, the occupation phases date to both the Later Bronze/ Early Iron Age and Roman periods; arguably the major discovery of this programme was that the stone building proved to be Roman. Other periods were, however, also represented, including scatters of Neolithic/Bronze Age flintwork, and both a middle and more substantial Later Iron Age occupation presence. Iron Age and Roman Republican coins were also recovered. Faced with the daunting scale of the site, for pragmatic reasons it was approached in the spirit of a formally structured 'New Archaeology-style' exercise. Applying statistical sampling techniques, the aim was to broadly characterise the parameters of its sequence (for its possible scheduling)-1
