The discovery of the timber platform in fen peat poses a number of questions relating to the local ecology and environment in which the structure was constructed and the character of local land use and economic subsistence. Pollen and plant macro-fossil studies were initiated at the outset of discovery of the Flag Fen structure. Special attention has been paid to the changing fen flora reflecting local hydrological changes as well as evidence for the vegetation and land use on areas of nearby dry land. In addition to the evidence for temporal changes in the environment obtained from ‘long peat cores’, sampling on a grid-square basis is being carried out on sands thought to have been floor covering. It is hoped that these will yield information on the use of the platform. Pollen data are now available from Flag Fen and the nearby Power Station site at Fengate. This short contribution seeks to provide a general summary of the vegetation and environment of the fen and its near terrestrial environs with data coming largely from the analysis of the ‘Mere Section’ (see Introduction, FIGURE 6 for location) although some reference will be made to the Fengate site (see Introduction, FIGURE6 for location). Results ofthe completed study will be presented in a forthcoming monograph.

The study of insect remains, particularly Coleoptera (beetles), from archaeological deposits can give palaeoecological information on a wide range of aspects of site environment and human activity. Some of this supplements other lines of evidence, for example the results of pollen analysis and the identifications of macroscopic plant remains can be supported by the food plant requirements of the more host-specific phytophagous Coleoptera present in a sample. However, there are other aspects of site ecology for which insects provide the main evidence. In particular, there are distinct faunas associated with various categories of structural, stored and decaying organic material.

The Flag Fen excavations of 1984 revealed a large number of oak timbers and it was immediately apparent that there was potential for dendrochronology. It was suggested that the work should initially comprise a major element in the Ph.D research of Nigel Holman, at the University of Cambridge. The direction of that research, however, moved away from Flag Fen, and it was not until 1987 that the Trust became aware of the situation. It was then decided, after much urgent consultation, that the best approach to the Flag Fen tree-ring analysis would be to undertake the study in-house (Pryor 1991: 111). The intention has been to keep costs low and (in view of the accumulated backlog) to produce as many measured samples as was reasonably possible. The dendrometer and computer facilities were initially provided by the Trust and the full implementation of the scheme has been supervised by the author. This work has been carried out under the supervision of the English Heritage funded Sheffield Dendrochronology Laboratory – to whom special thanks are due.

The method of excavating the platform is ostensibly simple, but every stage must be completed to a consistently high standard, particularly the maintenance of records. The amount of data generated every season is vast. For every level the procedure is the same: the wood is exposed and the surrounding peat removed so that the individual pieces are as clearly defined as possibIe whilst still in situ. The cleaned level is photographed from fixed positions in the roof of the excavation shelter. A plan is drawn, partly using the vertical photographs, but it is finished in the excavation, and amended where necessary during the lifting process. The wood is then lifted using teams of two or three excavators, at least one of whom needs to be very experienced. The wood is either allocated a number as it is lifted, or is sampled by metre square and spot height (O.D.). Provenance data, including grid reference, height O.D., and provisional measurements and description are noted on the field sheets.

The bone assemblages from the Flag Fen platform itself and from the contemporary post alignment revealed at the Fengate Power Station site are small (respectively c. 150 and c. 600 identified pieces), but may nonetheless shed some light on the nature of human activity at the two sites. For obvious reasons, this study is principally directed towards understanding the circumstances of bone deposition rather than towards a reconstruction of local animal husbandry.

The discovery of a small scatter of pottery sherds from around and to the south of the timber rows on the Fengate site and of more pottery from the Flag Fen platform enables us to undertake a comparative study which might throw some light on the history of activities associated with these two structures.

The assemblages discussed here are small, 103 sherds from Fengate and 293 sherds from the Flag Fen platform, and the procedures by which these assemblages were recovered differ. The Fengate pottery was excavated in the course of uncovering either the timbers or the metalwork. Consequently, the distribution of known material derives not from a detailed examination of the entire area but from the position of the timbers or the metal objects and where the position of the latter was indicated by metal-detector readings. These recovery procedures present us with some difficulties if we wish to seek a direct comparison between this assemblage and the material from the Flag Fen platform. The platform is being excavated fully and under closer archaeological scrutiny; the contrast is well illustrated by the fact that sherds of less than 20 mm in diameter are absent from the Fengate site but comprise about 17% of the Flag Fen assemblage.

For the purpose of discussion the platform and the alignment must be regarded as two separate sites which might in the future prove to have been related. Taken together the metal objects from the two sites represent a remarkable collection, not only by their context and numbers but also by the range of metals represented. Whilst the majority of the objects are in copper alloy (almost certainly bronze) there are also objects in iron, a white metal (some definitely tin, others, lead) and a single gold object.

On 14 February 1992, all the archaeological participants in the project met for a seminar at Flag Fen. Drafts of the papers published in this Special section were presented by their authors, and after each paper there was a lively general discussion.

This concluding paper does not pretend to be a synthesis of that discussion, because good archaeology cannot be done by committee. Having said that, I have tried to include the views of all dissenting voices, to give balance to the interpretations offered here. I have also tried to credit those individuals who put forward some of the excellent ideas that came from the meeting. In the discussion that follows ‘pers. comm.’ should be read as a coded reference to the St Valentine’s Day Seminar.