A prehistoric and Saxon site at Thrapston Road, Spaldwick by Susan Clelland and Lorraine Mepham with contributions by Catherine

Barnett, Lorrain Higbee and Chris J. Stevens Summary The excavation of an area adjacent to Thrapston Road, Spaldwick, revealed archaeological remains ranging in date from early prehistoric to post-medieval, focusing in particular on the Iron Age and Saxon periods. Ephemeral traces of a transient early prehistoric use of the site were succeeded in the Iron Age by a trackway and a succession of small enclosures, probably related to a mixed farming settlement. The site was apparently abandoned during the Romano-British period, but was reoccupied during the early/mid-Saxon period, at first for possible small-scale industrial use, before being formally reorganised in the late Saxon period into tofts, each containing rectangular post-built structures. Later, during the medieval period, the site became part of an agricultural field system. INTRODUCTION Wessex Archaeology was commissioned in 2010 by SD Construction and Developments Ltd to carry out an archaeological excavation at 33 Thrapston Road, Spaldwick, centred on NGR 512780 272917 (Figure 1). This followed on from an earlier desk-based assessment and evaluation of the site, both also carried out by Wessex Archaeology, in 20012 (Wessex Archaeology 2002a; 2002b). Site location and geology The site, which lies on the north-western fringes of the village of Spaldwick, comprised at the time of excavation a parcel of overgrown scrub covering approximately 0.12 hectares. Located on a gentle north-east facing slope, the site is bounded to the north-east by Thrapston Road, originally part of the road running from Huntingdon to Thrapston, which forms the village High Street, and which is now bypassed by the A14. The site is bounded to the north-west and south-east by

residential properties and to the south-west by further overgrown scrubland backing on to St James Church. The underlying geology is shown by the British Geological Survey (Sheet 2) as comprising Oxford Clays with Kellaway Beds. Archaeological and historical background The known archaeological and historical context of the site is limited, with little evidence for prehistoric settlement known from the immediate vicinity, although Iron Age ditches were recorded in a pipeline trench within 1km of the site. The Roman road from Leicester (Ratae) to Godmanchester (Durovigutum) lies 5km to the north. The manor of Spaldwick is first mentioned in the 10th century when it was granted to Ely Abbey by Brithnoth, earldorman of Essex, in AD 991. From 1109 to 1543 it was held by the Bishop of Lincoln who may have had a palace or hunting lodge to the east of the site, near to his deer park (the 12th century Forest of Spaldwick), of which there is now no trace (Wessex Archaeology 2002a). The banked enclosure around the presumed historic core of the village (Figure 1) is the site of the shrunken village of Danesfield, although the name Spaldwick (Spalduice) was in use by Domesday. The settlement of Danesfield may have been in existence before AD 991 when Spaldwick was granted to Ely Abbey. The enclosure is roughly oval with the south and west sides marked by a small surviving bank; the bank on the northern and eastern sides has been destroyed. The earthworks within the enclosure (see Figure 1) may range from building foundations, in the south, to a windmill mound or perhaps a very small motte in the west. Previous archaeological investigations within the village of Spaldwick found two Saxon or early medieval buildings succeeded by later medieval rubbish pits in yards with some industrial activity on a site on the northern side of Thrapston Road (CCCAFU 1996), and a medieval ditch and plough marks were recorded at Ferriman Road, to the south of the site (Murray 1998) (Figure 1).

The many 16th, 17th and 18th century Listed Buildings in Spaldwick may indicate the prosperity of the village, though none of these lie close to the site, and map regression shows that the village has changed little from 1775 (Wessex Archaeology 2002a, fig. 3). The line of the earthwork enclosure around Danesfield may be seen on the 1775 Enclosure map, curving around and running along the northern edge of the site (see Figure 1). EXCAVATION RESULTS The modern overburden generally comprised a mixed layer of made ground and topsoil, typically a mid- to dark grey-brown silty loam, above subsoil. Across the site this sequence of deposits varied in thickness from 0.4m at the northern end of the site up to 1m at the southern end. Archaeological features and deposits were noted both within the base of and below the subsoil (Figure 2). Early Prehistoric Evidence for a human presence on the site during the early prehistoric period (NeolithicBronze Age) was indicated by a small assemblage of worked flint, recovered as residual finds from features and deposits of later date. The assemblage, mainly comprising flake debitage, also included two scrapers likely to be of Neolithic date (topsoil finds) and one dating to the earlier Bronze Age (from Late Iron Age pit 663). In addition, a sherd of Middle/Late Neolithic Peterborough ware came from an otherwise undated pit (532), and single, abraded body sherd from a Beaker vessel was found in the topsoil. MidLate Iron Age Much of the complex series of intercutting features revealed on the site has been dated to the Iron Age (Figure 3). This includes a probable track- or droveway, field boundary ditches, a succession of small, subrectangular enclosures (784, 792, 794, 798), a possible ring-gully (790), and a number of pits. Stratigraphic relationships indicate a sequence of activity, in which four phases have been defined here, although the pottery evidence is insufficient to confirm this, or the timescale within which it could have occurred (see below).

Phase 1: Trackway Two parallel shallow gullies (averaging 0.3m wide and 0.1m deep), excavated in four sections (758, 774, 787, 796), form a probable track or droveway, running south-west to north-east across the site, and diverging at the south-western end. Eight sherds of pottery, four worked flints and two pieces of animal bone, all very small and abraded, constituted the only finds from these features. Phase 2: Field boundaries The phase 1 trackway ditches were not continuous, but their shallow nature suggests that they might have been truncated. One section (gully 758) was cut by a substantial ditch (783) running roughly east-west across the site. Ditch 783, together with a parallel ditch (628) partially revealed at the southern end of the site, appears to mark a formal demarcation of boundaries, approximately 30m apart, within the landscape. Ditch 783 was U-shaped, 2.5m wide and 1.1m deep (Figure 4). The associated fill sequence suggests that a bank, mounded along its northern edge, was later levelled into the ditch following the accumulation of a secondary deposit (733, 734) from which the majority of the finds from the ditch were recovered (73 pieces of animal bone; 1 piece of worked flint; 33 sherds of pottery). Ditch 783 may only have been in use for a short period of time before there was a change in land-use whereby the area appears to have been levelled and cleared. Phase 3: Enclosures and associated features Parts of three sub-rectangular enclosures were uncovered (784, 792, 794), with a possible fourth (798) just clipped by the southern end of the trench. Enclosure 792 may be the earliest its ditch was cut by both 798 and 794, and ended in a terminal at the eastern end, indicating an east-facing entrance. The ditch was on average 1.2m wide and 0.5m deep with two distinct episodes of infilling (Figure 4). The lower deposit (574) contained a concentration of charcoal and heat-affected flint nodules, which were mainly found along the base of the ditch. Overall, ditch 792 produced 111 fragments of animal bone and five sherds of pottery, as well as a fragment from a saddle quern, and the majority of this material derived from the upper fill of the ditch (575).

Ditch 798, of which only a small section was revealed at the southern end of the site, is thought to represent the north-eastern corner of an enclosure, cutting enclosure 792; it could conceivably have been a recutting of the latter enclosure. The ditch appeared to be on a similar alignment to 792, but had a profile differing slightly from the latter, and also from enclosures 784 and 794. There was a shallow outer lip to a deeper, Ushaped ditch, with a total width of 1.2m, and a depth of up to 0.4m (Figure 4). The ditch yielded a small assemblage of animal bone (50 fragments), pottery (13 sherds) and worked flint (three pieces), all from secondary fills. Enclosures 794 and 784, perhaps adjacent to each other, and differently aligned to 792 and 798, were exposed along the western side of the excavated area. Enclosure 784, of which the south-east corner was revealed, was aligned north-north-east to southsouth-west, with an east-facing entrance 3m wide. Cutting through enclosure 792, enclosure 794, to the south-west of enclosure 784, was aligned north-eastsouth-west and represented by the north-eastern corner. The north-eastsouth-west section of this ditch ended in a terminal at the south end, indicating a south-east-facing entrance. The dimensions, profiles and fill sequences of enclosure ditches 784 and 794 were similar in both instances to those of 792, but both enclosures were more productive in terms of pottery: 80 sherds from 784, and 40 from 794. A chalk spindlewhorl was also recovered from enclosure 784. An environmental sample taken from ditch 784 contained significant quantities of emmer and spelt wheat, as well as seeds of several wild species. The original size of the enclosures is unknown, although 784 appears to have been no more than 20m from north to south. Stratigraphic relationships indicate that enclosure 792 was earlier than both 798 and 794, but the chronological relationships of these three to 784 are unknown. The pottery is only marginally helpful, suggesting, in the absence of grog-tempered wares from 792, that this enclosure may have had a Middle Iron Age origin, the other three enclosures perhaps dating to the Late Iron Age, but quantities are too small to provide firm evidence here (only five sherds came from enclosure 792). Certainly the pottery groups from 784, 794 and 798 are very similar (mixed shell-tempered and grog-tempered), and suggest that all three enclosures could have been in use (or at least the ditches backfilled) over a relatively short timespan, perhaps no more than a century. Five Romano-British sherds were recovered from the

upper fill of 794, perhaps later finds in the top of the feature. The precise provenance of two early Saxon sherds from enclosure ditch 784 is unknown. The only internal features recorded within the enclosures comprise a group of five intercutting pits (800) of varying sizes inside the north-eastern corner of enclosure 794. These pits cut phase 1 gully 796, but only the latest pit (612) produced any datable finds (14 sherds of pottery, all shell-tempered). Outside the enclosures, a curvilinear feature (789) is of uncertain function. It had been recut along its southern side by gully 790. It could represent part of a roundhouse ring-gully, except that what was present appeared not to be circular or penannular, and it lies outside any of the identified enclosures. Seven pottery sherds, shell, fired clay fragments and animal bone were recovered from the gully fills. All seven pottery sherds are in shell-tempered wares, which could suggest a Middle Iron Age date, and therefore probably pre-dating enclosures 784 and 794. Gully 790 was also cut by another ditch (788), which ran parallel to the eastern side of enclosure 794, but whose function is uncertain; this ditch produced eight sherds of pottery, including both shelly and grog-tempered wares. Apart from pit group 800, nine other pits have been assigned to this period of activity (440, 549, 693, 695, 701, 762 and group 801). These were mostly small and contained single fills and very small quantities of finds, including animal bone and a few sherds of pottery. Several small abraded sherds of Iron Age pottery recovered from posthole group 802 may be residual. The function of this small group of at least four intercut postholes, each approximately 0.4m in diameter, is unknown. Late Iron Age/Romano-British The seven Romano-British sherds from enclosure 794 have been noted above. A further 16 sherds of Romano-British pottery were recovered from various features across the site; mainly as residual finds (for example, in Saxon pit 663, and medieval ditches 780, 782 and 791). No features could be definitively dated to this period, and it appears that the site was abandoned, possibly around the time of the conquest.

Early/mid-Saxon Evidence for activity in the early/mid-Saxon period comprises a number of pits (Figure 5). The base of large oval pit 663 (2.3m x 1.6m x 1m deep), adjacent to the eastern edge of the excavated area, cut through a sandy layer within the Oxford Clay, breaching the water table, and has been tentatively interpreted as a watering hole; the profile shows one steep side and the other more gradually sloping, which could have allowed access for animals to drink (Figure 6). The base of the pit was filled after going out of use through the gradual accumulation of laminated silt deposits (666 671), and this was followed by a series of dumped layers (6724); subsequently, late Saxon-medieval boundary ditch 791 cut through the top. Pit 663 produced 12 early/mid-Saxon sherds, as well as animal bone (64 fragments) and a possible iron smithing hearth bottom. Nine shallow, sub-rectangular fire pits, forming two clusters (433, 729 and 752 to the north; 435, 705, 715, 724 and 725 to the south) with one outlier (450), have been assigned to the early/mid-Saxon period, although the only dating evidence recovered comprised four early/mid-Saxon sherds from two intercutting pits (435, 725). Within the two clusters the pits were located close enough together to perhaps imply successive rather than contemporaneous use, a suggestion supported by the fact that pit 435 cut the corner of pit 725. The pits were all of broadly similar shape and size (1.21.8m long, 0.81.2m wide, 0.10.2m deep; near vertical sides and flat bases), and most had similar burnt material in their fills. The sides of most of the pits were heat-affected, although this was not consistent or continuous around the sides of any of the pits, while the pit bases were only slightly heat-affected. The depositional sequence recorded in pit 724 (Figure 6) is somewhat different to the others, and may provide an insight into how the pits may have functioned. A dense layer of charcoal (723), mainly consisting of large pieces of roundwood, lay at the base of the pit. This was overlain by a layer of slightly heat-affected flint nodules (722), apparently dumped into the pit rather than in situ. A deposit of accumulated debris comprising a fine dark silty loam with a distinct ashy grey hue and small fragments of charcoal (721) sealed the layer of stones. Layer 721 was sealed by an upper fill of topsoil-derived material (720). The remaining eight pits were filled with a mixed layer of silty loam including abundant charcoal and heat-affected flint nodules 7

of varying size (Figure 7). The density and size of the charcoal increased towards the bases of these pits, but there were no clearly differentiated charcoal and stone deposits as seen in pit 724. Apart from the four sherds of pottery noted above, other finds from the pits comprised 244 fragments of animal bone (mostly cattle and sheep/goat, but also a few pig, red deer and bird bones), 33 fragments of fired clay, 159 pieces of burnt, unworked flint, and one piece of ironworking slag. There was a concentration of finds in pit 433 (all of the burnt flint, most of the fired clay, and approximately half of the animal bone). The possibility that these pits may be slightly later in date, perhaps late Saxon, cannot be ruled out, although the southern cluster was cut by timber structure 799 (see below). Two other pits (455, 525), both on the eastern side of the excavation area, may also belong to this phase of activity; both produced early/mid-Saxon pottery (three sherds from pit 455, 20 from pit 525) and, although one sherd of medieval pottery was also recovered from 525, this was small and could well have been intrusive. Pit 525 also produced 125 fragments of animal bone and two pieces of ironworking slag. Late Saxon/early medieval Two right-angled gullies (539 and 793) appear to be related, forming a pair, and may be the earliest features in this phase (Figure 8). These two gullies ran parallel across the southern half of the site before turning to the south-west, though their full extent is unclear. Pottery indicates a late Saxon/early medieval date (three sherds of St Neots ware and one of early medieval shelly ware, alongside nine early/mid-Saxon sherds), and both gullies were cut by later boundary ditches (780 and 791 respectively).

The main focus of activity in this phase, however, was the construction of two timber structures, probably domestic buildings, and associated fences within clearly defined ditched boundaries (tofts). This marks a significant period of landscape reorganisation, and may represent the earliest evidence for the village of Danesfield

(Figure 8). The excavated area constitutes the majority of one toft, and part of a second, both oriented north-east to south-west, and respecting the present-day alignment of Thrapston Road along the north-east side of the site. The tofts were separated by ditches 780 and 791, with ditch 782 to the north. The main toft exposed extended for at least 37 m, and tapered from 17m wide at the northern end to 11m at the southern end. Ditch 782 along the northern side exhibited at least three episodes of recutting, by ditches 422 and 587 (Figure 9), resulting in a broader ditch (3.6 m wide, 1.1m deep) than the two forming the sides of the plot. Deposits within ditch 782 suggest the presence of a bank to the south, apparently backfilled during the medieval period. The two side ditches (780 and 791) were between 1m and 2m wide, and approximately 0.5m deep. Thirteen sherds of early/mid-Saxon pottery were recovered from the eastern ditch 791 (which cut possible waterhole 663), one medieval sherd from western ditch 780, and 20 Late Saxon/early medieval sherds from ditch 782 recuts 422 and 587 to the north. Within the central toft exposed lay a rectangular timber structure (799), just over 20m south of ditch 782, and following the same alignment as the side ditches, but closer to ditch 780. Structure 799 comprised 12 postholes, and measured 8m by 4m. The postholes varied in size and the profile and plan of several indicate that posts were replaced during the structures use. The two postholes on the east side may indicate the location of a centrally placed doorway in this wall. Three postholes produced early/mid-Saxon pottery (a total of 13 sherds), but the sherds are small (mean sherd weight 4g) and are considered likely to be intrusive here. Furthermore, one of the postholes in the east wall cut one of the earlier Saxon fire pits (715), and the projected wall line passed very close to three others. To the north of structure 799 was a probable yard. A number of postholes formed two sides of a fence alignment (786) running parallel to the eastern and northern boundary ditches (786 and 791 respectively), and at a distance of about 4m inside them. Two small pits (448, 490) and posthole 776 comprised the other contemporary features identified in this area, and these produced a total of only four sherds of late Saxon/early medieval pottery.

A second possible rectangular structure (803), on a similar alignment to structure 799, was more tentatively identified within the plot or toft to the west. Structure 803 comprised ten postholes forming a ground plan measuring an estimated 10m by 4m. No dating evidence was recovered from any of the postholes. Pit 543, probably a rubbish pit, to the north-west of structure 803, yielded two sherds of Late Saxon St Neots ware. Late medieval and post-medieval The few remains dated to the later medieval and post-medieval periods (Figure 8) suggest that the area reverted to agricultural use at this time. Late medieval feature 804 comprises an 8m wide series of ditches 10m to the south of and parallel with Thrapston Road. Parallel and less than 3m to the south was the late Saxon-medieval ditch 782 bounding the north side of the tofts. It is possible that feature 804 represents a hollow-way, with possible origins in the late Saxon period, comprising a sequence of linear hollows and ditches, the latter perhaps drainage ditches which broadly confined the shifting course of the route. Cobble patch 404 may also have been an element of this complex, and the gravel in the lowest fills (585 and 589) of the recuts (422 and 487) of ditch 782 may derive from similar metalled surfaces of which no other trace survived. During the post-medieval period the site was levelled, sealing feature 804. The remains of a street frontage garden/property wall (402), mirroring the alignment of Thrapston Road, lay at the northern end of the site, with a post-medieval pit (406) beyond. FINDS Pottery by Lorraine Mepham, with a contribution by Patrick Quinn The complete pottery assemblage recovered amounts to 558 sherds (5475g), and includes sherds of early prehistoric, late prehistoric, Romano-British, early/midSaxon, medieval and post-medieval date, although the majority belongs to the late prehistoric period. Only the prehistoric (271 sherds; 3042g) and Saxon material (109

10

sherds; 652g) is discussed here; details of all other pottery are held in the project archive. Methods of analysis The prehistoric and Saxon pottery has been subjected to full fabric and form analysis, following the standard Wessex Archaeology pottery recording system (Morris 1994), which accords with nationally recommended guidelines for the recording of prehistoric pottery (PCRG 2010), and post-Roman pottery (MPRG 2001). Fabrics have been defined and coded on the basis of predominant inclusion type; totals and summary descriptions are given in Table 1. To support the fabric analysis, samples of ten selected fabric types (four prehistoric and six Saxon) were submitted for thin section analysis by Dr Patrick Quinn (University College, London). His full report is held in the project archive, and a summary of his results is incorporated below. Early prehistoric Two early prehistoric sherds were recovered; both are small and abraded. The first is a body sherd in a very coarse, flint-tempered fabric, with traces of impressed whipped cord decoration on the external surface; this can be identified as Middle/Late Neolithic Peterborough ware. It was the only sherd from pit 532. The second is a body sherd in a grog-tempered fabric, with comb tooth decoration, clearly identifiable as Beaker. This came from the topsoil. Late prehistoric The condition of the later prehistoric assemblage is fair; mean sherd weight is 11.3g, but this is somewhat biased by the presence of a few large sherds; on the whole, sherds are small (<10g), and lightly to heavily abraded. The inclusions in some of the shelly fabrics have leached, leaving voids.

11

Fabrics by Lorraine Mepham and Patrick Quinn Eight fabric types have been defined for the later prehistoric assemblage; these include grog-tempered (Group GR), sandy (Group QU) and shelly fabrics (Group SH), although sandy fabrics are represented by just one small body sherd. Samples of two grog-tempered fabrics (GR1 and GR3) and two shelly fabrics (SH1 and SH4) were thin-sectioned. Both grog-tempered samples were found to be closely related petrographically, both characterised by abundant grog temper and fossil shell inclusions within a fine, silty, non-calcareous clay. The two samples vary only in the relative frequencies of grog and fossil shell within each. Some of the grog fragments contain fragments of fossil shell, suggesting the recycling of ceramics similar to the parent fabric, but none contain second-generation grog. The shell material in SH1 and SH4 is clearly related to that seen in GR3, and SH4 also contains sporadic grog fragments. There are clear similarities between the four fabrics petrographically, and all could have been produced using local resources, within the sedimentary rocks of the Jurassic Oxford Clay Formation. The Oxford Clay is currently extracted for brickmaking in several parts of the outcrop, and could therefore have been a suitable raw material for manufacturing ceramics. Shelly material could be sourced from the Stewartby Member of the middle Oxford Clay. Vince, in his analysis of other Iron Age ceramics from Cambridgeshire, suggested the Jurassic Cornbrash or lower Oxford Clay to be the source of the fossil shell used (Vince 1997; 2006; 2007a), but the Cornbrash is not present in the general area of Spaldwick. Other grog-tempered and shelly fabrics (GR2, SH2 and SH3) are clearly also related to this group, varying only in the frequency, size, and relative proportions of the grog and fossil shell inclusions. The single sandy sherd (QU1) stands out as anomalous within this group and, indeed, is not even definitively of Iron Age date it came from the topsoil.

12

Forms A restricted range of vessel forms was identified. Of the 26 rim sherds recovered, 21 are diagnostic at least to partial profile, and these fall into seven vessel forms (see Table 2). Other, less diagnostic rim sherds are likely also to belong to this range of forms.
1. 2. Vessels with beaded rims and convex, neutral profiles (R3, 7 examples; Fig. 10, 12) Vessels with plain or slightly thickened (proto-bead) rims and convex, neutral profiles (R7, 3 examples; Fig. 10, 34) 3. 4. 5. 6. 7. Vessels with everted rims and convex, neutral profiles (R2, 3 examples; Fig. 10, 5) Vessels with sharply everted rims and rounded, open profiles (R1, 2 examples, Fig. 10, 6) Vessels with upright, plain rims, shouldered (R8, 2 examples, Fig. 10, 7) Open vessels with plain, upright rims (R9, 1 example, Fig. 10, 8) Everted rim on necked form, profile unknown (R5, 3 examples, not illustrated)

Only eight vessels have measurable rims (examples of types 1, 2 and 4); these range from 140mm to 180mm, with a slight focus on the upper end of this range. These vessels could be said to fall within the small to medium size range, following work on vessel size and possible function on other mid- to late Iron Age assemblages from Cambridgeshire (e.g. Hancocks 2003, 901; Hancocks et al. 1998, 745). Decoration is confined to one type 6 vessel in fabric SH2, with a finger impression on the outside of the rim (Fig. 10, 8), and two pinched cordons on grog-tempered body sherds. Surface treatments, however, are more apparent in the assemblage. A number of sherds (29) carry surface scoring on external surfaces, although this may have as much of a functional as a decorative function. Scoring can be horizontal (Fig. 10, 1) or vertical; some vessels combine both (Fig. 10, 4). Both shelly and grog-tempered vessels are scored, and diagnostic scored sherds belong to vessel types 1, 2 and 5. One small body sherd in shelly fabric SH3 is externally burnished. No sooting or other surface residues were observed.

13

Discussion Shelly and sandy wares constitute the two main fabric groups in use during the middle to late Iron Age in Cambridgeshire (Percival 2008, 5); grog-tempered wares rarely appear before the late Iron Age. In this instance, sandy wares are virtually absent, and the assemblage is dominated by shelly wares (89% by weight) with a smaller proportion of grog-tempered wares (11%). This suggests that this assemblage may have an origin in the mid- Iron Age, but that there is a definite late Iron Age component. This is supported by the vessel forms represented mainly convex vessels with simple or pulled bead rims, occasionally scored, with a few necked and cordoned vessels. The slack-shouldered vessels, which are the most commonly represented form in most mid- to late Iron Age assemblages in Cambridgeshire (ibid., 3) are scarce here, and occur only in shelly wares. In terms of parallels with other mid- to late Iron Age assemblages from Cambridgeshire, Spaldwick shows little overlap with the mid- Iron Age assemblage from Cambourne, either in form or fabric (Leivers 2009), while other nearby sites along the route of the A428 provide parallels for the shouldered forms, but show the reliance on sandy and shelly fabrics characteristic of sites in the south of the county. Closer parallels can be found in the west of the county, from sites such as Little Paxton, Hinchingbrooke and sites along the route of Ermine Street (Hancocks 2003; Hancocks et al. 1998). At Little Paxton, the Spaldwick assemblage appears to overlap with phases 3 (Late Iron Age handmade) and 4 (Late Iron Age/Transitional), but lacks the wheelthrown wares of the latter phase. The period 1 (Late Iron Age) assemblage from Tort Hill West on Ermine Street contains similar proportions of grog-tempered and shelly wares (Hancocks et al. 1998, table 5). Overall, while the Spaldwick assemblage may have Middle Iron Age origins, it seems to fit best within the Late pre-Belgic Iron Age range of c. 100 BCAD 20 (ibid., 77). Distribution Late prehistoric sherds were recovered from 27 cut features either phased as prehistoric on stratigraphic grounds, or presumed to be of prehistoric date solely on datable pottery; a relatively small number of sherds (24) occurred residually in later

14

features and topsoil contexts. However, the numbers of sherds from individual features within this overall distribution is generally very low. Only three features produced more than 20 sherds (ditch 794, 26 sherds; ditch 783, 35 sherds; ditch 784, 78 sherds). The group from ditch 784 stands out not only by reason of its size, but also for a relatively high average sherd weight (16.9g, while ditches 783 and 794 have average sherd weights of 9.1g and 9.7g respectively). Vessel forms represented in ditch 784 comprise types 1, 2 and 4: Fig. 10, 13). There is some suggestion of a ceramic sequence through the ditchs fills: in one section, the primary fill contained only shelly wares, whereas grog-tempered wares appear in the secondary deposits. In terms of patterning across the site and through the stratigraphic sequence, however, there is little more that can be drawn from this small assemblage. Both shelly and grog-tempered wares occur from the earliest stratigraphic phases. One of the two examples of the shouldered vessel form (type 5), which might be considered an early form within this assemblage, came from the stratigraphically latest feature (enclosure ditch 794), which also yielded seven Romano-British sherds, including two of decorated samian. Early/mid-Saxon The identification of a Saxon element within the assemblage is somewhat problematic. A total of 109 sherds have been dated as early/mid-Saxon solely on the grounds of fabric; there are no clearly diagnostic sherds. Nearly all fabrics are sandy and are almost entirely unoxidised; these are harder-fired than those sandy wares identified as later prehistoric (although petrographic examination indicates that all were fired to less than 850C, the same as the prehistoric fabrics), and coarser. The condition of the sherds is fair; despite their hard-fired nature, sherds on the whole are smaller than in the prehistoric assemblage, and are at least slightly abraded. Mean sherd weight is 6.0g.

15

Fabrics by Lorraine Mepham and Patrick Quinn A total of eight fabrics has been identified (Table 1), and samples of six were thinsectioned (Q400, Q401, Q402, Q404, Q405, Q406). The Saxon fabrics are compositionally distinct from the prehistoric sherds analysed in that they are dominated by quartzose inclusions rather than shell and/or grog. In all but one sample (Q402) this material appears to have derived from sandstone. The source of this sandstone material is not entirely clear. Sandy layers occur within the Oxford Clay and in the Kellaway Beds below it. However, these are doggers or sandy clay rather than pure sandstones of the type seen in these fabric samples. Nevertheless, the calcareous sandstone seen in Q402 could have derived from a calcareous cemented sandy layer in the Oxford Clay or other Jurassic marine sedimentary unit. Vince found calcareous sandstone inclusions in Iron Age ceramics from Cambridgeshire and interpreted it as having a Lower Cretaceous origin (Vince 2006); he also found calcareous sandstone temper in Saxon ceramics from the county, which he interpreted as the Jurassic Spilsby Sandstone (Vince 2007b). Non-calcareous quartz arenite sandstone (as in Q400) does not occur as a primary outcrop in the region around Spaldwick and the nearest in situ sandstone outcrops may be some distance away. However, eroded sandstone clasts could be present in the glacial till or fluvio-glacial material that covers the Oxford Clay in places. Given that the sandstone clasts in several of the samples are thought to have been added as temper (Q400, Q401, Q406), it is possible to envisage the selection of glacial erratic for this purpose. Vince also attributed the presence of sandstone inclusions in Saxon ceramics from Cambridgeshire to the use of boulder clay (Vince 2007b and c). Forms There are only four rim sherds amongst the Saxon assemblage. One of the rims (from fire pit 725) is from a fairly thick-walled vessel with an upright, rounded rim. The other rims are simple and rounded in profile, but the rim orientation and overall vessel profile are unknown in all cases. These do not, therefore, help to confirm the dating of

16

the assemblage. No decoration is present, and only eight sherds (three in fabric Q401, four in Q403, and one in Q404) are burnished. Discussion The dating of this small group of sherds, then, has to rely heavily on parallels with the fabric types, and there are similarities with early/mid-Saxon wares identified at other Cambridgeshire sites, such as Cambourne New Settlement and Eynesbury (Seager Smith 2009; Mepham 2004), as well as the assemblages previously analysed by Vince (2007b and c). In the absence of a well-understood and well-dated early/mid-Saxon ceramic sequence for the region, the Eynesbury assemblage was dated broadly, on typological grounds, to the 5th7th centuries, and it is likely that the Spaldwick assemblage falls within the same date range. Only 16 Saxon sherds came from features stratigraphically phased as Saxon (12 from pit 663, 2 from fire pit 435 and 2 from fire pit 725). In addition, 20 sherds from pit 525 may serve to date this feature, if one small medieval sherd can be considered as intrusive, while three sherds from pit 455 have been taken as dating evidence, albeit tentative. A further 13 sherds formed the only dating evidence in three postholes in timber building 799, but their small size and highly abraded nature precludes their use as firm evidence. All other Saxon sherds came from later features and topsoil contexts.
List of illustrated vessels (Figure 10) 1. Beaded rim (R3); shallow horizontal scoring; fabric SH2. PRN [Pottery Record Number] 23, context 468, enclosure ditch 784. 2. 3. 4. Beaded rim (R3); fabric SH2. PRN28, context 470, enclosure ditch 784. Rim, convex vessel (R7); fabric SH3. PRN138, context 678, enclosure ditch 784. Rim, convex vessel (R7); deep multi-directional scoring; fabric SH3. PRN102, context 610, pit 612. 5. 6. 7. Rim, everted, on convex vessel (R2); fabric SH4. PRN197, context 773, ditch 788. Rim, everted, on rounded, open vessel (R1); fabric SH4. PRN15, context 432, ditch 783. Shouldered vessel with upright rim (R8); fabric SH4. PRN202, context 623, enclosure ditch 794.

17

8.

Open, convex vessel with plain rim (R9), single ?finger impression on outside of rim; fabric SH2. Context 575, enclosure ditch 792.

Animal Bone by L. Higbee A total of 1,348 fragments (or 10.768kg) of animal bone was recovered from the site; this is a raw fragment count and when adjusted to take account of refits the figure falls to 1,191 fragments. The majority of this material comes from Iron Age and early/midSaxon contexts (Table 3), but the number of identified fragments from each period is quite small and this limits its potential for detailed analysis. In terms of species proportions, the Iron Age and late Saxon/medieval phases appear to be sheep/goat dominated, while the early/mid-Saxon phases appear to be dominated by cattle. Whether or not this apparent difference in the relative importance of livestock species is real and reflects a shift in husbandry strategy between the local Iron Age economy and that of the Saxon period, and then back again, is uncertain given the small sample size. It is also difficult to assess how the Spaldwick assemblage fits with regional period trends since variations in the relative importance of sheep/goat and cattle have been noted for both main periods (Hambleton 1999, 46; Crabtree 2010). Analysis of body parts and age data indicates that livestock were raised locally and slaughtered on-site, and suggests that the local economy was a self-sufficient producer of meat and secondary products. Less common species include pig, horse, dog, red deer (mostly antler), domestic fowl and cat. The latter is from an Iron Age context and is likely to belong to a wildcat rather than a domestic animal. Other finds A small fragment of an antler comb was recovered from possible early/mid-Saxon pit 525; this is part of the tooth plate from a double-sided composite comb of a type

18

current from the Romano-British through to the medieval period (MacGregor 1985, fig. 51). Of the 48 pieces of worked flint recovered, most is flake debitage, and not chronologically distinct. Of the tools, the three scrapers are of varied date: a large end scraper made on a tertiary trimming flake in a distinctive banded flint is likely to be (perhaps later) Neolithic; an end scraper on a pale brown blade with neat edge damage on both margins is likely to be Early Neolithic or even Mesolithic; a somewhat shapeless end and side scraper on a secondary flake struck from a small pebble is likely to be earlier Bronze Age. There are also two gun flints of post-medieval date. Raw material was mostly a pale to dark brown inclusion-free flint, with a worn cortex indicating a source in the local drift geology. A small quantity of slag was recovered. Some of this clearly represents iron smithing slag, and includes one possible hearth bottom, from early/mid-Saxon waterhole 673. However, the small assemblage also includes pieces of a very light, vesicular material, grey in colour, which results from some kind of pyrotechnic activity, but not necessarily metalworking. Overall, the quantities of slag are insufficient to postulate on-site metalworking, or any other industrial activity. Fragments came from topsoil, and from both prehistoric and medieval features. Other finds comprise very small quantities of ceramic building material (medieval and post-medieval brick and roof tile); fired clay (probably structural, mainly from Iron Age contexts, including one fragment with wattle impressions); glass (post-medieval vessel and window); stone (fragment of saddle quern from Iron Age enclosure ditch 792, chalk spindlewhorl from enclosure ditch 784); and metalwork (including copper alloy buttons, buckle and pin, lead sheet fragment, iron nails, key and scythe, all medieval or later).

19

ENVIRONMENTAL Charred plant remains Chris J. Stevens Introduction Thirty-four bulk samples were taken and processed for charred plant remains and charcoal. On the basis of the assessment, which on the whole indicated few cereal remains or charred plant remains other than charcoal, six samples were selected for full analysis. The selected samples came from three middle to late Iron Age features (enclosure ditches 784 and 792; pit 612 from group 800), two early/mid-Saxon features (fire pit 433; pit 525), and one Late Saxon/medieval (ditch 782). Methods of analysis The samples were processed using standard flotation methods with the flot collected on a 0.5mm mesh. For the six samples selected for analysis all identifiable charred plant macrofossils were extracted from the flots, together with the 2mm and 1mm residues. Identification was undertaken using stereo incident light microscope at magnifications of up to x40, following the nomenclature of Stace (1997) for wild species and the traditional nomenclature as provided by Zohary and Hopf (2000, tables 3 and 5) for cereals. The results are presented in Table 4. In the case of the 0.5mm to 1mm fraction from ditch 784 the sample was extremely rich in small weed seeds and for this reason only one-third of the sample was examined. The results were then multiplied by three to provide estimates for the flot as a whole. Results Iron Age The Iron Age samples were dominated by remains of both emmer wheat (Triticum dicoccum) and spelt (Triticum spelta) wheat. These were particularly prevalent in the samples from pit 612 (group 800) and ditch 784. Generally spelt was better represented than emmer in the samples, with glume bases outnumbering grains.

20

Barley (Hordeum vulgare) grains were also present, although far outnumbered by those of hulled wheats. Only in one case, in the grain from pit 612, was hulled barley identified. Grains of free-threshing wheat were also present, although only in low numbers. No other crop remains were present in these samples, although remains of potential wild foods including hazelnut (Corylus avellana) shell fragments and a seed of bramble (Rubus sp.) were recorded. These samples also produced thorns of sloe/hawthorn (Prunus spinosa/Crataegus monogyna). The samples from pit 612 and enclosure ditch 784 contained several tubers of pignut (Conopodium majus). However, given that these same samples also yielded tubers/rhizomes of onion couch grass (Arrhenatherum elatius) and monocot (sedge, grass or rush) rhizomes, rootlets and stems, this might suggest they are more likely to have been associated with turf material rather than deliberately collected. The samples included reasonably large numbers of seeds of wild species, in particular from ditch 784. The latter sample was dominated by smaller seeds of orache (Atriplex sp.), fat-hen (Chenopodium album) and clover (Trifolium sp.). Other seeds present in all three samples included those potentially associated with grassland but equally possibly from arable fields, including dock (Rumex sp.), meadow grass/cats tails (Poa/Phleum sp.), brome grass (Bromus sp.) and oats (Avena sp.). The richer two samples also produced seeds of a large range of species associated both with grassland and arable fields; these included vetch/wild pea (Vicia Lathyrus sp.), medick (Medicago lupulina), selfheal (Prunella vulgaris), ribwort plantain (Plantago lanceolata), red bartsia (Odontites vernus) and perennial rye grass (Lolium perence). Species more closely associated with arable fields in these two samples included scentless mayweed (Tripleurospermum inodorum) and stitchwort (Stellaria sp.). Saxon and medieval Of the three Saxon and medieval samples, that from the early/mid-Saxon fire-pit 433 produced very little in the way of plant macros, being dominated by wood charcoal. However, those present did include several thorns of sloe/hawthorn (Prunus spinosa/Crataegus monogyna), along with three intact fruits of sloe (Prunus spinosa),

21

of which two had the stones clearly visible, and a well preserved stone. The only other remains in this sample were a single grain of rye and a seed of vetch/wild pea (Vicia
Lathyrus sp.).

The remaining two samples had a greater quantity of cereal remains, although only that from late Saxon/medieval ditch 782 might be deemed rich in cereal remains. Unlike the Iron Age samples, these samples produced no remains that might be associated with turves. Early/mid-Saxon pit 525 produced a few grains of barley (Hordeum vulgare sl) and a single rachis fragment and several grains of free-threshing wheat (Triticum aestivum/turgidum type). A possible grain of hulled wheat was recovered along with two extremely badly preserved glume bases. This sample also has a single fragment of hazelnut (Corylus avellana) shell. Seeds of wild species were relatively few but comprised a similar range of species to those seen in the Iron Age samples. These included fig-leaved goosefoot (Chenopodium ficifolium), dock (Rumex sp.), clover (Trifolium sp.), meadow grass/cats tails (Poa/Phleum sp.), oats (Avena sp.) and stinking mayweed (Anthemis cotula). The richer sample from late Saxon/medieval ditch 782 had a large number of grains of free-threshing wheat (Triticum aestivum/turgidum type), with a similar number of rachis fragments of free-threshing wheat. In most cases these rachises were not identifiable, but in a few cases they could be indentified as from hexaploid freethreshing wheat, (Triticum aestivum sl). Weed seeds mainly comprised both small and large seeded species. In the former group were those of fat-hen (Chenopodium album), orache (Atriplex sp.), dock (Rumex sp.) and stinking mayweed (Anthemis cotula). In the latter group were seeds of vetch/wild pea (Vicia Lathyrus sp.), cleavers (Galium aparine/tricornutum) and oats (Avena sp.).

22

Discussion Iron Age The cultivation of both spelt and emmer within other parts of East Anglia during the Iron Age is well testified (Stevens 2009a and b; Murphy 1997). The high numbers of glume bases to grain indicates that all three samples result from the charring of dehusking waste from the processing of spelt and/or emmer stored in the spikelet. Other aspects of crop-husbandry might be difficult to elicit from the samples. Given the presence of both tubers of pignut (Conopodium majus), and swollen tubers of onion couch grass (Arrhenatherum elatius), along with numerous rhizomes and stems of monocotyledonous plants, there is a question of whether the seeds of wild species recovered from the Iron Age samples arrived with turf material or were harvested with cereal remains. The range of species represented is for the most part indicative of neutral to calcareous, long, seldom-grazed grassland. Along with pignut and onion couch grass such species potentially include clover (Trifolium sp.), black medick (Medicago lupulina), vetch/wild pea (Vicia Lathyrus sp.), red bartsia (Odontites vernus), annual meadow grass/cats tails (Poa/Phleum sp.), and potentially indicative of slightly more disturbed grassland, ribwort plantain (Plantago lanceolata), dock (Rumex sp.), brome grass (Bromus sp.), cinquefoils (Potentilla sp.), self-heal (Prunella vulgaris), knotgrass (Polygonum aviculare) and oats (Avena sp.). While pignut is an unlikely component of arable fields, all the remainder, particularly the latter group, could equally represent arable weed seeds. Amongst those species whose seeds are found in the samples, but are less likely to be associated with turves from long-grassland are many seeds of orache (Atriplex sp.), fat-hen (Chenopodium album), chickweed (Stellaria media), black bindweed (Fallopia convolvulus) and scentless mayweed (Tripleurospermum inodorum). At the very least, most of these species would appear to be reflective of generally drier, lighter, neutral to calcareous soils, and hence suggestive that such local soils were under cultivation. Wetland species are rare but blinks (Montia fontana subsp.

23

chondrosperma) was recovered from ditch 792 and sedge (Carex sp.) from the other two samples, although in low quantities. High numbers of monocot stems and rhizomes along with tubers of onion couch grass (Arrhenatherum elatius) are common in parts of northern England where they were probably associated with the burning of turves (Hall 2003; Hall and Huntley 2007). The other alternative is that they derive from ground-breaking and vegetation clearance, which was then burnt perhaps during the formation of a fire-break (cf. Stevens 2008). Saxon and medieval The charcoal from fire pit 433 contained large quantities of sloe or blackthorn (Prunus spinosa) (Barnett, below), and it is probable that many of thorns derive from the burning of blackthorn or hawthorn (Crataegus monogyna) twigs and branches. The presence of several whole fruits of sloe (Prunus spinosa), while possibly representative of the collection of such fruits for consumption, are also likely to have come in with scrub/hedge material collected as fuel, still attached to the plant. Fruits of sloe gradually dry on the plant and become quickly wrinkled, and this would imply that the wood with the berries is likely to have been collected between September and November. More importantly, it suggests that at least some of the wood, if not a large part of it, was burned green with the berries still attached. The cereals from ditch 782 and pit 525 are relatively typical of Saxon and medieval assemblages with a high prevalence of free-threshing wheat, along with smaller amounts of barley and rye (Stevens 2009c). The high numbers of rachis fragments is of some significance, particularly since they are far more readily destroyed during charring than grains (Boardman and Jones 1990). The ratio of rachis fragments to grain is 1:26; as such, a higher ratio of rachis fragments, as seen here, is characteristic of threshing waste (van der Veen 1992, 82), in particular rachises removed by raking or coarse sieving. Both small and large seeds are relatively prolific in the samples, although it might be noted that around half of the small weed seeds are of stinking mayweed (Anthemis

24

cotula), which has a tendency to remain in the heads and therefore is often removed with coarse sieving waste. The samples are likely to come from the waste generated from processing sheaves for clean grain. It is quite possible that as such the crops were stored as sheaves after harvesting in summer. However, whether this was a common practice, or one that was only conducted in years when poor weather conditions restricted the processing of crops following harvest, is difficult to gauge from two samples. The presence of stinking mayweed (Anthemis cotula) can be associated with the cultivation of heavy clay soils. This species occurs in earlier periods, but only upon Romanised settlements, and its widespread occurrence in the Saxon period as seen here is quite probably related to the introduction of heavy mouldboard ploughs (Stevens with Robinson 2004). Wood Charcoal Catherine Barnett Introduction Five samples were analysed for charcoal, all from the early/mid-Saxon fire pits. All wood charcoal >2mm was separated from the processed flots and the residue scanned or extracted as appropriate. The samples proved rich and so were sub-sampled, with a number of fragments felt to be representative of the sample as a whole identified, normally 100 fragments. The fragments were prepared for identification according to the standard methodology of Leney and Casteel (1975, see also Gale and Cutler 2000). Each was fractured with a razor blade so that three planes could be seen: transverse section (TS), radial longitudinal section (RL) and tangential longitudinal section (TL). The pieces were mounted on a glass microscope slide using modelling clay, blown to remove charcoal dust and examined under bi-focal epi-illuminated microscopy at magnifications of x50, x100 and x400 using a Kyowa ME-LUX2 microscope. Identification was undertaken according to the anatomical characteristics described by Schweingruber (1990) and Butterfield and Meylan (1980) to the highest taxonomic level possible, usually that of genus, with nomenclature according to Stace (1997).

25

A list of taxa by period is given in Table 5. Individual taxa were quantified (mature and twig separated), and the results tabulated (Table 6). Results and Discussion As shown in Table 5, a minimum of eight woody species were represented overall. The charcoal from each of these five early/mid-Saxon features proved relatively similar, supporting the assumption that they are related in function and chronology. Three of the five were heavily dominated by blackthorn (Prunus spinosa) at 6086%, with common Pomoideae (usually indentifiable further as hawthorn, Crataegus type) at 1014%. Pomoideae, however, formed the dominant type in pit 435 at 70%, with ash (Fraxinus excelsior) in pit 724 at 46%, though blackthorn was still common in the latter. Oak (Quercus sp) occurred in three contexts, while field maple (Acer campestre), buckthorn (Rhamnus cathartica) and elder (Sambucus nigra) each occurred in one context. All the taxa found are relatively common deciduous types and most are tolerant of a variety of free-draining soils. However, in combination the taxa found are strongly suggestive of hedging or open scrub. In addition, for all of the species represented, there proved to be an overwhelming predominance of juvenile wood, twigwood and young roundwood, usually 310 years old when cut. Although there was clearly a substantial presence of roundwood, this varied in age, diameter and species, so there is no clear indication of management of woody resources, for example by coppice rotation. Instead, exploitation of substantial quantities of immature scrub and/ or hedges for fuel is indicated. Whether this was purposeful or as a result of extensive earlier deforestation in the area which limited availability of larger trees and shrubs or access to remaining resources is unclear. However, it is apparent that a very open landscape occurred locally during the Saxon period. Despite the fact that young narrow pieces dominated, a high temperature of burn was achieved with the selected fuel, as attested to by the glassy, vitrified appearance of many of the pieces. According to the experimental work of Prior and Alvin (1983), temperatures >800C are necessary to achieve this. It might be suggested that, given the young wood and types used, careful management of the fire, possibly within a

26

restricted area or structure would be needed to reach such temperatures, and casual domestic fires are not indicated. DISCUSSION Neolithic and Bronze Age There is very limited evidence for pre-Iron Age activity on or around the site, in the form of residual Neolithic and Bronze Age finds; the assemblage is largely indicative of transient occupation. Land on the Oxford Clays was only sporadically exploited prior to the Iron Age, and there are very few known pre-Iron Age sites in the area (e.g. Hall 1992, 96; Ellis et al. 1998, 107). Late Iron Age (1st century BC1st century AD) This period saw more permanent occupation of the site. The creation of a trackway implies the use of the area for pastoral agriculture and the movement of animals. From this period onwards the demarcation of boundaries and the construction of enclosures increasingly express the sense of tenure from which a sustained settlement emerges. Only one possible roundhouse structure was found in the excavated area, and the evidence is not very convincing. However, the occurrence of wattle impressions on a fragment of fired clay is indicative of structural remains, and more structures may have existed, possibly beyond the western and southern limits of the site. A variety of subsistence activities (cereal grain-processing, textile-working, iron smithing) are represented in the range and quantity of occupation debris recovered, implying a mixed farming economy. Emmer and spelt wheat were being cultivated in an increasingly deforested area; sheep/goat were the predominant animal species exploited, together with cattle in smaller numbers. There is slight evidence for the exploitation of the fenland in the form of blinks and sedge, possibly introduced to the settlement with fodder. The pottery suggests that while the settlement may have Middle Iron Age origins, the focus of activity is likely to lie within the range of c. 100 BCAD 20, and that the site was probably abandoned by the time of the conquest. Whether occupation spanned the whole of this date range, or was relatively shortlived, is not possible to determine from the ceramic evidence, although the intercutting of ditches implies more than one phase of enclosure.

27

The Spaldwick settlement is one of a number of such sites excavated within the county, and the environmental and economic background to Iron Age settlement in the region, including the fen-edge, has been increasingly well researched. The main strands of evidence indicate expansion of settlement on to the claylands during this period, possibly reflecting the exploitation of marginal lands as a result of population growth, or due to climate change leading to increased flooding of lower-lying areas (Haselgrove et al. 2001, 29; Abrams and Ingham 2008, 30). There is a general increase in activity along the fen-edge; there are more settlements, and more artefactual evidence, including an increase in traded products. There is little evidence of the latter from Spaldwick a quernstone and a chalk spindlewhorl. The closest parallels for the pottery appear to lie with sites in the west of the county, rather than those to the east or south, but there is no evidence to indicate that they were anything other than locally produced. The size of the Spaldwick enclosures is broadly comparable to those at Scotland Farm on the A428 in the south of the county (Abrams and Ingham 2008, fig. 2.1), but the agglomeration of enclosures seems to have grown more organically than at the latter site, while the construction of sub-rectangular rather than curving enclosure ditches can be paralleled also at Knapwell Plantation and Little Common Farm at Cambourne, also on the A428 (Wright et al. 2009, fig. 15, fig. 22), and these also lacked internal buildings, as at Spaldwick, though only small areas of the four enclosures were exposed on the latter site. The abandonment of the settlement may have coincided with the arrival of the Roman legions (cf Ellis et al. 1998, 107) or may be linked to increased flooding which led to many settlements in the Ouse/Nene region being abandoned by the 1st century AD (Dawson 2000, 11114). That some kind of presence persisted in the area, however, is implied by the sporadic occurrence of Romano-British pottery. Early/mid-Saxon (5th8th centuries) Activity on the site during this period is somewhat enigmatic. There are no definite structures, but this is in line with the general pattern across the region, where settlement was concentrated in the valleys and is rare on the clay uplands, which were

28

probably used primarily for pasture (Wright 2009, 1156). The presence of a probable waterhole certainly fits the latter supposition The nine sub-rectangular fire pits excavated are of uncertain function, and the dating evidence from them is very slight (four sherds of pottery). The charcoal found within them indicates the exploitation of large quantities of immature scrub or hedges for fuel. The relatively small size of the pits, the type of fuel burnt in them, and the presence of variable quantities of lightly burnt flint nodules renders their initial interpretation as charcoal-burning pits highly unlikely. The pitsteads used in the Saxon charcoal-burning industry were usually circular, often banked around the edges, and at least 4 metres across (see Steane 1985, 222 for medieval examples), and would have used very large quantities of mature wood (up to several tons in each episode of burning) which, given the evidence for the deforestation of the surrounding landscape, would not have been easily available. A link with metalworking is unsupported by any other firm evidence for on-site metalworking in the form of slag, but some kind of craft or industrial function still seems to be the best interpretation. The relatively large quantity of animal bone recovered from the pits should be noted (more than was recovered from the whole of the late Saxon/medieval phase), as should the concentration of this material, together with fired clay and nodules of lightly burnt flint, in one of the pits; it may be that these pits were used for cooking. Any connection with the later village development seems fortuitous. Late Saxon/early medieval (9th11th/12th centuries) The evidence from the late Saxon to early medieval period is of particular interest in throwing some light, albeit somewhat dim, on the origins of the modern village. Comparable evidence is scarce within the county. This period saw the creation of formalised land divisions or tofts which extended south from what is now Thrapston Road, a medieval route which perhaps originated as a hollow-way through the settlement. These tofts appear to have contained rectangular post-built timber structures with fences sub-dividing the yard areas, and perhaps represent the beginnings of the village. Evidence of the cultivation of the clay soils of the area, probably related to the introduction of heavy mouldboard ploughs, comes in the form of seeds of stinking mayweed recovered from one of the boundary ditches.

29

The area at this time formed part of Mercia, with Cambridge as a possible frontier burh, but it is assumed that the arrival of the Danes in 875, and a settlement at Cambridge from c. 889, must have stimulated some reorganisation of land ownership, from multiple estates with dependent tenures, to the Danish system of free tenancies (Ellis et al. 2001, 103). In this context it is perhaps significant that the early village was named Danesfield. After the reconquest of the Outer Danelaw in 9167, serfdom and the manorial system were reintroduced. It is more likely that the laying out of tofts in Spaldwick is linked to the latter phase, given the pottery evidence, which includes St Neots and Stamford wares and other 10th12th century wares from the boundary ditches, although the rectangular buildings produced only residual early/mid-Saxon sherds. The village may, however, have slightly later origins. A possible parallel can be seen at Bassingbourn in the south of the county, where there is a suggestion of an earlier (early/mid-Saxon) origin to the village, with subsequent changes in layout taking place in the 10th/11th century and possibly linked to land ownership changes after the Norman conquest (Ellis et al. 2001, 123). In terms of pottery, the late Saxon and early medieval village shows less reliance on local supplies, with regional wares (from St Neots and Stamford) appearing from the 10th century; this may reflect the position of the settlement on what may have been even at that time a major route between Cambridge, Huntingdon and Thrapston. There is little else, however, to illustrate the material culture of the settlement. In contrast to the early/mid-Saxon period, cattle were predominant rather than sheep/goat (although quantities are very small), while weed seeds suggest that the heavy clay soils of the area were being increasingly cultivated; crops include wheat, barley and rye. Acknowledgements Wessex Archaeology would like to thank Steve Daniels of SD Construction and Development Ltd for his assistance with the project, and also Dan McConnell, Historic Environment Officer for Cambridgeshire County Council, for his guidance and support. The fieldwork was directed by Susan Clelland, assisted by Ross Lefort, Virginia Meszaros, Jeff Muir, Virginia Vargo and Marius Wisniewski. The fieldwork was

30

managed by Brendon Wilkins, and the post-excavation analysis and publication by Andy Crockett. The illustrations were drawn by Elizabeth James. The archive is currently stored at the offices of Wessex Archaeology under the project code 75070, but in due course will be deposited with the Cambridge Archaeological Store under the accession code ECB3445.

31

Bibliography Abrams, J & Ingham, D 2008 Farming on the Edge: archaeological evidence from the clay uplands to the west of Cambridge. Albion Archaeology/East Anglian Archaeology Report 123, Bedford: Albion Archaeology Boardman, S & Jones, G 1990 Experiments on the effects of charring on cereal plant components. Journal of Archaeological Science 17: 111 Butterfield, BG & Meylan, BA 1980 Three-Dimensional Structure of Wood: an Ultrastructural Approach. London and New York: Chapman and Hall CCCAFU 1996 Land on and behind Thrapston Road, Spaldwick, Cambridgeshire. Cambridge County Council Archaeology Field Unit report 127/1996 Crabtree, P 2010 Agricultural innovation and socio-economic change in early medieval Europe: evidence from Britain and France. World Archaeology 42(1): 122 136 Dawson, M 2000 The Iron Age and Romano-British period: a landscape in transition. In M Dawson (ed), Prehistoric, Roman, and Post-Roman Landscapes of the Great Ouse Valley. Council for British Archaeology Research Report 119, York: 10730 Ellis, P, Hughes, G, Leach, P, Mould, C & Sterenberg, J 1998 Excavations alongside Roman Ermine Street, Cambridgeshire, 1996: the archaeology of the A1(M) Alconbury to Peterborough Road Scheme. Birmingham University Field Archaeology Unit Monograph Series 1/British Archaeological Report British Series 276, Oxford: Archaeopress Ellis, P, Coates, G, Cuttler, R & Mould, C 2001 Four Sites in Cambridgeshire: excavations at Pode Hole Farm, Paston, Longstanton and Bassingbourn, 19967. Birmingham University Field Archaeology Unit Monograph Series 4/British Archaeological Reports British Series 322, Oxford: Archaeopress

32

Gale, R & Cutler, D 2000 Plants in Archaeology. Westbury and Royal Botanic Gardens, Kew Hall, D 1992 The South-western Cambridgeshire Fenlands. East Anglian Archaeology/Fenland Project No. 6, Cambridge Hall, A 2003 Recognition and characterisation of turves in archaeological occupation deposits by means of macrofossil plant remains. Centre for Archaeology Report 16/2003, Portsmouth: English Heritage Hall, AR & Huntley, JP 2007 A review of the evidence for macrofossil plant remains from archaeological deposits in Northern England. Research Department Report Series 872007, Portsmouth: English Heritage Hambleton, E 1999 Animal husbandry regimes in Iron Age Britain: a comparative study of faunal assemblages from British archaeological sites. British Archaeological Report British Series 282, Oxford: Archaeopress Hancocks, A 2003 The Iron Age pottery from Little Paxton near Bedford. In A Gibson (ed.), Prehistoric Pottery: People, Pattern and Purpose. Prehistoric Ceramics Research Group Occasional Publication 4/British Archaeological Report International Series 1156, Oxford: Archaeopress, 10356 Hancocks, A, Evans, A & Woodward, A 1998 The prehistoric and Roman pottery. In Ellis et al. 1998, 3479 Haselgrove, C, Armit, I, Champion, T, Creighton, J, Gwilt, A, Hill, J D, Hunter, F & Woodward, A 2001 Understanding the British Iron Age: an Agenda for Action. Salisbury: Prehistoric Society/Wessex Archaeology Leivers, M 2009 Prehistoric pottery. In Wright et al. 2009, CD-Rom Leney, L & Casteel, RW 1975 Simplified procedure for examining charcoal specimens for identification. Journal of Archaeological Science 2: 53159 MacGregor, A 1985 Bone, Antler Ivory and Horn: the technology of skeletal materials since the Roman period. London and Sydney: Croom Helm

33

Mepham, L 2004 Saxon pottery. In CJ Ellis, A Prehistoric Ritual Complex at Eynesbury, Cambridgeshire: Excavation of a multi-period site in the Great Ouse Valley, 20002001. East Anglian Archaeology Occasional Paper 17/Wessex Archaeology, Salisbury: Wessex Archaeology, 537 Morris, EL 1994 The Analysis of pottery. Wessex Archaeology Guideline 4, Salisbury: Wessex Archaeology MPRG 2001 Minimum Standards for the Processing, Recording, Analysis and Publication of Post-Roman Ceramics. Medieval Pottery Research Group Occasional Paper 2 Murray, J 1998 Land off Ferriman Road, Spaldwick, Cambridgeshire: an archaeological evaluation. Hertford Archaeological Trust Report 298 Murphy PJ 1997 Environment and economy. In S Bryant, Iron Age, in J Glazebrook (ed.), Research and Archaeology: a Framework for the Eastern Counties, 1. Resource Assessment. East Anglian Archaeology Occasional Paper 3, 301 PCRG. 2010 The Study of Later Prehistoric Pottery: general policies and guidelines for analysis and publication. Prehistoric Ceramics Research Group Occasional Paper 1/2 (3rd ed) Percival, S 2008 Pottery (Bronze Age and Iron Age), in Abrams and Ingham 2008: CD-Rom, Appendix 5 Prior, J & Alvin, KL 1983 Structural changes on charring woods of Dictostachys and Salix from Southern Africa. International Association of Wood Anatomists Bulletin 4(4): 197206 Schweingruber, FH 1990 Microscopic Wood Anatomy. Birmensdorf: Swiss Federal Institute for Forest, Snow and Landscape Research (3rd ed.) Seager Smith, R 2009 Saxon pottery. In Wright et al. 2009, CD-Rom Stace, C 1997 New Flora of the British Isles. Cambridge: Cambridge University Press (2nd ed.)

34

Steane, JM 1985 The Archaeology of Medieval England and Wales. London and Sydney: Croom Helm Stevens CJ 2008 Cereal agriculture and cremation activities. In MJ Allen, M Leivers & CJ Ellis, Neolithic causewayed enclosures and later prehistoric farming: duality, imposition and the role of predecessors at Kingsborough, Isle of Sheppey, Kent, UK. Proceedings of the Prehistoric Society 74: 2969. Stevens, CJ 2009a The local and wider Iron Age landscape. In Wright et al. 2009, 6972 Stevens, CJ 2009b The Iron Age agricultural economy. In Wright et al. 2009, 7884 Stevens, CJ 2009c The Anglo-Saxon agricultural economy. In Wright et al. 2009, 116 Stevens, CJ with Robinson, M 2004 Production and consumption: plant cultivation. In G Hey, Yarnton: Saxon and Medieval Settlement and Landscape. Thames Valley Landscape Monograph, Oxford: Oxford Archaeology, 812 Van der Veen, M 1992 Crop Husbandry Regimes: An archaeobotanical study of farming in northern England 1000 BC AD 500. Sheffield Archaeological Monograph 3/JR Collis Publications, Sheffield: University of Sheffield Vince, A 1997 The petrology of some shell-tempered pottery from Cambridgeshire (Site LHP97). Alan Vince Archaeological Consultancy (AVAC) Report 1997/004 Vince, A 2006 Characterisation studies of Iron Age pottery from the A428 Caxton to Hardwick Improvement Scheme, Cambridgeshire (CH1131). Alan Vince. Archaeological Consultancy (AVAC) Report 2006/74 Vince, A 2007a Characterisation of Iron Age and Roman shell-filled pottery from Earith, Cambridgeshire. Alan Vince Archaeological Consultancy (AVAC) Report 2007/34

35

Vince, A 2007b Petrographic analysis of Saxon and medieval pottery from Stratton, Bedfordshire, and Cambridgeshire: early to mid Anglo-Saxon wares. Alan Vince Archaeological Consultancy (AVAC) Report 2007/119 Vince, A 2007c Characterisation studies of Cambridgeshire Anglo-Saxon and medieval pottery: Colne Ware. Alan Vince Archaeological Consultancy (AVAC) Report 2007/112 Wessex Archaeology 2002a Land adjacent to 33 Thrapston Road, Spaldwick, Cambridgeshire: desk-based assessment. Wessex Archaeology Report 51568E Wessex Archaeology 2002b Land adjacent to 33 Thrapston Road, Spaldwick, Cambridgeshire: archaeological evaluation. Wessex Archaeology Report 51568.02 Wright, J 2009 Anglo-Saxon settlement, in Wright et al. 2009, 1056 Wright, J, Leivers, M, Seager Smith, R & Stevens, CJ 2009 Cambourne New Settlement: Iron Age and Romano-British settlement on the clay uplands of west Cambridgeshire. Wessex Archaeology Report 23, Salisbury: Wessex Archaeology Zohary, D, & Hopf, M 2000 Domestication of plants in the Old World: the origin and spread of cultivated plants in West Asia, Europe, and the Nile Valley. Oxford: Clarendon Press (3rd ed.)

36

List of Figures Figure 1: Site location Figure 2: Overall view of site during excavation, view from the south-west Figure 3: Iron Age features phase plan Figure 4: Sections of selected Iron Age features (enclosure ditches 783, 792 and 798) Figure 5: Early/mid-Saxon features phase plan Figure 6: Sections of selected early/mid-Saxon features (waterhole 663, fire pit 724) Figure 7: Fire pit 433 after partial excavation Figure 8: Late Saxon/medieval and post-medieval features phase plan Figure 9: Sections through late Saxon/medieval boundary ditch 782 Figure 10: Prehistoric pottery

List of Tables Table 1: Pottery fabric totals and summary descriptions Table 2: Iron Age vessel forms by fabric Table 3: Animal bone: number of identified specimens present (or NISP) by period Table 4: Charred plant remains Table 5: Charcoal species list Table 6: Charcoal from early/mid-Saxon fire pits

37

Table 1: Pottery fabric totals and summary descriptions

Fabric Code GR1* GR2 GR3* QU1 SH1* SH2 SH3 SH4* PREHISTORIC Beaker Peterborough ware Abundant, subangular grog <2.5mm; sparse fossil shell fragments <1mm; in fine silty clay matrix Common, poorly sorted, subangular grog <3mm; lumpy texture; no shell visible in hand specimen As GR1, but grog less common and fossil shell more common Rare subrounded quartz <0.5mm; in fine clay matrix. One small body sherd only Abundant, poorly sorted fossil shell fragments <4.5mm; rare quartz; rare burnt plant matter; in fine clay matrix As SH1, but shell finer and better sorted Sparse fossil shell fragments <1mm; sparse to moderate, well sorted, subangular grog <1mm; in fine clay matrix Sparse fossil shell and calcite fragments <2mm; sparse subangular grog <1mm; rare quartz; in fine clay matrix sub-total prehistoric SAXON Quartz inclusions <0.5mm, derived from arenitic sandstone in slightly calcareous clay matrix; crushed sandstone possibly added as temper Sandstone inclusions (quartz, polycrystalline quartz, and rock fragments) <1mm and fossil shell <1mm in fine, noncalareous clay matrix; both possibly added as temper Coarse fabric: rounded quartz and polycrystalline quartz <2mm, some possibly added as temper, in fine, noncalcareous clay matrix; sparse carbonate inclusions Fine fabric: finer variant of Q400; inclusions <0.25mm Quartz-rich sandstone inclusions <2mm (quartz and polycrystalline quartz), poorly sorted, probably naturally occurring; possible plant temper, in non-calcareous clay matrix Abundant quartz, polycrystalline quartz and sandstone <2mm (derived from coarse arkosic sandstone); sparse carbonate inclusions, whether temper or naturally occurring uncertain; in non-calcareous clay matrix Sand-sized inclusions of quartz and polycrystalline quartz <0.5mm (well sorted, probably added as temper) in silty non-calcareous clay matrix; sparse chert and sandstone inclusions Organic-tempered fabric: moderate, fairly well sorted organic inclusions <3mm; rare quartz sub-total Saxon OVERALL TOTAL No. sherds 1 1 42 3 2 1 73 48 25 75 271 Weight (g) 3 7 245 63 14 1 1025 565 265 854 3042

QU400*

28

184

QU401*

44

222

QU402* QU403

13 13

92 58

QU404*

16

QU405*

11

QU406* VE400

3 5 109 380

37 32 652 3694

* indicates fabrics samples for petrographic analysis

38

Table 2: Iron Age vessel forms by fabric

Vessel Form Type 1: Convex vessels with beaded rims Type 2: Convex vessels with plain or proto-bead rims Type 3: Convex vessels with everted rims Type 4: Rounded, open vessels with sharply everted rims Type 5: Shouldered vessels with plain, upright rims Type 6: Open vessels with plain rims Type 7: Everted rim, necked form TOTALS GR1 1 GR2 1 SH1 3 SH2 2 SH3 SH4 Total 7 3 3 2

2 2 1

1 1 1

1 1 1 2 1 4 1 7 2

2 1

1 5

3 21

39

Table 3: Animal bone: number of identified specimens present (or NISP) by period
pig cattle horse dog cat rat sheep/goat red deer domestic fowl Unidentifiable Total 3 109 482 391 138 21 47 1191 113 44 8 16 360 150 29

Period/Species 3 12 28 54 10 6 2 112 168 46 9 3 8 12 1 1 831 7 4 1 2 31 2 13 26 7 1 94 36 13 2 3 5 278 86 21 3 3 3 4 1 1 332 11 1 3 1 1 80

Iron Age phase 1

Iron Age phase 2

Iron Age phase 3

Early/mid-Saxon

Late Saxon/medieval

Modern

Undated

Total

40

Table 4: Charred plant remains

IA 792 ditch 515 21 8 110
25 20 65 3 30 10

Phase Feature Number Feature type Context Sample Size/Litres Flot Roots 2 2 5 33 4 1 2 cf.1 2 1 1 1 45 1 1 1h 1 2 6 111 2 63 2 1 14 48 est. 160 3 est. 16 28 3 3 1 cf.1 2 7 2 1 127 105 3 1 13 2 -

IA 612 pit 609 37 8 100

IA 642 ditch 678 24 17 180

E/M Saxon 433 fire pit 434 5 17 16501 LSax/med 782 ditch 748 27 2 20

E/M Saxon 525 pit 524 18 18 100

Cereal Hordeum vulgare sl (grain) H. vulgaresl (rachis frags) Triticum sp. (grains) Triticum dicoccum/spelta (grain) T. dicoccum/spelta (glume bases) T. dicoccum/spelta (spikelet forks) Triticum dicoccum(glume base) T. dicoccum(spikelet fork) Triticum spelta (glume bases) Triticum spelta (spikelet fork) Triticum aestivum/turgidum (grains) T. aestivum/turgidum (rachis frags)

T. cf. aestivum (hexaploid rachis frags) Secale cereale (grain) Secale cereale (rachis frags) Cereal indet. (grains) Cereal indet. (est. whole grains from frags.) Cereal culm nodes Species Corylus avellana Ranunculus subg. Ranunculus (arb)

Common Name barley barley wheat emmer/spelt wheat emmer/spelt wheat emmer/spelt wheat emmer wheat emmer wheat spelt wheat spelt wheat bread/rivet wheat bread/rivet wheat hexaploid rachis fragment rye rye cereal cereal cereal Common Name hazelnut buttercup

41

Fumaria sp. Chenopodium ficifolium Chenopodium album Atriplex sp. Stellaria media 3 1 2 1 1 8 3 1 1 1 est. 9 1 1 5 20 2 2 est. 4 2 est. 31 est. 90 4 20 1 11 est. 3 est. 21 est. 9 3/1 ++ 1 3 7 4 8 8 est. 28 est. 1 2 8 2 36 5 1 1 1 1 est. 3 -

2 1

2 3 3 1

1 est. 60 est. 220 est. 3

1 1 -

19 36 -

Stellaria palustris/graminea Montia fontana subsp. chondrosperma

Persicaria maculosa/lapathifolium Polygonum aviculare Fallopia convolvulus Rumex sp. Brassica sp. Rubus sp.

Rubus/Rosa type sp. (thorn) Potentilla sp. Prunus spinosa (fruit/stones) P. spinosa/ Crataegus monogyna (thorns) Medicago lupilina Trifolium sp. Galium aparine/tricornutum Vicia./Lathyrus sp. cf. Lathyrus nissola Conopodium majus (tubers) Prunella vulgaris Plantago lanceolata Odontities vernus Sherardia arvensis Sambucus nigra

Ateraceae indet. >2.5mm Tripleurospermum inodorum

fumitory fig-leaved goosefoot fathen orache chickweed marsh/lesser stitchwort blinks redshank/pale persicaria knotgrass black bindweed dock black mustard etc. bramble bramble/rose type thorns cinquefoil/tormentil sloe sloe/hawthorn thorns black medick clover cleavers vetch/pea grass-pea pignut selfheal ribwort plantain red bartsia field madder elder Indet. Daisy/thistle type scentless mayweed

42

Anthemis cotula Monocot Root stems/culms/internodes Lemna sp. Carex sp. lenticular Carex sp. (trigonous) Poaceae large (culm node) Lolium perenne L. 2 1 2 3 1 2 43 12 2 20 30 10 10 4 1 9 est. 12 2 33 -

2 -

12 1 2 2

est. 170 1m 3 est. 6

1 -

24 1 4

Poa/Phleum sp.

Arrhenathermum elatius var. bulbosum Avena sp. (grain) Avena L./Bromus sp. Bromus sp. Seed indet. small indet root/tuber Charred rodent dropping

stinking chamomile grass/sedge stems duckweeds sedge flat seed sedge (trigonous) grass culm node rye grass meadow grass/cats'tails false oat-grass (tuber) oat grain oat/brome brome

43

Table 5: Charcoal species list

Species Acer campestre Corylus avellana Fraxinus excelsior Pomoideae Crataegus type Prunus sp. Prunus spinosa Quercus sp. Rhamnus cathartica Sambucus nigra Common Name Field Maple Hazel Ash Pomaceous fruits e.g. apple, whitebeam, hawthorn Hawthorn Cherry type, includes bird cherry, wild cherry, blackthorn Blackthorn Oak Buckthorn Elder

44

Table 6: Charcoal from early/mid-Saxon fire pits

Charcoal 4/2 mm Prunus spinosa Quercus sp. Acer t Corylus avellana Fraxinus excelsior Pomoideae Pomoideae, Crataegus Prunus sp. Rhamnus catartica Comments Sambucus i Unidentified 5 twd 3 7, 1 twd Total no frags used 100 100 100 100 100

Feature Context

Sample

Size Litres

Flot Size ml

700/500 ml

2 rwd

14*

83 rwd**

433

434 950/425 ml 6 3 70 1

17

16501

12

435 150/150 ml 10 rwd * 10 rwd* * -

436

11

1700

60 rwd***

2 rwd

450

452 160/50 ml -

16

425

724

46 rwd *

16

18 rwd*

20 rwd* *

723 240/100 ml

32

250

13, 1 twd

80 rwd*, 6 twd

729

730

V large sample, big pieces, occ. fissured and vitrified. *0.8 2.2cm 513 yrs, **12.2cm 522 yrs Rich sample, large pieces (esp Pomoideae), occ. vitrified Moderate sample, several warped and vitrified but friable and fragmentary, all young rwd 11.5cm diameter, *35 yrs, **5 yrs, 310 yrs Moderate sample but large pieces (esp Pomoideae), occ. fissured and vitrified. *13 cm 38 yrs, **1cm 45 yrs Moderate sample but inc v large pieces. *inc 1.22.2 cm 8 11 yrs

Key:

34 2 400 rwd roundwood; twd twigwood

45

512500

ne Ne R.

Peterborough

Ol

dN

en

eR

513000

273000

The Site
1996 Evaluation
A5

Cambridgeshire
Huntingdon
R. Ouse

Spaldwick

Boundary of 'Danesfield' late Saxon village

04

St James' Church

Godmanchester Little Paxton Eynesbury

EET STR INE ERM

Cambourne Cambridge

Spaldwick
m

R. Ca

272000

Bassingbourne

1998 Evaluation

Earthworks
Digital Map Data (2004) XYZ Digital Map Company

The Site

272900

512800

Site outline Excavation area Evaluation trench Archaeological feature

0
Contains Ordnance Survey data Crown copyright and database right 2012.

50m

Tree-throw hole

Site location

Figure 1

Figure 2

Enclosure 784

Gully 787

440

Gully 758 Boundary ditch 783 Enclosure 794 762 Gully 789 801 549 Gully 774 693 695 701

612

Gully 790 Ditch 788

Rubbish pits 800 Gully 796

802

Enclosure 792 Enclosure 798 Boundary ditch 628

Excavation area Iron Age: Phase 1 Iron Age: Phase 2 Iron Age: Phase 3 Later phase/undated

10

20m

Section line

Iron Age phase plan

Figure 3

23.97m OD

400

736

735 734

733

Boundary ditch 783

732

NW 577 575

SE 24.39m OD

574

Enclosure 792
0 1m

Enclosure 798

Sections of selected Iron Age features

Figure 4

Pit 450

Pit 455

Pit 752 Pit 433 Pit 729 Pit 525 Waterhole 663 Pit 724 Pit 435 Pit 725

Pit 715 Pit 705

Excavation area Early-mid-Saxon 0 10 Other phase/undated 20m Section line

Early/mid-Saxon phase plan

Figure 5

SE 674 676 673

NW
23.65m OD

791
670 675

672

669 671

668 666

667

Flint Charcoal flecks Ceramics

Waterhole 663

NNE 720 721 722 723

SSW
23.80m OD

Burnt flint Low charcoal density Medium charcoal density

Fire pit 724

0 1m

High charcoal density

Sections of selected Saxon features

Figure 6

Figure 7

Pit 406 Surface 404 Wall 402

?Hollow way 804

Pit 448 776

Fence 786

Ditch 782

Ditch 780 Pit 490 Pit 543 Structure 803 Fence 786

Gully 793 Structure 799

Gully 539

Ditch 791

Excavation area Late Saxon-medieval Post-medieval 0 10 Other phase/undated 20m Section line

Late Saxon/medieval and post-medieval phase plan

Figure 8

S 584 586 583 588

23.17m OD

585

Ditch recut 422

582

Ditch recut 587

Ditch 782 0 1m Flint Charcoal flecks

Sections of selected medieval features

Figure 9

50

100mm

6 7 Figure 10