Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire 9781407316574, 9781407323275

Table of contents :
Cover
Title page
Copyright
Acknowledgements
Contents
List of Plates
List of Figures
List of Tables
List of Graphs
List of Contributors
Abstract
Introduction
The Results of the Excavation
The Artefactual and Environmental Evidence
Discussion
Summary and Conclusions
Bibliography

Citation preview

Andrew A. S. Newton studied archaeology at the University of Bradford, carrying out an MPhil on the relationship between politics and archaeology at the same institution. He has worked for Archaeological Solutions since 2005, contributing to numerous post-excavation projects in East Anglia and the south-east of England.

‘This is a comprehensive and thorough report of a fascinating site. It adds to a growing knowledge of the Iron Age of Cambridgeshire and the fen-edge, facilitating the construction of a detailed understanding of prehistoric settlement and life in the region.’ Dr Jody Joy, Senior Curator (Archaeology), Museum of Archaeology and Anthropology, University of Cambridge ‘Provides a detailed, well written and engaging introduction to the site, with a thorough overview of the fen edge environment the site sits within [and] a wealth of specialist information. … A surprisingly engaging read.’ Michael Bamforth, Project Manager, POSTGLACIAL project, University of York

BAR  640  2018  NEWTON  IRON AGE FEN-EDGE SETTLEMENT AT BLACK HORSE FARM, SAWTRY, CAMBRIDGESHIRE

Black Horse Farm is situated on the Cambridgeshire fen-edge. During the Iron Age and early Romano-British period it occupied a low promontory reaching out into the surrounding wetland. This volume describes the archaeological excavation of the site and the Iron Age settlement and Romano-British activity that was recorded there. The wetland of the fen would have been a prominent part of everyday life at Black Horse Farm and the book examines the way in which the site’s inhabitants utilised and exploited it. Fluctuations between dry and damp conditions were also a prominent aspect of life at this marginal location and the later sections examine how the population responded to these conditions. The book examines themes including the organisation of space within the roundhouse, the role of ditches and banks as flood defences versus their social and defensive function, and offers alternative interpretations for some commonly observed features at contemporary sites.

B A R

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Andrew A. S. Newton

BAR British Series 640 2018

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Andrew A. S. Newton with contributions by Beta Analytic Inc., Jane Cowgill, Nina Crummy, Julia E. Cussans, Val Fryer, Andrew Peachey, Ruth Pelling, Carina Phillips, Rob Scaife and Maisie Taylor Illustrations by Kathren Henry, Charlotte Davies and Caroline George

BAR British Series 640 2018

Published in by BAR Publishing, Oxford BAR British Series Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire © Andrew A. S. Newton Site location plan. The Author’s moral rights under the UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reser ved. No par t of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any for m of digital for mat or transmitted in any for m digitally, without the written per mission of the Publisher.

ISBN 9781407316574 paperback ISBN 9781407323275 e-format DOI https://doi.org/10.30861/9781407316574 A catalogue record for this book is available from the British Library

BAR titles are available from: BAR Publishing Banbury Rd, Oxford, [email protected] + ( ) + ( ) www.barpublishing.com

,

Acknowledgements The Black Horse Farm, Sawtry project was undertaken by Archaeological Solutions Ltd (AS). The Phase 1 project was commissioned and funded by Scandstick UK Ltd. Michael Eyres Partnership and Consultancy monitored the project on behalf of Scandstick UK Ltd. The Phase 2 project was commissioned and funded by Davies Street (Sawtry) Ltd. Post-Excavation analysis and Publication was funded by Tesco Pension Trustees Ltd. AS wishes to express their thanks to these organisations. AS would like to thank Kasia Gdaniec of Cambridgeshire County Council Historic Environment Team for her input and advice. AS would like to thank Christopher Evans of the Cambridge Archaeological Unit for permission to recreate images appearing in the volume Process and History. Prehistoric Communities at Colne Fen, Earith. The Project was managed by Claire Halpin. Fieldwork was conducted by Iain Williamson and Phil Weston. Finds were co-ordinated by Claire Wallace. Illustrations are by Charlotte Davies, Kathren Henry, and Caroline George.

Contents List of Plates ...................................................................................................................................................................... vi List of Figures ................................................................................................................................................................... vii List of Tables...................................................................................................................................................................... ix List of Graphs..................................................................................................................................................................... x List of Contributors .......................................................................................................................................................... xi Abstract............................................................................................................................................................................. xii Chapter 1. Introduction ..................................................................................................................................................... 1 I. Introduction ................................................................................................................................................................ 1 II. Background ................................................................................................................................................................ 3 III. Archaeological and Historical Background ............................................................................................................... 3 IV. Topography, Geology, Soils and the Fenland Environment ....................................................................................... 6 Chapter 2. The Results of the Excavation.......................................................................................................................11 I. Phasing ..................................................................................................................................................................... 11 II. Phase 1: The Middle Iron Age .................................................................................................................................. 11 III. Phase 2: The Later Iron Age ..................................................................................................................................... 41 IV. Phase 3: Romano-British Activity; the Mid 1st Century AD .................................................................................... 51 V. Phase 4: The Alluvial Deposits ................................................................................................................................ 54 VI. Phase 5: Roman Activity after Alluvial Deposition ................................................................................................. 56 VII. Phase 6: Roman Ploughsoil ...................................................................................................................................... 58 VIII. Activity Surrounding the Settlement; Evidence from Trial Trench Evaluation ....................................................... 59 Chapter 3. The Artefactual and Environmental Evidence ........................................................................................... 63 I. The Pottery (by Andrew Peachey) ............................................................................................................................ 63 II. The Daub and Ceramic Building Materials (by Andrew Peachey) .......................................................................... 75 III. The Slag (by Jane Cowgill) ...................................................................................................................................... 78 IV. The Small Finds (by Nina Crummy with Maisie Taylor) ......................................................................................... 81 V. The Animal Bone (by Carina Phillips and Julia E. Cussans).................................................................................. 83 VI. The Human Bone (by Carina Phillips) .................................................................................................................... 94 VII. The Plant Macrofossils (by Val Fryer and Ruth Pelling) ....................................................................................... 104 VIII. Pollen Analysis (by Rob Scaife) ............................................................................................................................. 106 IX. Radiocarbon Dating (by Beta Analytic Inc. and Andrew A. S. Newton)................................................................. 107 Chapter 4. Discussion .....................................................................................................................................................111 I. The Environment .....................................................................................................................................................111 II. The Agricultural Economy ..................................................................................................................................... 113 III. The Domestic Structures ........................................................................................................................................ 114 IV. Enclosures .............................................................................................................................................................. 120 V. A Shift in Domestic Occupation............................................................................................................................. 125 Chapter 5. Summary and Conclusions ........................................................................................................................ 129 Bibliography ................................................................................................................................................................... 131

v

List of Plates Plate 1. The Black Horse Farm site under excavation ......................................................................................................... 1 Plate 2. View north across the interior of the Iron Age enclosure during excavation of Roundhouse 3 ........................... 10 Plate 3. Roundhouse 1 during excavation. In this view to the east-north-east, Gullies F2027 and F2174 are visible in the foreground, while in the top right, Phase 4 Gully F2309, which cut clay floor L2270, is visible ............... 15 Plate 4. View north across excavation area AS857 showing, in the near centre of the photograph, investigation of the features and deposits forming Roundhouse 2 .................................................................................... 18 Plate 5. A view north-east across Roundhouse 3 during excavation.................................................................................. 20 Plate 6. Wall Footing F2477 following removal of overlying deposits ............................................................................. 25 Plate 7. A view north-west across Roundhouse 4 during excavation. The inner and outer ring ditches are clearly visible. This photograph illustrates the wet conditions and high levels of groundwater encountered during excavation .......................................................................................................................................... 38 Plate 8. A view across Roundhouse 6, looking to the west ................................................................................................ 39 Plate 9. Removal of Layers L2459 and L2420 from Roundhouse 3, viewed from the north-west. Wall Footing L2477 is visible in the right foreground ............................................................................................................... 50 Plate 10. View west across excavation area AS857 showing, in the foreground, the southern end of the delineated trackway ........................................................................................................................................................... 56

vi

List of Figures Figure 1. Site location plan .................................................................................................................................................. 2 Figure 2. Detailed site location plan .................................................................................................................................... 5 Figure 3. Iron Age land surface............................................................................................................................................ 8 Figure 4. Southern excavation area (AS857) all features plan .......................................................................................... 12 Figure 5. Northern excavation area (AS1111) all features plan ......................................................................................... 13 Figure 6. Phase 1 plan ........................................................................................................................................................ 14 Figure 7. Roundhouse 1 and associated features ............................................................................................................... 16 Figure 8. Structure 3012 .................................................................................................................................................... 17 Figure 9. Roundhouse 2 ..................................................................................................................................................... 19 Figure 10. Roundhouse 3 after complete removal of L2494 ............................................................................................. 21 Figure 11. Roundhouse 3 after partial removal of L2494 .................................................................................................. 22 Figure 12. Roundhouse 3 after removal of L2459 ............................................................................................................. 23 Figure 13. Roundhouse 3 after removal of L2420 ............................................................................................................. 24 Figure 14. Features forming and associated with Roundhouse 3 ...................................................................................... 26 Figure 15. Roundhouse 3, distribution of finds within F2324 ........................................................................................... 28 Figure 16. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature) ......... 29 Figure 17. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature) ......... 30 Figure 18. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature) ......... 31 Figure 19. Distribution of animal bone by weight (g) within internal features of Roundhouse 3 ..................................... 32 Figure 20. Distribution of animal bone by weight (g) within internal features of Roundhouse 3 ..................................... 33 Figure 21. Distribution of animal bone by weight (g) within internal features of Roundhouse 3 ..................................... 34 Figure 22. Burial 2374 ....................................................................................................................................................... 35 Figure 23. Roundhouse 4 plan and sections....................................................................................................................... 36 Figure 24. Distribution of finds in the ring gullies of Roundhouse 4 ................................................................................ 37 Figure 25. Roundhouse 5 and 6 ......................................................................................................................................... 40 Figure 26. Phase 2 plan ...................................................................................................................................................... 42 Figure 27. Boundary ditches F2325, F2378, and F2808.................................................................................................... 43 Figure 28. Boundary ditches F2152 and F2226 ................................................................................................................. 46 Figure 29. Skeleton 2332 ................................................................................................................................................... 48 Figure 30. Burial F2985 ..................................................................................................................................................... 49 Figure 31. Phase 3 plan ...................................................................................................................................................... 52 Figure 32. Structure S2273 ................................................................................................................................................ 54 Figure 33. Delineated trackway ......................................................................................................................................... 55 Figure 34. Phase 5 plan ...................................................................................................................................................... 57 Figure 35. Plan showing the extent of Phase 6 deposit L2002 .......................................................................................... 59 Figure 36. Trial trenches and excavation area AS857........................................................................................................ 60 vii

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Figure 37. Trial trenches and excavation area AS1111 ...................................................................................................... 61 Figure 38. Pottery............................................................................................................................................................... 70 Figure 39. Pottery............................................................................................................................................................... 72 Figure 40. Spatial distribution of small finds .................................................................................................................... 80 Figure 41. Small finds ........................................................................................................................................................ 81 Figure 42. The wooden board or paddle ............................................................................................................................ 82 Figure 43. Distribution of significant deposits of bone...................................................................................................... 92 Figure 44. Comparison of the roundhouses from Black Horse Farm .............................................................................. 115 Figure 45. Comparison of the Iron Age enclosure system at Black Horse Farm with those at Site 1, Colne Fen and Haddenham Sites V and VI (after Evans 2013, figs. 5.46 and 5.47; Evans and Hodder 2006, 270) ........................ 124

viii

List of Tables Table 1.Simplified deposit model for excavation area AS857 ............................................................................................. 9 Table 2. Simplified deposit model for excavation area AS1111 .......................................................................................... 9 Table 3. Summary of phasing ............................................................................................................................................ 11 Table 4. Finds recovered from Ditches F2325, F2738 and F2808..................................................................................... 44 Table 5. Quantification of pottery in phased groups by sherd count, weight (g) and R.EVE ............................................ 63 Table 6. The distribution of fabrics by % sherd count in phased groups and the total assemblage ................................... 65 Table 7. Quantification of form types in phased groups by R.EVE/minimum no. of vessels ........................................... 67 Table 8. Distribution of sherds in Ceramic Group 1 .......................................................................................................... 68 Table 9. Distribution of sherds in Ceramic Group 2 .......................................................................................................... 68 Table 10. Distribution of sherds in Ceramic Group 3 ........................................................................................................ 71 Table 11. Distribution of sherds in Ceramic Group 4 ........................................................................................................ 73 Table 12. Distribution of sherds in Ceramic Group 5 ........................................................................................................ 73 Table 13. Distribution of sherds in Ceramic Group 6 ........................................................................................................ 73 Table 14. Distribution of sherds in Ceramic Group 7 ........................................................................................................ 74 Table 15. Distribution of sherds in Ceramic Group 9 ........................................................................................................ 74 Table 16. Quantification of daub and CBM by phase ........................................................................................................ 75 Table 17. Quantification of daub in Phase 1 feature groups ............................................................................................. 76 Table 18. Quantification of daub in Phase 2 feature groups ............................................................................................. 76 Table 19. Quantification of daub in Phase 3 feature groups ............................................................................................. 77 Table 20. Quantification of daub in Phase 4 feature groups ............................................................................................. 77 Table 21. Quantification of daub in Phase 5 feature groups .............................................................................................. 77 Table 22. Summary of the count and weight of the slag types by context......................................................................... 79 Table 23. NISP by Phase .................................................................................................................................................... 84 Table 24. MNI for domestic mammals by phase ............................................................................................................... 85 Table 25. Butchery type by species and phase ................................................................................................................... 87 Table 26. Withers heights for sheep/goat, cattle and horse by phase ................................................................................. 90 Table 27. Partial skeletons and skulls ................................................................................................................................ 91 Table 28. Key archaeobotanical samples selected from Phase 1 ....................................................................................... 98 Table 29. Key archaeobotanical samples selected from Phase 2 ..................................................................................... 100 Table 30. Key archaeobotanical samples selected from Phase 3 ..................................................................................... 102 Table 31. Archaeobotanical sampling of Phase 4 deposit L2060 .................................................................................... 105 Table 32. Key archaeobotanical samples selected from Phase 5 ..................................................................................... 105 Table 33. Pollen profiles .................................................................................................................................................. 108 Table 34. Calibration of radiocarbon age to calendar years............................................................................................. 108 Table 35. Estimate of labour (in days) required to excavate enclosure ditches at significant Iron Age sites in Cambridgeshire using Startin’s (1982) estimate of 5.43 – 8.15m³ for a four person team per day (after Evans 2003, tables 70 and 71) ......................................................................................................................................... 123 ix

List of Graphs Graph 1. Total weight (g) of pottery recovered from Roundhouse 3, plotted by quadrant................................................ 35 Graph 2. Total weight (g) of animal bone recovered from Roundhouse 3, plotted by quadrant ....................................... 35 Graph 3. % NISP for principal domestic mammals by phase ............................................................................................ 85 Graph 4. Cattle survivorship based on tooth wear data from Phase 1 ............................................................................... 86 Graph 5. Sheep/goat survivorship based on tooth wear data for the 5th-2nd centuries BC and late 1st century BC-mid 1st century AD. .................................................................................................................................... 87 Graph 6. Sheep/goat survivorship based on bone fusion data for Phase 1. Age stages follow those given by O’Connor (1989, fig. 28) .............................................................................................................................................. 88 Graph 7. Percentage of butchered bone fragments by phase ............................................................................................. 88 Graph 8. Butchery by species, all phases combined .......................................................................................................... 89 Graph 9. Pollen profile from Ditch F2738 .......................................................................................................................109 Graph 10. Pollen profile from Buried Romano-British Soil L2002................................................................................. 110

x

List of Contributors Beta Analytic Inc. Radiocarbon dating laboratory Jane Cowgill Independent Industrial Residue Specialist Nina Crummy Independent Small Finds Specialist, Colchester Julia E. Cussans Osteoarchaeologist, Archaeological Solutions Ltd Charlotte Davies Former Graphics Officer, Archaeological Solutions Ltd Val Fryer Independent Environmental Specialist Caroline George Former Graphics Officer, Archaeological Solutions Ltd Kathren Henry Graphics Manager, Archaeological Solutions Ltd Andrew A. S. Newton Assistant Projects Manager (Post-Excavation), Archaeological Solutions Ltd Andrew Peachey Pottery, CBM and Worked Stone Specialist, Archaeological Solutions Ltd Ruth Pelling Independent Environmental Specialist Carina Phillips Osteoarchaeologist, formerly of Archaeological Solutions Ltd Rob Scaife Environmental and Palynological Specialist, McDonald Institute for Archaeological Research Maisie Taylor Independent Ancient Wood Specialist

xi

Abstract Several phases of archaeological work, conducted by Archaeological Solutions Ltd, at Black Horse Farm, Sawtry, Cambridgeshire, have revealed middle Iron Age to early Romano-British activity representing a small, at least partially enclosed, settlement, followed in later phases by enclosures and systems of land division used for agricultural purposes. During the Iron Age, the site was situated on the fen edge and occupied a low promontory reaching out into the fen. The history of occupation at the site is punctuated by alluvial events, associated with changes in the local fenland environment and it is likely that occupation was also hampered by a generally high watertable.

diminution of activity with each episode of flooding until it became incorporated into a wider area of agricultural land in the generally drier Romano-British period. Due to this history of inundation, Black Horse Farm is a very well preserved site. Located along the Iron Age fenedge, and effectively sealed by a series of alluvial events, it has provided a rare opportunity to investigate a domestic Iron Age settlement that has escaped the worst impacts of modern deep ploughing. Although the architectural components and artefactual assemblages recorded at Black Horse Farm are by no means unique, it is the level of preservation that makes the site of particular significance; indeed, this makes the site an exemplar of its type. Of particular significance is the preservation of Roundhouse 3 and the detailed ‘house-history’ that can be constructed from a sequence of architectural footings, floor layers, internal fixtures, and artefact distributions and which offer a notable opportunity to examine the way in which domestic space was structured and arranged in the Iron Age.

The initial phases of occupation were brought to an end by flooding of the site during the later Iron Age, represented by an alluvial deposit covering a large part of the southern area of excavation, but domestic activity appears to have been relocated slightly earlier, perhaps due to an increasingly high water table. Some aspects of these early phases may represent the responses of the site’s inhabitants to the increasingly wet conditions. Following this, domestic habitation appears to have been moved elsewhere, although the artefactual assemblage suggests that it must have remained nearby. The part of the site subject to excavation appears to have been used only for agricultural purposes in these later phases but a notably drier interlude immediately following the late Iron Age inundation allowed for apparent horticultural cultivation to be carried out here; a marked departure from the agricultural regime of earlier phases. The subsequent history of the site sees the gradual

More broadly, the site serves to highlight the significance of fenland archaeology and the potential for well-preserved sites in this landscape. It also represents an excellent opportunity to understand the Iron Age population’s relationship with the natural environment and why the fen edge was deemed a suitable location for settlement. Some responses to the natural environment may appear to be counter-intuitive to a modern mindset but these may have been subject to, or influenced by, other pressures which remain archaeologically invisible.

xii

1 Introduction

Plate 1. The Black Horse Farm site under excavation

I. Introduction

posits is, however, not just in their effect on the population who lived through the events that led to their deposition. These deposits, especially the later ones, have effectively sealed much of the site, contributing to its preservation and helping it to escape from the worst impacts of modern deep ploughing. French (1992, 730) has noted that alluviated fen-edge landscapes such as this are of notable importance due to the potential of the alluvial deposits to seal and protect the underlying archaeology and, as the landscapes are sealed relatively intact, they allow the best version of the complete picture of human land-use to be discovered in one area; waterlogging of such sites often provides an excellent degree of preservation of organic materials. The southern part of the site, and Roundhouse 3 in particular, was preserved, and therefore recorded, in notable detail providing detailed information about the way in which the building and its immediate surroundings developed. Black Horse Farm serves to highlight the significance of fenland archaeology and the potential for well-preserved sites to exist in this landscape.

Archaeological Solutions Ltd, in its former guise as the Hertfordshire Archaeological Trust, first became involved in archaeological work at Black Horse Farm in Sawtry, Cambridgeshire in 2002, compiling an archaeological desk-based assessment as part of a planning application for development of the site. Since then, two phases of trial trench evaluation, two phases of archaeological excavation, and a programme of post-excavation analysis have been undertaken. This archaeological work has revealed settlement activity at this location beginning in the later part of the middle Iron Age and stretching into the early Romano-British period. Black Horse Farm is located on the edge of Sawtry Fen. Although the extent of the fen has fluctuated over time, during the period of occupation represented by the archaeology it appears to have gradually drawn closer. A series of alluvial deposits, laid down mostly later in the history of the site, represent flooding events which appear to have led to changes in the way that the site was used, or at least that part of it viewed through the windows afforded by the areas subject to excavation. The significance of these de-

This document serves to describe the archaeology recorded at Black Horse Farm and to characterise the activity that was taking place at the site during the later Iron Age and 1

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 1. Site location plan

2

Introduction evaluation of this part of the site conducted in 2006 (Doyle and Weston 2006; Figs. 1 & 2). This evaluation mapped the northern extent of the former promontory of higher ground on which the Iron Age settlement activity was recorded to the south. Sparse activity of prehistoric date was recorded during the evaluation in the southern part of this area. This comprised traces of an Iron Age roundhouse not unlike those to the south, identified in Trench 10 at the point where the promontory began its northern descent, and linear features representing possible boundary ditches. Further north, at the northern edge of the promontory on which this activity was recorded, alluvial deposits associated with the fen were identified.

early Romano-British periods. The recorded archaeology is put into regional context and compared to relevant sites elsewhere in eastern England. The fen-edge location and the history of inundation at the site provide a major theme in the understanding of the site and, throughout, the relationship of the fenland environment and the prevailing damp conditions are considered in respect to the development, layout, and functioning of the settlement. II. Background Black Horse Farm lies just east of Sawtry, which is located about 25km south of Peterborough and 14km north-west of Huntingdon. The site lies 750m south-east of the present village of Sawtry, on low ground, that slopes away from the old Great North Road, towards Sawtry Fen to the east (see Figs. 1 & 2). The site occupies a position just east of the course of the old Great North Road (B1043 and A1(M)), formerly Roman Ermine Street. At the time of excavation, the site comprised 4.1ha of arable and vacant land. The central part of the site was crossed by a large open drain running south-west to north-east. Overhead power cables traversed the southern part of the site from south-east to north-west. It is bounded to the east by the Middle Level Catchwater Drain and to the south by a further drainage ditch, beyond which lies arable land and the A1 southbound access roundabout. To the north, the site is demarcated by Black Horse Drain, orientated west-southwest to east-north-east.

III. Archaeological and Historical Background The known archaeology and history of the Sawtry area Known prehistoric sites are scarce along this part of the fen edge. Many Mesolithic trees, or ‘bog oaks’, have been recovered from Sawtry Fen (Hall 1992, 33), indicating a wooded, dry-land landscape in this period. Flint concentrations indicate Mesolithic and Neolithic human activity in the area of Wood Walton, which lies on a glacial gravel-topped clay promontory c. 5km west of the Black Horse Farm site. Further Neolithic occupation of the area is represented by several stray flint implements of this date recorded by J R Garrood (1937a). Further examples of such items were recorded in the area by the Fenland Survey.

Archaeological work at this site comprised two separate areas of open-area excavation (Figs. 1 & 2). The more southerly of these areas (AS857) was excavated by AS between November 2004 and January 2005 and was focussed on an area of dense archaeological activity, identified within Trenches 2, 5, 5B and 11 of the evaluation which preceded it (Eddisford, O’Brien and Williamson 2004). The evaluation revealed significant archaeological features of later prehistoric date, sealed and well-preserved beneath thick layers of later alluvium. Evidence suggested occupation of a raised sand promontory or ‘island’ above the surrounding wetland (focussed in the central southern part of the area of proposed development; that part of the site which was later subject to excavation). More widespread evidence of ditched enclosures and possible field systems was also present in the areas surrounding this focus of occupation. Finds included direct evidence of occupation, including pottery sherds, daub, animal bone and burnt stone.

The Bronze Age fen edge in this area comprised a complex series of bays and promontories but at this time the Black Horse Farm site would have been located c. 900m west of the closest fen inlet, on the clay skirtland (Hall 1992, 33 and fig 18). The Bronze Age is rather poorly represented in the area surrounding the site. There is little evidence of Bronze Age activity on the major lithic-producing sites, and there are no concentrations of finds within the immediate area that might suggest occupation sites. Finds are restricted to chance discoveries but are known from both the upland and fen areas of Sawtry parish. Although many Iron Age sites are known on the Boulder Clay plateau to the south, around, for example, Abbot’s Ripton and King’s Ripton, there is relatively little known evidence of contemporary activity in the Sawtry area. A settlement at Stocking Close, near Monks Wood, 4km to the south, was excavated by Garrood during the 1930s and produced evidence for activity from the La Téne period (3rd century BC) through to the Romano-British period (Garrood 1937b). Belgic material has been reported around Grange Farm, 3km to the south-east, but no site was encountered here in the course of the Fenland Survey. Late Iron Age finds have also been found on several sites where occupation continued into the Roman period.

During the evaluation a dynamic fenland palaeoenvironment was recorded. Deeper deposits of alluvium indicative of a fen embayment were found in the lower-lying parts of the site away from the sandy rise. Evidence of the early form of the lower-lying parts of the site was suggested by a thin peaty layer of probable Neolithic or earlier date, indicative of a contemporary freshwater lagoon.

Close to the line of the Great North Road, at Tort Hill, is a small ditched enclosure that has produced Iron Age and Roman pottery (Cambridgeshire HERs 11666; MCB13711). Excavations in the 1990s (Welsh 1994,

The northern area (AS1111) was excavated by AS between November 2007 and February 2008. It lay c. 80m to the north of the previous area of excavation. This followed an 3

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire edge at Peterborough’s ‘Eastern Industry’ (e.g. Nicholson 2012; Pryor 2001). It was suggested that the evaluation had revealed regionally significant evidence of fen edge exploitation during the later prehistoric period, with occupation and possible landscape division, and further important environmental evidence of landscape change with later inundation of the former dryland, before the intensive draining and farming of the fens from the end of the medieval period onwards.

Roberts 1995) revealed further evidence of very late Iron Age occupation to the west of the A1 and Roman occupation either side of it. The Black Horse Farm site is situated to the east of Roman Ermine Street. The road appears to provide the focus for Roman settlement around Sawtry, although there is also a site c.4km east of the road around Wood Walton (Hall 1992). To the east of the A1, excavations revealed features which are thought to have marked the boundaries of plots fronting Ermine Street, as well as 2nd to 4th century pits, cobbled areas and a pottery kiln. This was a peripheral area of the Roman rural roadside settlement used for rubbish disposal and possibly used for the small scale industrial processing of metal and leather (Welsh 1994; Roberts 1995). The extent of this occupation along Ermine Street is unknown, but it may have been part of the same settlement as the Tort Hill site excavated in the 1940s (Garrood 1940, 1947).

Excavation, focussed on the areas of occupation identified during the evaluation, revealed a less variable deposit model but one which still betrayed the fen-edge location of the site. Immediately beneath the topsoil was an alluvial subsoil, possibly analogous to the stiff grey alluvium which is considered typical of the western fen margins. This is the result of rising sea-levels on a regional scale which caused increased waterlogging through ponding back of local river systems, a rising groundwater table and the deposition of alluvium. In general this is recorded overlying fen peat deposits but at this site it clearly only accumulated above the earlier ground surface. Beneath this was a ploughsoil of Romano-British date that sealed all of the earlier archaeological layers and features. Underlying this ploughsoil was the natural Oxford Clay, topped by a band of mottled grey and brown silty clay. Localised deposits were recorded stratified between the natural clay and this overlying ploughsoil. These have been variably identified as anthropogenic layers and alluvial deposits.

Excavations in Sawtry at St Andrews Church (Pearson and Murray 2000) and St Andrews Way, found sparse Roman features as well as medieval features. Other evidence for Roman activity in the Sawtry area is mainly in the form of stray finds of pottery and coins, although antiquarian sources note that a number of cremation urns were found ‘in Sautre-field, near Ermine Street, about a mile from the village’ in 1722 (HER 1339; Page et al. 1932, 268). Two Barnack Stone coffins have also been discovered, possibly representing roadside burials (HER 1332). Previous archaeological work at Black Horse Farm

At the northern end of the site (within the Davies Street (Sawtry) site, site code AS1111), the evaluation (Doyle and Weston 2006) mapped the northern extent of the former promontory of higher ground that is present to the south of this site and at which the previous archaeological work (AS 875) demonstrated significant occupation remains of the later prehistoric and Romano-British periods. The stratigraphy recorded during evaluation in this northern part of the site was fairly uniform, with only limited variation. Beneath the topsoil was a layer of made ground consisting of mid-grey clay and orange sandy gravel which was identified as being of modern date. Apart from a localised sandy subsoil, across much of the site this made ground overlay a clay-rich alluvial deposit, similar to that identified beneath the topsoil at the southern end of the site. The nature of this layer suggests that it was deposited during wet, flooded conditions. Across the northern two thirds of the area subject to trial trench evaluation at the north of the site a sand-rich alluvial deposit, most likely formed by run-off from the ‘island’ or promontory to the south, was present. This deposit thinned out to nothing towards the promontory of higher ground upon which archaeological activity was situated and was at its thickest to the north of the site. Beneath this was the natural glacial till which, at this end of the site, comprised flint gravels in a silty sand matrix with pockets of clay, forming the area of higher ground upon which Iron Age occupation was identified.

Trial trench evaluation at the southern end of the site (under site code AS857) recorded a complex fen-edge and fenland depositional sequence (Eddisford, O’Brien and Williamson 2004). Beneath the topsoil was a thick alluvial layer, which was deepest at the edge of the promontory, particularly in Trench 8. Beneath this was a deposit into which several archaeological features were cut and this overlay what were identified at the evaluation stage as ‘buried soils’ but which in reality are likely to consist of both alluvial deposits and other elements. Some of these overlay a humic deposit considered to represent vegetation at the fen-edge. This Phase of evaluation concluded that the site contained significant archaeological features of later prehistoric date, including occupation of a raised sand ‘island’ above the surrounding wetland. This occupation comprised ‘at least one, and possibly two roundhouses or former barrows seen as ring ditches. Co-axial possible field boundaries, indicative of exploitation of the fen edge in the later prehistoric period’. Finds were considered to represent ‘direct evidence of occupation, including pottery sherds, daub, animal bone and burnt stone’ (Eddisford, O’Brien and Williamson 2004). Pottery evidence from the evaluation suggested that occupation ceased in the early Roman period It was considered that the ditched enclosures and possible field systems could represent a further significant, and previously unknown, example of later prehistoric field systems as recorded over a wide area of the fen

The evidence for Iron Age occupation recorded during this phase of evaluation was sparse and comprised only evidence for boundary ditches and a ring-ditch. In light of the 4

Introduction

Figure 2. Detailed site location plan

5

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire ments’; a view which Coles and Hall (1998, 1) rail against, stating that the area contains ‘wondrously rich and varied landscapes and ancient sites’. It is natural that those with a connection to the fenlands express romantic notions about its importance and uniqueness, the inhabitants of almost every landscape do likewise. These notions are perhaps best encapsulated by Coles and Hall’s (1998, 1) retelling of an exchange between the Cambridge botanist Harry Godwin and a fenman who opined that ‘it takes a man of discernment to appreciate the fens’. The fenland is, however, like any other landscape in the United Kingdom insomuch as it is the result of a combination of naturally occurring events and human intervention and because it has offered its inhabitants, both now and in the past, a variety of problems and benefits which they have had to overcome and to harness. Undoubtedly, though, the fenland contains a variety of well-preserved archaeological sites and environmental information (Hall 1987, 1) and the riverine and marine systems and events which characterise the history of the fens have bestowed upon the area a legacy of soils which are unique to the fens, comprising a variety of silts, clays, and peats (Sly 2003, 4).

evidence previously recorded under site code AS857, the significance of this archaeology was clear, potentially representing further elements of the settlement to the south. IV. Topography, Geology, Soils and the Fenland Environment The known topography, geology and soils of the Sawtry area The village of Sawtry lies on the western edge of the fenlands, on low-lying ground beneath a scarp that rises sharply to the west. The modern land surface at the Black Horse Farm site is recorded at a little over 5m AOD, falling gently away to the east and south-east in the direction of Great Common and Sawtry Fen, parts of which lie below sea level. The land rises sharply to the west of Sawtry, which is separated from more westerly parishes by a ridge of higher land. The site lies on the fen edge between the major watercourses of the Ouse and the Nene to the south and north respectively. Sawtry Brook runs past the western side of Sawtry, c.1.7km to the west of the site. The site is bounded by fen drains; Black Horse Drain lies to the north and Middle Level Catchwater Drain to the east. It was, at the time of excavation, traversed by an open field drain (Figs. 1 & 2). These drains flow into Monk’s Lode, which empties into the Twenty Foot River, the river Nene and the Great Ouse. Although modern drainage systems and flood defences are in place, the site lies within an indicative fluvial floodplain (Anon. 2004).

The fenland basin comprises the largest area of Holocene deposits in Britain and has a complex palaeoenvironmental history (Smith et al 2012). At the commencement of the Postglacial interval, the Fen basin was dry land drained by a series of rivers flowing into a major outlet that ran out through the Wash. Early Mesolithic hunter-gatherers camped adjacent to ponds and lakes on the river terraces and probably used the river valleys as their main access routes. Rises in sea level have caused periodic incursion of water into the surrounding flat fen plains, resulting in the deposition of silts and estuarine clays. The environment created as a result was one of extensive salt-marshes with fresh water fens and large expanses of peat generated under waterlogged, anaerobic conditions (Seale 1975; Hall 1996).

The site lies in an area of Oxford Clay (SSEW 1983) which is overlain by deep and variable layers of alluvium and first river terrace gravels, which rise from the Great Ouse to the east and south of the site. On the high ground beyond the fen-edge scarp is a thick clayey till with occasional patches of glacial gravel. The scarp is deeply indented, forming a series of bays and promontories. Deep peat, containing large numbers of ‘bog oaks’, covers most of the fen area. Near the fen edge post-Roman alluvium spreads out from brooks issuing from the higher ground and covers earlier deposits. The majority of Sawtry Fen consists of this material at its surface (Hall 1992, 33).

The deposition of Holocene deposits began with the marine transgression that flooded ‘Doggerland’ at around 8000– 6000BP (Ward et al. 2006; Weninger et al. 2008). During the Mesolithic, with sea-level rise, water tables rose c. 16m and consequently the Fenland Basin became flooded as the drainage systems ceased to be effective; early sites in the fen basin are likely to be deeply buried (Reynolds 2000). This allowed for the development of freshwater reed beds in some areas (Zalasiewicz 1986; Smith et al. 2010; Smith et al. 2012). Natural and Anthropogenic factors induced a changing environment in Neolithic Cambridgeshire. In the north of the county this is characterised by an encroachment of fen environments, creating varied landscapes of reedswamp, fen carr and patchy woodland. Elsewhere the mixed deciduous woodland landscape was broken by man-made clearance, intended to create areas of pasture and cultivation plots (Pollard 2000). By the later Neolithic, the Fenland was transformed by both fresh and marine waters, peat was forming in many riverine backwaters and in deeper channels, just as marine and brackish-based silts had encroached on much of the lowlands around the Wash. By this time vast amounts

The character of the fenland The fenland basin of eastern England, in which the site lies, covers an area of approximately 4000km² (Wheeler and Waller 1995, 223). It extends some 120km from Lincoln in the north to Cambridge in the south. It is low lying, with little of the land surface exceeding 3.5m OD in height; through it flow the rivers draining the greater part of central England, an area more than five times the size of the fens themselves (Sly 2003). To some, the fenland is ‘one of the most distinctive landscapes in Britain’ (Honnor and Lane 2002). To others it is ‘an unendingly flat landscape of dark fields, sluggish rivers, dykes and banks, with infrequent scattered settle6

Introduction on the floor of the Fenland was overwhelmed by extensive ‘marine’ beds of clays and silts before renewed peat growth took place above the clays and silts.

of forested dry land had already begun to be lost to the swamping peats (Coles and Hall 1998, 15). A major phase of inundation in the late Neolithic and early Bronze Age (occurring c. 2500-1800BC) led to the loss of dry land around March and Chatteris and the isolation of the two main fen islands (Last 2000). Marine-based deposition of silts dominated the whole of the Fenland area, overriding the peaty expanses and clogging the valleys. In the embayments, where rivers flowed, and along the margins of the fen, peat remained exposed and continued to form due to the severely impeded drainage. Fenland water levels peaked in the Iron Age, at approximately the time that the Black Horse Farm site was occupied. The main rivers of the region followed the courses that they had previously but a band of seaward marine deposits caused an increased expanse of peat-fen to accumulate. Extensive flooding in the central regions led to the deposition of coarse material that would become the silt fens and salt marshes, and mud flats encroached over peat and over areas that had previously been dry land. The southern fen became an immense expanse of freshwater wetland in which peat formed up to 2.5m above mean sea level (Coles and Hall 1998, 41). During the Roman period, it is considered that watertables were generally lower than in the Iron Age, at only 1.5m above mean sea level (Coles and Hall 1998, 49). The peat level fell somewhat and the fen edge is considered to have occurred at 2m OD (Hall 1992, 8). However, it has been noted in some parts of the fens that there is a mid 3rd century gap in occupation, attributed to changing climate conditions and a rise in the watertable (Upex 2008, 178). For several centuries in the immediate post-Roman period there was little change in fenland watertables and a period of some stability occurred. However, by 700AD conditions began to deteriorate and there followed periods of inundation by the sea and by inland waters, forming the extensive fens that were described by monastic writers of the 12th century (Coles and Hall 1998, 61). Post-Roman alluvial deposits are considered to cover much of the surface of Sawtry Fen (Hall 1992, 33). By the time that these monks were describing the fens, however, there had ceased to be any active marine phases as seabanks provided protection from inundation from the Wash. This allowed the siltlands in the north of the fens to be developed, leaving the peat fens as wetland (ibid. 68).

The depositional sequence recorded during archaeological excavation at Black Horse Farm Hall (1992, figs. 18 & 20) indicates that although the Black Horse Farm site lay at some distance from the fen edge in the Bronze Age, by the Roman period it was, although lying close to the fen edge, almost subsumed within the fen. This suggests that during the period represented by the archaeology recorded here, changing environmental conditions and the encroachment of the fen resulted in a notable effect on life in this area. These changes may be evident in the deposits recorded during the archaeological work carried out at Black Horse Farm. Recorded overlying the natural Oxford Clay were deposits of brickearth. During the Iron Age, the brickearth and gravel areas formed a land surface with wet fen expanding from the east as a result of rising sea-levels and the local ground water table. The Iron Age archaeology was therefore cut into what was, at the time, a dry terrestrial landscape (Scaife 2006). Information recorded during the preceding trial trench evaluations indicates the presence of a humic layer, judged to be a remnant of prehistoric fenland reed beds lying at a height of 2.65m AOD. However, Ditch F1121, a feature identifiable as being of Iron Age date, was observed to cut a palaeochannel cut into the natural substrate at a height of 2.80m AOD. This indicates that the fen edge must have lain between 2.65m and slightly below 2.80m AOD during the Iron Age occupation of the site and is broadly consistent with Hall’s (1992, 8) suggestion that the Iron Age fen edge occurred at approximately 2.5m (but contradicts information given by the same author (fig. 20) that the site would have been within the fen wetland during the Roman period). Using this information, in conjunction with the known topography of the area, has led to the development of a postulated model of the maximum extent of the Iron Age fen edge (Fig. 3) in the immediate vicinity of Black Horse Farm. This reveals the approximate form of the area of higher land upon which occupation occurred. It also suggests that a second area of higher land existed to the south upon which further archaeological features exist, cut in to alluvial deposits recorded in Trenches 8 and 9 of the 2004 evaluation (Eddisford, O’Brien and Williamson 2004).

Brew et al. (2000, 270) offer a simplified version of the evolution of the fenland, suggesting that it was dominated by three main events. The first of these events was the initial post-glacial transgression over the pre-Holocene land surface. This was followed by a second phase comprising the sedimentary infilling of the embayment with rising sea-level; this was characterised by the deposition of clastic sediments alternating with peat accumulation, with the final stages dominated by peat formation. The third event comprised renewed expansion of tidal flat areas forming the final clastic fill, which mostly occur within the northern parts of the fens at the margins of the Wash. This is similar to the general, classic vertical stratigraphy of the fenland demonstrated by Waller (1994, 14), and which is broadly accurate for the southern fens, although localised complexities and variations do occur: Initial peat growth

The deposit model recorded during the excavation phases of archaeological work at Black Horse Farm was less complex than that recorded during the preceding evaluation phases. During the excavation of the southern area of excavation (AS857), the archaeology was recorded cutting the natural Oxford Clay, which was capped by a band of mottled grey and brown silty clay. Overlying this were a variety of localised deposits that appear to have been of alluvial origin (see below). The earliest of these was L2459 which covered the majority of the extent of the large Roundhouse 3 and appears to have been associated with its abandonment. In the south-eastern corner 7

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 3. Iron Age land surface

of this excavation area deposit L2206 was identified covering a small area between Iron Age ditches; this coarse sandy silt was considered to represent some kind of flooding or alluvial event. These both appeared to be earlier in date than L2420, a silty clay layer which overlay L2459

and contained a variety of pottery. Overlying L2420 were L2060, a mid to light grey brown highly silty clay located within the enclosure formed by Ditches F2226 and F2152, and L2211, a very similar deposit identified in the southwestern part of this excavation area. These both appear to 8

Introduction Table 1. Simplified deposit model for excavation area AS857

Layer

Description

Depth

L2000

Topsoil

0.32m

L2001

Alluvial subsoil

0.90m

L2002

Buried soil layer. Early Roman ploughsoil

0.26m

L2060

Localised alluvial deposit within late Iron Age enclosure

0.28m

L2420

Localised deposit associated with abandonment of roundhouse

c. 0.10m

L2459

Localised deposit associated with abandonment of roundhouse

c. 0.10m

L2494

Localised deposit. Thick floor or mound associate with roundhouse

0.22m

L2003

Natural Oxford Clay capped by a band of mottled grey and brown silty clay

-

Table 2. Simplified deposit model for excavation area AS1111

Layer

Description

Depth

L5000

Topsoil

0.25-0.50m

L5011

Modern made ground

0.45m

L5012

Alluvial layer

0.26m

L5010

Natural sandy gravel

0.28m

have been of later 1st century date and probably represent the same episode of alluvial activity. These deposits were overlain by L2002 which appeared to be a topsoil that had developed during or following the final phase of human activity. All archaeological layers and features were sealed by L2002. L2002 was subject to a test pit survey, designed specifically to date the deposit. Test pits were hand dug on a five-metre grid in the southern part of the excavation area and on a ten-metre grid in the northern part, where the deposit was less well developed. The test pits produced a large quantity of dateable material, the vast majority of which indicated that the deposit had developed towards the end of the mid 1st century BC to mid 1st or 2nd century AD. Deposit L2002 was overlain by L2001, a plastic, mid orange-brown silted clay which measured up to 0.90m in depth. L2001 appeared to be an alluvial deposit consistent with the layers of post-Roman alluvium considered to cover much of the surface of Sawtry Fen (Hall 1992, 33). The topsoil (L2000), a friable, mid to dark brown clayed silt which was up to 0.32m in depth, was present overlying L2001.

Clay, revealed at a depth of c.2m below existing ground level. During the excavation phase of work at this northern area, it was observed that that this bluish grey clay was present throughout the site in small pockets within the natural substrate. Generally, L5010 sloped downwards in a northerly direction. A level of 4.03m AOD was taken at the top of F5074 Segment B, in the south-west corner of the excavated area. It was observed at 3.69m AOD in the vicinity of F5111 in the north-west corner. Looking beyond the boundaries of the excavated area into the larger evaluation area, L5010 was identified at 2.16m AOD at the northern end of Trench 6 and at 1.96m AOD at the western end of Trench 5. These levels underscore the position of the site on the northern slopes of a gently rising fen-edge promontory. All of the archaeological features within this excavation area cut L5010 and were sealed by L5012, which was a clay rich alluvial deposit. It appears that this deposit was consistent with L2001 recorded in the southern excavation area and is, therefore, representative of post-Roman alluvial deposition. Overlying L5012 was L5011, a mid grey clay and orange sandy gravel that contained occasional red, frogged bricks and plastic items. This, in turn, was overlain by a topsoil consistent with those of the Hanslope Association (SSEW 1983).

The natural substrate recorded during the excavation of the northern area (AS1111) was glacial till, seemingly consistent with that which Hall (1992, 33) states is found on the high ground beyond the fen-edge scarp. This deposit (L5010) comprised a moderately fine flint gravel in an orange silty sand matrix. During the evaluation that preceded this phase of excavation (Doyle and Weston 2006), the natural substrate observed in Trench 1 was Oxford 9

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Plate 2. View north across the interior of the Iron Age enclosure during excavation of Roundhouse 3

10

2 The Results of the Excavation I. Phasing

character of activity at the site as a direct result of local environmental factors.

Archaeological features and deposits representing activity of middle/later Iron Age to Romano-British date were recorded in the southern excavation area (AS857; Fig. 4). Middle/later Iron Age features were recorded in the northern excavation area (AS1111) but identifiable Romano-British activity was largely absent (Fig. 5).

II. Phase 1: The Middle Iron Age Introduction Phase 1 (Fig. 6) represents activity in the middle to late Iron Age. The pottery evidence that characterises this period can be assigned a date between the 5th and 2nd centuries BC. However, given the character of the activity assigned to Phase 1 and its apparent relationships with activity that can be assigned a later Iron Age date, it is most likely that the activity represented by Phase 1 occurred in the latter part of the date range indicated by the ceramic evidence.

Analysis of the artefactual assemblages, stratigraphic sequence, spatial and functional relationships between features and the use of scientific methods of dating have identified six distinct phases of archaeological activity at the site (Table 3). These are characterised by distinct changes in the ceramic tradition current at the site during a particular time period or by distinct changes in the Table 3. Summary of phasing

Date

Phase

Details

Middle Iron Age (c. 5th to 2nd centuries BC)

1

Roundhouse and related features on southern slope of promontory. A second roundhouse, encircled by wide, deep ring ditch, associated ovens and possible working surfaces. Smaller ‘roundhouse’ structure to immediate south-west with possible ancillary function. Evidence for possible contemporary enclosure surrounding structure. A roundhouse with two smaller ancillary round structures on the northern slope of the promontory with a possible associated boundary ditch. Ditches and gullies. Occasional small pits towards southern edge of site. Possible inundation layer.

Later Iron Age (c. 1st century BC to 1st century AD)

2

Large enclosure ditches, regularly maintained and cleared out, remaining open in to Phase 3 and possibly later. Human burial radiocarbon dated to the 1st century BC to 2nd century AD. Pit oven cut into, and deposit, of possible alluvial origin, overlying the final abandonment layer of the large roundhouse, initially established in Phase 1. A small number of discrete pits and short gullies to the east and west of the enclosure on the southern slope of the promontory and in the vicinity of the roundhouse on the northern slope and a more substantial ditch entering the excavated area from the southwest.

Romano-British (c. Mid 1st century AD)

3

Beam slot structure cutting deposits within enclosed area and strip field systems on higher land at the centre of the promontory. Broadly parallel ditches, possibly forming a delineated trackway, approaching the area of the enclosure, from the north-west.

4

Alluvial deposits sealing earlier archaeological features on the southern slope of the promontory

5

Pits cutting, and surfaces overlying, Phase 4 inundation deposits Three separate ditches cutting features on the southern slope of the promontory and containing ceramic evidence to date them to the mid 1st century AD

6

Buried soil L2002 (test pitted layer)

Mid 1st to 2nd century AD

11

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 4. Southern excavation area (AS857) all features plan

Roundhouse 1

Phase 1 represents the first evidence of settlement at this location. It is possible, however, that some of the linear features recorded to the west, during the evaluation that preceded excavation, and which are considered to represent fields or enclosures close to the fen edge, may slightly predate this occupation activity. Stratigraphic evidence indicates that, in the southern excavation area, Phase 1 saw the establishment of an initial roundhouse and the subsequent development of a possible enclosure and the apparent replacement of this first roundhouse. In the northern part of the site, a group of three round structures, comprising a roundhouse and two probable ancillary structures, was established during Phase 1.

Clay Floor L2270 covered much of the area enclosed by the ring ditch (see Fig. 7). This represented a floor surface or occupation layer. L2270 sealed several features which were suggestive of domestic activity within the roundhouse. Domestic activity appears to have continued following the deposition of L2270. A number of features cut the floor surface, including possible ovens or hearths (F2383 and F2431) and a variety of postholes and/or small pits. Due to later truncation by Ditch F2325, the purpose of many of these features remains uncertain; no convincing structural configuration of postholes was apparent. 12

The Results of the Excavation

Figure 5. Northern excavation area (AS1111) all features plan

Other features, with no stratigraphic relationship to L2270, existed within the area encircled by the ring ditch of Roundhouse 1. Some of these represent domestic activity, such as possible Hearths F2462 and F2442, while others would appear to have had a structural function. A line of postholes in the south-western part of the structure (F2231, F2233, F2235 and F2237; Grid Square G5) may have represented some kind of elaboration to the entrance of the structure or a form of internal division.

Several pits were identified within Roundhouse 1 (F2174, F2188, F2444 and F2470) which, despite the possible wet ground conditions in this fenland environment, displayed similarities to the storage pits recorded in Roundhouse CS20 at Danebury (Cunliffe 1986, 98-99). Storage pits have been recorded in other environments with high and potentially fluctuating water levels, such as at Ingham Quarry, Suffolk (Newton and Mustchin 2012). This suggests that such conditions were not considered problematic and it is possible that such pits were lined with wicker-work or other materials to make them water-tight (Harding 1974, 78).

Many of the remaining features comprised post or stakeholes and are likely to have had a structural function. 13

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 6. Phase 1 plan

14

The Results of the Excavation earlier phase of the ring ditch. This may indicate reworking to form a new entranceway at this point. It is difficult to state with any certainty whether the ring gully defining Roundhouse 1 was structural, representing the location of the building’s walls, or if it acted as an enclosing drainage gully (Pope pers. comm.). This is in contrast with the other roundhouses recorded at this site, especially Roundhouse 3, the defining ring ditches of which appeared to be enclosing, rather than structural, features perhaps with a drainage or water-management function.

Evidence for the ‘division of diurnal activities within the Roundhouse’ which ‘mimics the movements of the sun, with the tasks of the daytime carried out in the south and the activities of the night-time in the north’ (Parker Pearson 1999, 49; Fitzpatrick 1994) was not evident in Roundhouse 1. The pattern of finds distribution within Roundhouse 1 may, however, be biased by the obscuration of the southern and south-eastern parts of the structure by the large Phase 3 Ditch F2325. Roundhouse 1 (Fig. 7) was defined by a composite ring gully, formed of F2027, F2110, F2044 and F2190. The ring gully that these features formed varied little from a depth of 0.2m and a width of 0.6m. The area that they enclosed had a diameter of c. 12m.

Possible structure S3012 A group of postholes (F2010, F2012, F2014, F2018, F2020 and F2024; Grid Square F5; Figs. 6 & 8), to the west of Roundhouse 1, were arranged in a configuration that suggests that they are likely to have formed some kind of simple structure (S3012), possibly associated with the features that lay close by. At many Iron Age sites, arrangements of four or five postholes are interpreted as granary structures (Reynolds 1979, 80). Alternatively, similar, but slightly less substantial structures may be considered to be ricks, in which corn or barley, that had been cut damp, could be stored and dried prior to threshing (Cunliffe 1986) or drying racks for grain or skins (Megaw and Simpson 1981, 382). In the fenland, a structure such as this may

F2110, which formed the north-western section of the ring ditch, terminated close to the end of Boundary Ditch F2325, which truncated a large proportion of the southern and south-eastern part of the structure (Fig. 7). The position of the terminus of F2110 may indicate a gap in the ring gully, suggestive of an entranceway, in the south-east of its circumference. At the south-west of the circumference of the ring ditch, the termini of F2027 and F2044 opposed each other across a 0.65m gap. Both of these termini truncated F2190, which appeared to be a short section of an

Plate 3. Roundhouse 1 during excavation. In this view to the eastnorth-east, Gullies F2027 and F2174 are visible in the foreground, while in the top right, Phase 4 Gully F2309, which cut clay floor L2270, is visible.

15

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 7. Roundhouse 1 and associated features

16

The Results of the Excavation Although its west-south-westerly continuation was obscured by Phase 2 features, the position of F2385 suggests that it cut features forming Roundhouse 1. This implies that Roundhouse 1 was no longer in use when the enclosure around Roundhouse 3 was created. However, it is not possible to confirm whether Roundhouse 3 was extant when the enclosure ditches were cut or if it was constructed after the enclosure was established. Ditches F2200 (Grid Squares I6, J6) and F2836 (Grid Squares J5, K5; see Figs. 2, 4 & 6) were located on the very eastern edge of the excavated area and appeared to run roughly parallel to one another. The western terminal ends of both of these features were truncated by Phase 2 Boundary Ditch F2152. The linear form of these features, coupled with their width (comparable to Ditch F2385), has led to the tentative conclusion that they formed part of the same enclosure system as F2385. F2274 and F2287 (Grid Square H7, I7, H8, I8) a pair of roughly parallel ditches, Ditch F3005 (Grid Squares F9, G9), the only feature of Phase 1 date located in the northern half of the southern excavation area, Ditch Terminus F2176 (Grid Square E8), which was associated with a series of postholes that may represent an associated palisade or fenceline, and Ditches F2075 (Grid Squares G7, H7) and F2080 (Grid Square G6) may also have formed part of a system of enclosure surrounding or associated with the Phase 1 settlement. Truncation by later features has obscured some of the more meaningful spatial relationships making it unclear quite how these features functioned in association with one another.

Figure 8. Structure 3012

Roundhouse 2 A group of features, including gullies (F2418, F2455 and possibly F2457), numerous post and stakeholes (F2522, F2524, F2554 F2555 F2757, F2591, F2692, F2694, F2696, F2698, F2700 and F2759), pits (F2526, F2536, F2539, F2569, F2571, F2601, F2712, F2713 and F2731), and several surfaces (L2502, L2514, L2563, L2564 and L2689), were located in Grid Squares H3, I3, H4. Together, these represent a second circular structure (Fig. 9).

conceivably have been used for storing and drying reeds for thatching or basket weaving. Pit F2008, which lay immediately to the north-west of structure S3012, and Pit F2006, which lay a short distance to the south, were both tentatively interpreted as cooking pits due to evidence for burning in their basal fills. It is possible that Structure S2012 represented a small shelter or preparation area associated with these ovens.

On the whole, finds were present in low quantities in this structure. Only two features, F2571 and F2731, were found to contain dateable material. This material consisted of 4 sherds (52g) of pottery, consistent with a Phase 1 date. However, the position of this structure within the Phase 2 enclosure and its proximity to Roundhouse 3 suggests a possible ancillary function, possibly indicating that this structure may have had a similar lifespan to the larger structure to the north which appears to have remained in use into later phases.

The Phase 1 enclosure and possibly associated features Roundhouse 1 appears to have fallen into disuse prior to the time that Ditch F2385 (Grid Square H6; Figs. 2, 4 & 6) was cut. This feature appears to represent part of an enclosure associated with Roundhouse 3. It is possible that unphased Ditch F2176 (Grid Square H7) also formed part of this enclosure. Postholes F2389 and F2450 were cut into the base of F2385, suggesting the presence of a fenceline, serving to emphasise the boundary. Gully F2387 (Grid Square H6; Figs. 2, 4 & 6) ran parallel to Phase 1 Gullies F2385 and F2386, suggesting some kind of relationship although this was not identified stratigraphically. During excavation, it appeared that F2387 may have been cut by F2385 but this was not conclusively proven.

The fragmentary nature of the features forming Roundhouse 2 does not detract from its interpretation as a roundhouse (or similar structure). The presence of roundhouses has been interpreted on the basis of much more ephemeral evidence in the cases of, for example, 17

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Plate 4. View north across excavation area AS857 showing, in the near centre of the photograph, investigation of the features and deposits forming Roundhouse 2.

Structures 24-26 and 30-32 at Site IV, Colne Fen, Earith (Evans 2013, fig. 5.11), Structures II and III at Wardy Hill, Ely (Evans 2003, figs. 30 & 31, 44), and Buildings 5, 8, 9, 10 and 11 at Cedars Park, Stowmarket, Suffolk (Nicholson and Woolhouse 2016, figs. 14, 23 & 24).

3.5 and 4.5m from its two termini; these were F2708 and F2831. Finds from Ditch F2324 comprised pottery (6929g), CBM (165g), animal bone (4965g), shell (5g), flint (36g), burnt stone (16g), clay/fired clay (13g) and slag (621g). Roundhouses with similar enclosure ditches to that surrounding Roundhouse 3 have been identified elsewhere in Cambridgeshire (e.g. Evans and Hodder 2006, fig. 5.46; Evans 2003, figs. 29, 30, 31) and in neighbouring Northamptonshire (e.g. Woodward and Hughes 2007; Jackson 1983). The function of the ring gully in roundhouse construction is uncertain in many cases and construction appears to have varied over time, between regions, between sites, and even within sites (cf. Allen et al. 1984; Cunliffe 1991, 242-246). They are usually interpreted either as wall trenches or as drainage features. Silty fills and u-shaped profiles, as displayed by F2324, were characteristic of the ring gullies identified at Crick Covert Farm, Northamptonshire. These were interpreted as drainage features (Woodward and Hughes 2007, 190). There appears to be a correlation between ring gullies of this type and impermeable clay soils (ibid.). It has been suggested that such features are storm-water gullies designed to hold water during rains and also to keep the ground surface dry.

Roundhouse 3 and associated features The encircling ring-gully Features comprising Structure S2441 (Roundhouse 3; Grid Squares H4, I4, H5, I5; Figs. 10-14) lay within an area enclosed by ring-gully F2324. This feature had a diameter of c. 13.8m, making the structure, as a whole, slightly larger than Roundhouses 1 and 2. F2324 was c. 1.3m wide and 0.35m deep with irregular sides and a concave base (Plate 5). Its fill (L2315) was a dark browngrey firm silty clay. A single entranceway was apparent, a gap of c. 3.1m between the ring ditch’s termini in the south-eastern part of its circumference. The part of the ring ditch immediately south of the entranceway seems to have been a double ditch, with the number F2679 applied to the inner of the two ditch bases. North of the entranceway, the terminus of the ring ditch had been remodelled at some point, the number F2773 being applied to the recut. Two postholes were cut into the base of the ring ditch between

When it was first created, Ring Ditch F2324 would have been cut into the underlying sandy gravels and brickearth 18

The Results of the Excavation deposits forming the promontory on which all of the Iron Age occupation at this site was located. The fen edge would have been comparatively distant and the area would have been a dry terrestrial surface (Scaife 2006). Later, however, as the level of the fen rose, this ditch would have been inundated regularly (as the later alluvial deposits indicated) and, as has been demonstrated, even contained

standing water later in its lifespan (see Fryer and Pelling, Ch. 3.VII). The effects of a rising water table are likely to have left silty deposits within the ring ditch and its profile suggests that it has been recut or cleared out. This re-cutting or, as is more likely, regular scouring out of the ditch is crucial to understanding the lifespan of Roundhouse 3; spatial and, in one case, stratigraphic relationships with

Figure 9. Roundhouse 2

19

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Plate 5. A view north-east across Roundhouse 3 during excavation

later features suggest that Ring Ditch F2324, and therefore probably the rest of the building as well, was maintained in this way well into Phase 2.

cal reason for this large gap between it and the structural components of the building. The ditches surrounding the roundhouses at Haddenham, Wardy Hill and Brigstock may have had secondary functions denoting boundaries around their internal structures; it is possible that Ditch F2324 was created as a particularly emphatic expression of this function.

Ditches similar to F2324 have been recorded encircling, or partly encircling, Roundhouses 3, 6 and 7 at the Haddenham V site in Cambridgeshire (Evans and Hodder 2006, fig. 5.46). Six circular structures have been identified at the Wardy Hill complex at Coveney, Ely, from the presence of similar circular or sub-circular ditches (Evans 2003, figs. 29, 30, 31). At Wardy Hill, no structural evidence for the presence of the buildings was recorded. However, at Haddenham V, the outer wall of the roundhouse buildings was around 1.5m from the centre of the ‘eaves-gully’ (Evans and Hodder 2006, fig. 5.46; Evans and Serjeantson 1988, fig. 3). At the Iron Age site at Brigstock, Northamptonshire, the single roundhouse was encircled by a penannular drainage gully, the centre of which lay some 1.4m from the outer wall of the house, around most of its circumference (Jackson 1983, 17). At over 3m, the distance between the outer wall of Roundhouse 3 and the centre of F2324 is much greater than these other examples. If it is understood that the ground surface upon which this structure was built was dry at the time of construction, and the function of the ring ditch was only to deal with rain water and surface run-off, rather than to protect against the encroaching flood-waters, there seems little practi-

F2773, a pit-like feature, was a recut of the northernmost terminus of the circular Enclosure Ditch F2324, on its inner margin. The part of the ring ditch immediately south of the entranceway seems to have been a double ditch, with the number F2679 applied to the inner of the two ditch bases (Weston and Nicholson 2006, 19). Alternatively F2679 may be understood as a recut of the inner part of F2324 at this point. Both of these features were found to contain near-complete pottery vessels. These vessels were both East Midlands Scored Ware jars, although they were of differing forms and displayed different types of decoration (see Peachey, Ch. 3.I). The discovery of a pair of similar vessels from two features on either side of the inner part of the entranceway of Roundhouse 3, suggests that they were deposited as part of some kind of symbolic act. The presence of middle Iron Age vessels surviving, intact, in a feature that is suggested, by other evidence, to have been maintained through re-cutting or scouring-out, into the later Iron Age appears contradictory. It seems that these 20

The Results of the Excavation

Sections of features with Roundhouse 3

Figure 10. Roundhouse 3 after complete removal of L2494

21

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Sections of features with Roundhouse 3

Figure 11. Roundhouse 3 after partial removal of L2494

22

The Results of the Excavation pottery vessels must have been deliberately curated and deposited within the ring ditch when the roundhouse itself was abandoned later in the period represented by Phase 2. Symbolic deposition of objects into watery contexts is a well-recognised Iron Age behaviour; metalwork and other objects recovered from bogs and rivers are likely to have been offerings deliberately dedicated to the spirits of these places (Cunliffe 2005, 566). Deliberate deposition into man-made watery contexts has also been noted. At Crick Covert Farm in Northamptonshire, deliberately water-filled roundhouse ring-ditches have been identified as having served as a watery context for the deliberate deposition of artefacts (Woodward and Hughes 2007, 191). It would appear that these artefacts, which would have been antique at the time of deposition, were placed into F2324 as closure deposits when the roundhouse was abandoned.

Internal features Amongst the stratigraphically earliest features (Fig. 10) there is little clear structural configuration. The majority of these features were clustered in the area adjacent to the entrance within the surrounding ring-gully F2324. A number of postholes (F2748, F2969, F2907, F2909, F2911, F2913, F2996 and F2998), a postpad (F2963) and a pit (F2917), were located close to the centre of the roundhouse. With the exception of F2917, which produced 80g of animal bone, none of these features contained any finds. These features may have had a structural function in the earliest part of the lifespan of Roundhouse 3, or may represent posts inserted to provide support during construction. In the north-western quadrant of the roundhouse, Gully F2506 was observed cut into L2003. Gully F2506 appears to have been a structural component of Roundhouse 3. Its form and position suggest that it probably held a line of wattle and daub walling. As well as cutting L2003, it was

Figure 12. Roundhouse 3 after removal of L2459

23

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire it continued into this sector of the structure. These pits and postholes may have held posts that supported a wooden wall in this area, although only F2646, F2673, F2785 and perhaps F2861 displayed profiles that suggested that they were capable of supporting a post. Shallow postholes F2675, F2648, F2655, F2657 and F2823 formed a line running south-east to north-west and may represent some kind of internal division or possibly some kind of small structure associated with Hearth F2663. Finds from features cut within L2494 were limited, with 9g, 28g and 10g of animal bone coming from F2673, F2781 and F2850 respectively, while flint (11g) and burnt clay (17g) were recovered from F2779 and F2785. Four of the pits (F2787, F2755, F2721 and F2714) contained assorted finds while F2729 and F2671 were devoid of artefactual evidence. None of these features contained more than seven sherds of pottery and no more than 66g of animal bone was recovered from any feature other than Posthole F2823.

observed to cut through from the top of deposit L2494 and at higher stratigraphic levels (Figs. 11 & 12) it appeared to arc round further to the west and south-west. It is therefore not clear if F2506 was a later development in the lifespan of the building, deliberately cut through L2494 into the underlying natural, or if it was an early structural component which remained in place as L2494 accumulated, or was deliberately built up, around it. A similar lack of clear structural configuration was evident amongst the features recorded cut within L2494 (Fig. 11). Many of these features were pits which are unlikely to have had structural functions, although it is interesting to note that a group of postholes (F2646, F2781, F2779, F2861, F2850, F2785 and F2673) and pits (F2787, F2755, F2729, F2721, F2714 and F2671) extended around the western/ north-western part of the structure’s circuit, loosely following the line that gully F2506 would have followed had

Figure 13. Roundhouse 3 after removal of L2420

24

The Results of the Excavation remainder of the circumference was formed by wattle and daub walling without this stone footing or reinforcement. The stratigraphic relationship between cobbles F2477 and L2494 indicates that F2477 represents a later phase of structural development in the building’s lifecourse. Many Roman buildings that appear to be made of stone only had stone footings to inhibit the rotting of the timber superstructure (de la Bédoyère 2001, 22) and this may denote a similar approach. F2477 may represent an attempt to raise the wooden superstructure of the roundhouse off the damp floor; possibly a response to the increasingly damp conditions. Stone elements within a roundhouse would appear to be unusual for the fenland area, where, as with most other areas of south-eastern England, Iron Age buildings were mostly of wooden wattle and daub construction (Cunliffe 2005, 242). It remains unclear, however, why stone was only used in the north-western part of the building.

Of the stratigraphically latest internal features of Roundhouse 3, i.e. those overlying or cut into the top of layer L2494, amongst the most significant was F2477, a collection of river cobbles and flints. Although the possibility of stone robbing makes interpretation difficult, several explanations have been offered for this spread of stone; it may represent the remains of a drystone wall that originally ran around the entire structure, it was possibly all that remained of a more extensive stone floor, or may have been part of a stone sill built around the inside of the building and used to keep beds or produce off the damp floor (Jackson 1983, 14). The tight, compact and clearly localised appearance of F2477 following the removal of Layer L2459 (Fig. 12), suggests that it was the remnant of a stone wall, which lay on top of L2494 and was present solely in the north-western quadrant of the structure. Further river cobbles and flints spread throughout Layer L2459 may indicate that F2477 originally stood to a greater height; its constituent material appears to have been spread over a wider area following the abandonment of the structure (Fig. 13). The presence of postholes beneath F2477 suggests that, at least initially, the stones were piled up against and around the structural timbers of ST2441 to form the wall. F2477 indicates that Roundhouse 3 had a stone-built element and the recovery of fragments of daub from within it may support the suggestion that it had a structural function (see Peachey, Ch. 3.II). The recovery of daub from other features comprising Roundhouse 3 and the presence of structural postholes and Gully F2506 suggest that the

Overlying L2494 was L2459, a dark brown-grey, compact silty clay with charcoal and many sub-rounded river cobbles and flint nodules. It was present across the majority of the interior of the Roundhouse 3, absent in only a narrow band in the eastern/north-eastern part of the structure (Fig. 13). Postholes F2558 and F2542 probably represent structural supports for a wall in the southern part of the building’s circumference. Both of these features contained East Midlands Scored ware pottery and animal bone. Other

Plate 6. Wall Footing F2477 following removal of overlying deposits

25

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Sections of features forming Roundhouse 3

Sections of features associated with Roundhouse 3

Figure 14. Features forming and associated with Roundhouse 3

26

The Results of the Excavation greater concentration towards the termini of the feature and notably higher weights of animal bone recovered from Segments Q and V. The distribution of CBM/daub did not conform to this pattern, occurring in only the northern half of the feature’s circumference (Fig. 15).

postholes (F2549, F2547, F2551 and F2605) lay to the north of these, appearing to be too far inside the wall’s circumference to form part of it, though a structural function remains possible. Gullies F2515 and F2439, Postpad F2512 and Posthole F2534 formed a line that led from within the entrance of the ring ditch, turning through 90 degrees, to Pit F2433. This may represent a fence line augmenting the entranceway to the Roundhouse.

Distribution of pottery within internal features of Roundhouse 3 (Figs. 16-18) showed a slight bias towards greater concentrations in the southern half of the building with large concentrations (over 2000g) recovered from F2773, a recut of the northern terminus of F2324, and F2679, a feature which may represent a re-cut of F2324 or form a double ditch with it, causing a noticeable bias towards the south-eastern quadrant. The distribution of animal bone (Figs. 19-21) followed a similar pattern with the high concentrations from Pits F2517 and F2497, the possible structured deposits, causing a noticeably greater weight of material within the south-eastern quadrant. Figures 16-18 and 19-21 show a fairly uniform spread of both pottery and animal bone throughout the internal features of the structure, with an increase in the south-eastern quadrant. However, when the total weights of these classes of finds are plotted by quadrant on a graph, there is clearly a far greater quantity of material in the south-eastern quadrant, where the entrance to the structure is located, than anywhere else (Graphs 1 and 2). There was also a high concentration of pottery in the south-western quadrant of the structure; this is notable as this contained proportionately fewer features than the other quadrants.

Distribution of artefactual remains within Roundhouse 3 Like Roundhouses 1 and 4, the entrance of Roundhouse 3 was located in the south-eastern quadrant of its circumference (Grid Square I4). It is often noted that the orientation of roundhouse doorways is broadly to the east or southeast in examples from the British Iron Age (Oswald 1997). Parker Pearson (1996, 119 and 127) suggests that houses were orientated in this direction as the east or south-east was a propitious or sacred direction, related to sunrise and the daily rebirth of the cycle of light and darkness revolving around the house. However, other theories suggest that more practical concerns dictated the orientation of roundhouse doorways with Hingley and Miles (1985, 63) suggesting that entrances were placed as best to avoid the prevailing winds. Parker Pearson (1999) suggests that the division of diurnal activities within the roundhouse structure will lead to greater distribution of artefacts within the southern or south-eastern parts of the structure due to the apparently standard practice of placing the entrance at the east or south-east of the structure. Pope (2007, 222), argues that any suggestion that such a pattern of artefact distribution is due to cosmological beliefs regarding the symbolic significance of south-eastern orientation is methodologically unsound and overlooks context, disregards taphonomy, agency and regional variation and does not consider the environment sufficiently. Webley (2007) has presented evidence to indicate that such patterning of finds does not relate to everyday use of roundhouse buildings but does not discount the role of symbolic activity in the formation of such patterns, suggesting that it is in fact the result of a particular form of ‘ritualised’ house abandonment behaviour. Pope (2007, 216, fig. 8) indicates that the archaeological evidence for north and central Britain, as opposed to the Wessex area upon which Parker Pearson’s (1999) model is based, reveals a general trend towards the use of front and backspace in the roundhouse which is related to the way in which light falls within a circular structure.

The biases towards the south-eastern quadrant of the roundhouse structure are caused by the large deposits of pottery in F2773 and F2679 and the deposits of articulated sheep/goat remains in Pit F2517 and F2497. These deposits appear be associated with symbolic or ritual practices and the pottery vessels in the features associated with F2324 are considered to represent abandonment or termination deposits. This indicates that their presence in these locations is not representative of day to day domestic activity or due to the practical considerations of the light being greater in the part of the structure closest to the entrance as they appear not to represent the kinds of working practice that would require this extra light. It is impossible, however, to state whether or not these apparently symbolic deposits of pottery and animal bone were placed in these locations because the south-east was a propitious or sacred direction (Parker Pearson 1996, 119 and 127) or because they were positioned in relation to the entrance which coincidentally (and perhaps for practical reasons; c.f. Hingley and Miles 1985) was positioned in the southeastern part of the structure. The distribution of flint within Roundhouse 3 is very slightly biased towards the south-eastern quadrant. It did not occur in any great quantities or in any contexts that may be considered to be structured or ‘special’ deposits. It seems likely that this pattern of distribution is a result of the light being greater closer to the entrance and this is, therefore, a more practical place to carry out tasks utilising flint tools; the lack of debitage may be reasonably considered to preclude the possibility that flintworking occurred

The distribution and density of pottery (by weight) within the enclosing ring ditch F2324 was mostly quite uniform, but there was a marked increase in concentration in Segments Q and V, which lay close to, but not immediately at, the termini of the features that marked the probable entrance (Fig. 15). The distribution of animal bone within F2324 (Fig. 15) followed a similar pattern with a slightly 27

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire as the distribution of artefacts within the internal features of the roundhouse. This may be due to the postulated origins of L2494, representing the repeated relaying of a floor or rebuilding of a mound, into which the building’s structural elements were anchored, using material scoured from the encircling ring-ditch.

within Roundhouse 3. The distribution of daub within the internal features of Roundhouse 3 mostly occurs in features located in the approximate area of the walls of the structure; the only exception to this was the small quantity recovered from F2655 at the centre of the structure. Pottery recovered from layer L2494 was only recovered from the southern half of the roundhouse, and seemed to be concentrated on the approximate area through which the south-west to north-east aligned running baulk was maintained during excavation. The distribution of animal bone within L2494 was focussed on the eastern side of Roundhouse 3 with the greater concentrations occurring in the north-eastern quadrant. The distribution of artefacts within this deposit does not reflect a bias towards the south-eastern part of the overall structure quite so clearly

While there was clearly evidence suggestive of symbolic deposition of artefacts and other material within Roundhouse 3, the distribution of these items, and indeed of the other artefacts present within the structure, might be interpreted to support any of the postulated theories regarding artefact distributions within roundhouses. The biases towards the south-east quadrant in the distribution of both pottery and animal bone within Roundhouse 3, as indicated by Graphs 1 and 2, are heavily influenced by the

Figure 15. Roundhouse 3, distribution of finds within F2324

28

The Results of the Excavation Roundhouse 3 after complete removal of L2494

Figure 16. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature)

major deposits of these materials which have been identified as symbolic or ‘special’ deposits and which might be linked to the position of the entrance to the structure but which cannot be satisfactorily linked to the south-east being a direction of cosmological significance. The presence of symbolic or special deposits associated with the entrance to a structure might be anticipated as this is one of the most symbolically charged elements of a building, marking the transition from interior to exterior and vice versa. The entrance to the structure would have such symbolic importance no matter what its orientation.

The distribution plans (Figs. 16-21), although suggesting the presence of greater quantities of material in the southern half of the building, show no great bias towards the south-eastern quadrant or the entrance of the structure. This might simply be a reflection of the way in which practicality dictated that activities were subject to zonation within the roundhouse; for example it makes sense that, for a variety of reasons, activities involving fewer material objects, such as sleeping, occurred at the furthest point from the entrance. This might also be a result of other processes; for example, during cleaning it is most likely that the interior of the roundhouse would have been swept from inside to

29

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Roundhouse 3 after partial removal of L2494

Figure 17. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature)

outside, meaning that any small objects left lying around are likely to have been moved towards the doorway of the building, finding their way into structural, or other, features in this part of the roundhouse. On this basis, it may be argued that the distribution of finds within roundhouses in general might be considered to be influenced by the position of the entrance, symbolic factors associated with the entrance, and to be the result of taphonomic and other factors influenced by the position of the entrance. There was no evidence from Roundhouse 3 to indicate that the south-eastern portion of the structure was more symbolically important than any other part of the structure beyond the presence of the entrance here. Neither did Roundhouse

3 produce any evidence to indicate whether the entrance to the structure was placed in the south-eastern part of the building for cosmological reasons or practical reasons. What all of the theories expounded about roundhouse entrances fail to consider is that their position may simply have been dictated by social convention, as distinct from symbolic or cosmological conventions. Infant burial F2374 Pit F2374 (Figs. 6 & 22) lay directly opposite the entrance to Roundhouse 3, at a distance of 4m. It contained no dateable artefacts and had no stratigraphic relationships with 30

The Results of the Excavation Roundhouse 3 after removal of L2459

Figure 18. Distribution of pottery by weight (g) in internal features of Roundhouse 3 (total quantity per feature)

domestic dwellings is known from ancient Greek and Roman sites. Their burial within areas of settlement in Britain is observed from the middle Iron Age, often in shallow pits cut into ramparts or enclosure ditches, as foundation burials associated with structures, or as components of structured pit deposits (Moore 2009, 34). Human remains are potent symbolic objects and are used as meaningful objects by a variety of different cultures. During the late Bronze Age, human remains may have been used to give meaning to specific locations, perhaps being used to reinforce community identity or to denote ancestral and kinship rights to particular land or resources and it is possible that SK2375 represents a similar practice (Brück 1995, 260-261).

any other features, although it was sealed by the Phase 4 layer L2060, indicating that the feature had a terminus ante quem of mid 1st century AD. Like the earliest features of Roundhouse 3, it cut L2003, the natural substrate, and its position in relation to the roundhouse might be considered to suggest that the two were related. The feature contained a partial infant skeleton (SK2375). Analysis indicates that the individual was a neonate (see Phillips, Ch. 3.VI and Fig. 22). The burial of neonates in domestic contexts has a long history in Europe. Examples are known from the Neolithic and Bronze Age and the burial of neonates in proximity to 31

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Roundhouse 3 after complete removal of L2494

Figure 19. Distribution of animal bone by weight (g) within internal features of Roundhouse 3

Working surfaces and pits

postholes indicating a shelter or similar structure, such as the Phase 1 Structure S3012, were recorded.

A group of features to the north and north-east of Roundhouse 3 (Fig. 6) comprising three pits (F2135, F2157 and F2335), with a deposit interpreted during excavation as an occupation layer (L2478), a layer of burnt stones (L2635), and a clay floor (L2124) (Grid Squares H5, H6) have been interpreted as working surfaces and associated features. L2478 and L2635 contained material interpreted as cooking waste, including pottery and animal bone. East Midlands Scored Ware pottery was recovered from all of these features/layers. It is suggested that this group of features represents an open-air working area; no

Roundhouses 4, 5 and 6 Introduction The density of archaeological features present within the northern excavation area (AS1111) was not as great as that in the southern excavation area (AS857); however, amongst these are the three roundhouse structures Roundhouses 4, 5 and 6 (Figs. 5 & 6). These were located in close proximity to one another with the larger 32

The Results of the Excavation Roundhouse 3 after partial removal of L2494

Figure 20. Distribution of animal bone by weight (g) within internal features of Roundhouse 3

Roundhouse 4 positioned with the two smaller structures to its eastern side.

is potentially interesting. For example, these structures may represent a shift in the focus of settlement when Roundhouse 1 was replaced by Roundhouse 3 and its apparent ancillary structure. They could represent expansion of the settlement, perhaps occurring concurrently with the development of Roundhouse 3. Alternatively, they could have stood throughout the lifespans of both Roundhouses 1 and 3 and not been affected by the changes occurring to the south.

The spatial relationship that Roundhouse 4 displayed with Roundhouses 5 (F1086) and 6 (F1091 and F1093) (see Fig. 5) suggests that the three formed a group. Roundhouses 5 and 6, being smaller in size, might have been secondary structures within the group. A similar spatial arrangement of roundhouse structures was evident at Wardy Hill, between Structures I, III and V (Evans 2003, fig. 17).

Roundhouse 4

The development of these structures in comparison to the rearrangement of the domestic structures to the south

Roundhouse 4 (Grid Squares E20, F20, E21, F21; Fig. 23) was composed of an outer ring ditch (F5050) and an inner 33

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Roundhouse 3 after removal of L2459

Figure 21. Distribution of animal bone by weight (g) within internal features of Roundhouse 3

ring ditch (F5054) with several postholes present within the area enclosed by F5054.

in greater concentrations of weight in the south-eastern quadrant of the circumference of its ring ditches, close to the entrance to the area that these features enclosed. This conforms to the widely observed pattern of greater concentrations of artefacts occurring in the southern parts of the roundhouse, close to the entrance. The evidence, in the case of Roundhouse 4, appears too limited to make any contribution to the debate surrounding whether this pattern is due to cosmological/symbolic factors or considerations of practicality. The widely varying dates of the pottery recovered from this structure, however, might indicate that the distribution of artefacts is more closely related to the ‘ritualised’ house abandonment behaviour suggested by Webley (2007) and hint at longevity of occupation.

Ring Ditch F5050 yielded pottery of middle to late Iron Age date, later Iron Age, and early Roman date. It also contained cobbles, daub and animal bone. Ditch F5054 yielded a tiny amount of struck flint, pottery of middle to late Iron Age and later Iron Age date, further cobbles and animal bone. Most of the artefactual evidence from Roundhouse 4 was recovered from the upper fills of its ring ditches. The distribution of pottery and animal bone (Fig. 24) within the ring ditches of Roundhouse 4 clearly occurred 34

The Results of the Excavation

Graph 1. Total weight (g) of pottery recovered from Roundhouse 3, plotted by quadrant

Graph 2. Total weight (g) of animal bone recovered from Roundhouse 3, plotted by quadrant

case, it may represent an act of symbolic deposition comparable to that observed in Roundhouse 3. The double-ring arrangement of the gullies of Roundhouse 4 may be seen to suggest that the inner gully, F5054, had a structural function, similar to that suggested for Roundhouse 1. Ring-grooves to carry tightly packed upright timbers are a recognised form of roundhouse construction (Adkins and Adkins 2003, 74) and it has been noted at the reconstructed roundhouses at Butser Ancient Farm that rats tunnelled underneath the walls, palpably altering the archaeological evidence and leaving the walls represented by a ring gully as opposed to the expected ring of postholes (Reynolds 1979, 36). However, examination of the profiles of F5050 and F5054 demonstrates a pattern of infill that suggests that they had both been re-cut, possibly due to becoming silted up, perhaps making an enclosure or drainage function a more likely explanation than a structural function for both of these features.

Figure 22. Burial 2374

The pottery recovered from excavated Segment M, the terminus, of F5054 comprises a highly fragmented East Midlands Scored Ware jar, which may have been intact when it was deposited (see Peachey, Ch. 3.I). If this is the

The internal postholes of Roundhouse 4 displayed limited structural configuration. F5064, located close to the centre of the structure, may represent a central pillar. Posthole F5060, located close to the inner ring ditch in the 35

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 23. Roundhouse 4 plan and sections

36

The Results of the Excavation

Figure 24. Distribution of finds in the ring gullies of Roundhouse 4

south-eastern quadrant of the feature, contained a piece of worked sandstone which has been interpreted as a postpad. F5060 formed a pair with Posthole F5061 located just within the entrance to Roundhouse 4, perhaps suggesting a function associated with the door or entrance.

All of the pottery recovered from F5086 came from Segment A of this feature. The south-western end of Segment A was rounded and contained burnt material. It is possible that this implies that this was the original terminus of the ring ditch (rather than the result of truncation) and that there was, therefore, an entrance on the northwestern side of this Roundhouse, in contradiction to the more traditional south-east.

Roundhouse 5 Roundhouse 5 (Grid Squares G19, H19, H20; Fig. 25) was located c. 14m to the south-east of Roundhouse 4. It was represented by a fragmentary ring ditch, F5086, from which 3 sherds of East Midlands Scored Ware pottery (18g) were recovered. Only the northern half of the circuit of the ring ditch was present.

Roundhouse 6 The sixth circular structure (Fig. 25) recorded at Black Horse Farm (Roundhouse 6) was represented by a fragmentary ring gully (F5091; Grid Squares G21, H21, G22, 37

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Plate 7. A view north-west across Roundhouse 4 during excavation. The inner and outer ring ditches are clearly visible. This photograph illustrates the wet conditions and high levels of groundwater encountered during excavation

H22) and a single posthole (F5093; Grid Squares G21, H21, G22, H22). Neither feature produced any dateable finds.

Roundhouses 4, 5 and 6 displayed a marked lack of finds and firm evidence for domestic occupation, in comparison to Roundhouses 1 and 3. It is possible that this group of Roundhouses were simply not as affluent as the Roundhouses to the south or that they were not domestic structures and were not utilised in the same way. However, it is possible that the differences in deposit model between the area in which Roundhouses 4, 5 and 6 lay and that to the south offer an explanation. The northern area, despite displaying a layer of clay and gravel containing modern material and an alluvial deposit beneath the modern topsoil and above the natural into which the features were cut, did not display the buried ploughsoil that was present to the south. This buried ploughsoil, L2002, accumulated between the mid 1st and 2nd century AD and may have protected the Iron Age archaeology present in the southern part of the site from disturbance due to agricultural activity in later periods. The lack of internal floors and the truncation to later features F5109 and F5111 suggests a lesser degree of preservation among the northern features.

Roundhouse 6 had a diameter of c. 6m, approximately the same as that of Roundhouse 5, which lay to the south. This similarity, and the proximity of Roundhouse 4, is sufficient to suggest that Roundhouse 6 formed part of a coherent group with these other Roundhouse structures. Summary Within the similar group of structures at Wardy Hill, the smaller structures (to which Roundhouses 5 and 6 correspond) were termed ‘minor buildings’ (Evans 2003, 39). These minor buildings displayed some evidence of domestic use and cannot be shown to be directly contemporary to the larger Structure I (which is comparable to Roundhouse 4 at Black Horse Farm). The same, of course, may be said of the structures at Black Horse Farm. Indeed, Roundhouse 6 produced no dateable material and therefore cannot even be conclusively dated to the same phase as Roundhouses 4 and 5, although it seems likely that it was related. Their proximity to one another and convincing appearance as part of an interrelated group, suggest that, if they were not all built at the same time, they were at least all standing within the lifespan of the others.

The identification of two distinct groups of roundhouses (Roundhouses 4, 5 and 6 and Roundhouses 2 and 3) situated on the same fen-edge promontory is reminiscent of a settlement plan recorded at other Iron Age fen-edge sites. At the Upper Delphs, Haddenham (Evans and Hodder 2006, 38

The Results of the Excavation

Plate 8. A view across Roundhouse 6, looking to the west

270-271) and at Site I and the Camp Ground at Colne Fen, Earith (Evans 2013, 240-241) paired compounds lying at a short distance (c. 100m) from one another and containing evidence for Iron Age settlement occupation have been recorded. At Site I, Colne Fen, these compounds contained a similar pattern of buildings to those recorded at Black Horse Farm. The possibility that these two distinct groups of roundhouses at Black Horse Farm represent a further example of this settlement pattern suggests that the two groups were contemporary with one another.

Posthole F5082 (Grid Square G19) was sub-oval in shape, with steeply sloping sides and an irregular base. Its fill (L5083) was mid grey-brown clayey silt, with occasional flint inclusions. F5082 cut the north side of Ditch F5080. A group of four postholes (F5099, F5101, F5103 and F5105; Grid Square H18) lay to the south of Roundhouse 5, on the southern side of Ditch F5074. They formed a loose line running north to south. However, this alignment did not form a convincing structural configuration. Immediately adjacent (to the north-east) of the southernmost of these postholes, lay short linear feature F5107 (Grid Square H18).

Features and contexts in proximity to Roundhouses 4, 5 and 6 Irregular Ditch F5097 (Grid Square H18, H19; dimensions: 2.40m long x 0.80m wide x 0.17m deep; Figs. 5 & 6) cut Ditch F5080 at the point at which it intersected Boundary Ditch F5074. Ditch F5097 clearly post-dated the two Phase 1 ditches but the lack of artefactual evidence from this feature, and its lack of a clear functional relationship with any feature, meant that it could not be assigned to a particular phase. Its function was equally difficult to determine. It may have been a recut of Ditch F5080, but as this had no clear function and did not appear to have been part of the boundary system formed by Ditches F5074 and F5076, the function of F5097 remains uncertain.

Isolated Pit F5047 (Grid Square F20) lay c. 3.20m to the south-east of Roundhouse 4. It yielded no finds. Its upper fill (L5048) comprised dark brown-grey silty clay, with orange mottling, while its lower fill (L5049) comprised dark grey-brown silty clay. Its location may indicate that it was in some way associated with Roundhouse 4 but, due to the lack of associated dateable artefacts, this remains speculative. Two undated layers, L5095 and L5096 (Grid Square G22), were recorded to the north-west of Roundhouse 6. Neither of these layers produced any finds. Despite being recorded as distinct layers, Doyle and McCall (2008, 30) suggested 39

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Roundhouse 5 (see Fig. 5)

Roundhouse 6 (see Fig. 5)

Figure 25. Roundhouse 5 and 6

40

The Results of the Excavation after L2206 formed (which is most likely in the case of F2152) or that they were cleared out following the event that caused its deposition. L2206 was cut by Posthole F2311, which was undated but, as it was sealed by L2002, must be earlier than the 2nd century AD. L2206 sealed F2318 (Grid Square J4), which was a sub-circular pit. It lay c. 0.25m to the east of Phase 2 Ditch F2152, and c. 3m to the south of Ditch F2836. The location of F2318 and the features surrounding it, give no indication as to its function. It contained a small amount of animal bone (17g), in addition to early to middle Iron Age pottery (11g).

that these may in fact have been pockets of the overlying alluvial subsoil, which survived the initial machine stripping of the site, as they were trapped in undulations within the natural substrate. The northern enclosure ditches An east to west aligned ditch (F5074) traversed the site to the south of Roundhouses 4, 5 and 6. F5074 (Grid Squares D19, E19, E18, F18, F19, G19, H19; Figs. 5 & 6) measured in excess of 40m in length and was 1m wide at its widest point. The minimal artefactual evidence suggests that this feature was contemporary with the roundhouses that were located a short distance to the north.

III. Phase 2: The Later Iron Age Introduction

Although of much smaller dimensions than the enclosure ditches surrounding Roundhouse 3 and its ancillary structure Roundhouse 2 to the south, perhaps in part due to the differing preservation conditions between the two parts of the site, it is possible that this ditch represents part of an enclosure associated with Roundhouses 4, 5 and 6. The northern parts of this enclosure could potentially have been represented by Ditch F5089 (Grid Squares H21, H22; see Figs. 5 & 6) which ran on a northnorth-west to south-south-east alignment to the north-east of Roundhouse 4, and Ditches F5111 and F5109 which lay to the west of Ditch F5089 (Grid Square E22). They ran parallel to one another and were aligned north-east to south-west.

The horizon between Phase 1 and Phase 2 is marked by the shift from a pattern of ceramic consumption which may be described as middle Iron Age in character to one that may be considered to be typical of the later Iron Age. This shift is marked by the introduction of Belgic fabrics to the pottery assemblage. The earliest recognisable act within this second phase of activity is the creation of large ditches to the north and west of the area in which the Phase 1 Roundhouse 3, and its ancillary structure Roundhouse 2, were situated, possibly replacing the earlier system of ditches represented by F2385 (Fig. 26). Following this came the creation of further ditches surrounding Roundhouse 3, this time on its eastern side.

F5074 was cut by the undated north to south aligned Ditch F5076 (Grid Square F18). The similarities in the dimensions and fills of the two ditches and the perpendicular angle at which F5076 met F5074 suggest that they may have formed part of the same boundary system. F5080 (Grid Squares F19, G19, H19) was an irregular curving ditch, located 6.2m north of F5074 towards the west of the excavation area. It continued for a distance of c. 16m, at which point its eastern end met F5074. It then appeared to continue on a north to south alignment, parallel to F5076, suggesting that this also formed part of the same boundary system.

Embellishment and augmentation of the boundary system Ditches to the south and west of Roundhouse 3 Despite the fact that only 7.2m of its extent was visible before it was truncated by Ditch F2325, Ditch F2385 is considered, along with F2200 and F2836, to have formed an enclosure associated with Roundhouse 3. It appears that these features were the first element in a boundary system that was later enlarged and augmented. Ditches F2325 (Fig. 48; Grid Squares F4, G4, F5, G5, G6, H6) and F2738=F2808 (Grid Squares H3, G3, G4, F4, F5, E5; see Figs. 26 & 27) would appear to represent that process of enlargement.

Alluvial deposit Context L2206 (Grid Squares J3, J4, K4) was a layer located in the very south-eastern corner of the site. It was a dark blackish, greyish, brown coarse gravelly sandy silt; typical characteristics of an alluvial deposit (Whitten & Brooks 1975, 22). Such a deposit is likely to have been deposited during an episodic rise in the local groundwater table or by overbank flooding from a nearby watercourse, which, given the topography of the site, is most likely to have been present to the south of the main area of excavation.

These ditches formed a boundary system which extended beyond the limits of the excavated area. Indeed, F2738 continued for some way to the south; it was identified as Ditch F1121 in Trench 7 of the preceding trial trench evaluation (see Fig. 2). It is possible that these ditches may also have formed a sub-enclosure with possible Ditch Terminus F2105 (Grid Square G6), in the area of the modern drainage ditch, which bisected the site and with Ditch Terminus F2176, which lay to the north-east, on a similar alignment to Ditch F2325.

L2206 contained a small quantity of middle to late Iron Age pottery and was not recorded beyond Phase 2 Ditch F2836, which bounded it to the north or Phase 2 Ditch F2152, which bounded it to the west. There were no comparable fills in these ditches, indicating that they were cut

Ditch F2385 was much narrower than F2325, which truncated it, but the earlier feature would have been substantial 41

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire

Figure 26. Phase 2 plan

42

The Results of the Excavation

Figure 27. Boundary ditches F2325, F2378, and F2808

43

Iron Age Fen-Edge Settlement at Black Horse Farm, Sawtry, Cambridgeshire Table 4. Finds recovered from Ditches F2325, F2738 and F2808

Feature

Context

Finds

F2325

L2327

-

L2326

Animal bone 116g

L2328

Pottery (22) 424g; CBM 74g; Animal bone 1425g

L2366

Pottery (13) 214g; Animal bone 313g; Fired clay (1) 16g; Stone (1) 63g

L2367

Pottery (1) 32g; animal bone 179g; Fired clay (35) 300g; Burnt clay (8) 84g

L2382

Pottery (6) 101g; CBM 1143g; Animal bone 105g

L2373

Pottery (1) 14g; Animal bone 55g

L2682

Pottery (16) 98g; Animal bone 1066g

L2702

-

L2372

Pottery (6) 112g; Animal bone 568g

L2377

Pottery (17) 468g; Animal bone 652g; Burnt bone 1g; daub (1) 4g; Possible Quernstone Fragment (1) 1116g

L2361

Pottery (56) 958g; Animal bone 396g; SF4 Cu Alloy Fragments (2)