A R C H A E O L O G Y. N o. 3 NORTH WEST CAMBRIDGE (2) EXCAVATIONS Assessment Report Craig Cessford and Christopher Evans

Size: px
Start display at page:

Download "A R C H A E O L O G Y. N o. 3 NORTH WEST CAMBRIDGE (2) EXCAVATIONS Assessment Report Craig Cessford and Christopher Evans"

Transcription

1 NORTH WEST CAMBRIDGE A R C H A E O L O G Y EXCAVATIONS Assessment Report Craig Cessford and Christopher Evans N o. 3 (2)

2 NORTH WEST CAMBRIDGE ARCHAEOLOGY University of Cambridge Excavations - Romano-British - (NWC Report No. 3; Pt. 2) Craig Cessford & Christopher Evans With contributions by Katie Anderson, Grahame Appleby, Guy Bettley-Cooke, Lawrence Billington, Steve Boreham, Richard Darrah, Natasha Dodwell, Charles French, David Hall, Vicki Herring, Jacqui Hutton, Mark Knight, Rob Law, Quita Mould, Vida Rajkovača, Ian Riddler, Hayley Roberts, David Smith, John Talbot, Simon Timberlake, Anne de Vareilles and Nick Wells Illustrations by Vicki Herring and photography by Dave Webb CAMBRIDGE ARCHAEOLOGICAL UNIT UNIVERSITY OF CAMBRIDGE March 2014/Report No HER Event Number: ECB 4111

3 SECTION 3: THE ROMANO-BRITISH PERIOD The archaeology of the Romano-British period defined here as including the Latest Iron Age (50BC 50AD) consists of two Romano-British settlements (RB.1 2), some routeways (Ways 1 3) a double-ditched boundary and a few isolated features (Table 3.1; Fig. 3.01). The presentation of the Romano-British archaeology will be based upon these units. One of the Romano-British settlements (RB.2) is a complex entity which has been divided into four spatial units (RB.2A 2). The phasing of the Romano-British archaeology is based on a number of factors. These include stratigraphic sequence, spatial logic and dateable material culture. In general the stratigraphic resolution at the site was relatively poor and, although a significant number of stratigraphic relationships were identified, in approximately 40 50% of instances they were either unclear or non-existent. Spatial logic proved particularly useful as there were frequent instances where one feature or group of features clearly respects the location or alignment of another feature or group of features and must therefore post-date them. The most frequently present type of dateable Romano-British material culture is pottery, although a high proportion of the assemblage consists of coarsewares that can only be broadly dated to the Romano-British period. This has been supplemented by the rarer, but frequently more precisely dateable, coins. Other material culture that provides some dating evidence includes metalwork, vessel glass, leather and worked antler and bone. Although running the risk of creating circular arguments initial analysis indicated that certain types of material culture notably iron slag and stone roof tile are either entirely absent or extremely infrequent at the sites until around the mid-3rd century AD. It also appears that a distinctive dark coloured midden-like or extremely dirty soil layers do not appear until the 3rd century AD. The only significant potential dating evidence that is not available at this stage is dendrochronology, if successful this will more precisely date the construction of a few features but may have wider implications as well. No. % Way 1/ Way Double-ditch Routeways etc total RB RB.2A RB.2B RB.2C RB.2D RB.2 total Settlement total Isolated features Total 1183 Table 3.1: Feature breakdown for all Romano-British features by site. 105

4 Phasing After the Romano-British archaeology was divided into large-scale units such as sites, routeways etc. each was independently phased. Some of the entities such as routeways generally lack the necessary body of evidence to allow detailed phasing so this has not been done. The complexity of the archaeology, particularly of parts of Sites RB.2A D, and the organic nature of much of the development mean that phases are in part artificial analytical constructs. Phases are therefore defined as representing major transitions in the overall character of an area. Criteria for defining phases include: 1) A shift from using an area principally for burial purposes to occupation or vice versa. In some instances these may effectively represent a form of precursor or successor phase (see below) 2) The creation of a ditched enclosure system where one had not existed before 3) A major expansion or contraction of the area of a site, typically >50% 4) A significant shift in the alignment of a ditched enclosure system, typically >10 degrees 5) A change from a curvilinear to a rectilinear ditched enclosure system or vice versa. In the interests of comprehensiveness a number of additional phases have been created. These are principally: 1) Precursor phases where there is some extremely limited evidence for Romano- British activity in an area prior to the commencement of occupation. 2) Successor phases where there is evidence that a site was largely abandoned, but that a few essentially relict features continued in use. This relates principally to wells and the fact that banks which had not left any direct surviving archaeological traces may have continued to exist as significant topographic features after the associated ditches had entirely infilled. 3) Intermediary phases where there is evidence for a hiatus between significant earlier and later phase of Romano-British activity. As the phases cover relatively long periods some individual features can be more precisely dated within a phase. As this never exceeds 10% of the features assigned to a phase it was not deemed worth creating formal subphases, however this more precise dating has been taken into account where appropriate. These individual site based phases were then subsequently combined into an overall phasing system; this consists of three broad phases with eight subphases (Table 3.2; Fig. 3.02) and the site phases are broadly categorised (Table 3.3). 106

5 RB2D Later Quarries Early/Mid 1st Century Mid 1st - mid 2nd Century Mid 2nd - mid 3rd Century Mid 3rd - mid 4th Century RB2C Way 4 Double-ditched system RB2A RB2B Way 3 Way 1 0 metres 100 Way 2 RB1 Figure Phase plan of Romano-British archaeology, including selected evidence from evaluation

6 RB2D Double ditched system Double ditched system Way 4 RB2A RB2B Way 3? Way 3 RB2B Way 3 RB2B Way 1 Way 1 Way 2? Way 2 RB1 0 RB1 200 metres Early 1st century Mid 1st century Early 2nd century RB2D RB2C RB2A RB2A Way 4 RB2C RB2C RB2A Double ditched system Way 3 Way 3 Way 1 Way 3 Way 1? Way 2 Way 2? RB1 Late 2nd century Late 3rd century Figure Phased sequence of Romano-British archaeology 4th century

7 Overall phasing Date Way 1 Way 2 Way 3 Double-ditched boundary RB.1 RB.2A RB.2B RB.2C RB.2D 1.I 1.II 1.III 2.I 2.II 3.I Mid 1st BC mid 1st century AD Mid late 1st century AD Late 1st mid 2nd century AD Mid late 2nd century AD late 2nd early 3rd century AD Mid late 3rd century AD Prob. Way 1.1 Way 1.2 Way 1.2 Poss. Poss. Poss. Way 2.1 Way 2.2 Poss. Poss. Way 3.1 Way 3.1 Way 3.1 Way 3.2 Way 3.2 D- D.1 D- D.1 D- D.2 D- D.2 Est. (1.1) Cont. (1.1) Cont. (1.1) Relict well (1.2) Relict well (1.2) Burial only (2A.1) Est. (2A.2) Mod. (2A.3) Cont. (2A.3) Cont. (2A.3) Est. (2B.1) Cont. (2B.1) Mod. (2B.2) Cont (2B.2) Presettlement (2C.1) Est. (2C.2) Cont. (2C.2) Mod. (2C.3) Burial only (2D.1) Est. (2D.2) Mod. (2D.3) Cont. (2D.3) Relict well (2D.4) Early mid Way Mod. Cont. 3.II 4th century AD 3.2 (2A.4) (2C.3) Mid-4th century Way 3.III AD onwards 3.3. Table 3.2: Phasing of Romano-British sites etc, bold entries indicate settlement proper (Cont. continuity, Est. established, Mod. major modification/change to settlement, Poss. Possibly in existence, Prob. Probably in existence). Site Phase Date (AD unless otherwise stipulated) Description RB.1 1 mid 1st late 2nd century Initial ditched enclosure system RB.1 2 late 2nd late 3rd century Successor phase: relict well RB.2A 1 mid late 1st century Precursor phase: cemetery RB.2A 2 late 1st mid 2nd century Initial ditched enclosure system RB.2A 3 mid 2nd late 3rd century Modified and expanded ditched enclosure system RB.2A 4 late 3rd mid 4th century Contracted ditched enclosure system. Possibly successor phase. RB.2B 1 mid 1st century BC mid 1st century AD Initial ditched enclosure system RB.2B 2 mid 1st century late 2nd century Modified and expanded ditched enclosure system RB.2C 1 mid 1st mid 2nd century Precursor phase: limited presettlement activity RB.2C 2 mid 2nd mid 3rd century Initial ditched enclosure system RB.2C 3 mid 3rd mid 4th century Modified and contracted curvilinear ditched enclosure system RB.2D 1 early 1st mid 1st century Precursor phase: isolated small enclosures and cremations RB.2D 2 late 1st mid 2nd century Initial ditched enclosure system RB.2D 3 mid 2nd late 3rd century Modified and expanded ditched enclosure system RB.2D 4 late 3rd early/mid 4th century Successor phase: relict well Way 1 1 mid 1st century Ditched routeway Way 1 2 mid 1st late 2nd century Banked routeway Way 2 1 late 1st/early 2nd century mid 2nd century Ditched routeway Relict well (2D.4) 109

8 Way 2 2 mid late 2nd century AD Banked routeway Way 3 1 mid 1st century BC/early 1st Intermittently metalled and ditched century AD late 2nd century AD routeway Way 3 2 late 2nd century mid 4th century Intermittently metalled and ditched routeway Way 3 3 mid 4th century onwards Successor phase: routeway D-D 1 mid 1st mid 2nd century Double-ditched boundary D-D 2 mid 2nd early 3rd century Double-ditched boundary Table 3.3: Summary of phases of Romano-British sites, routeways, etc. Component Parts - Feature Types The Romano-British features can be broken down by type (Table 3.4). Ditches dominate the Romano-British archaeology at just over half the total number of the period s features, whilst this is largely genuine there is an element of this that is related to the multiple numbering of complex ditches where it was impossible to relate particular individual re-cuts between different slots. The other common elements are pits, which constitute around a quarter of the features and postholes which represent over an eighth. In the interests of avoiding unnecessary repetition the general form and range of Romano- British feature will be briefly described as a group. Individual features that, for whatever reason, merit individual description will be dealt with subsequently on a site-by-site basis. Feature type No. % Postholes Pits Pits (specialised) Ditches Ditches (ring) Ditches (wheel ruts) Wells Burials (cremation) Burials (inhumation) 16 (+2?) 1.5 Corn driers Beam-slots Metalling Tree throws/root boles Buried soils Hollows Later infilling Miscellaneous Total 1183 Table 3.4: Feature breakdown for all Romano-British features by feature type. Structures, plus Miscellaneous Postholes and Beam-slots Only seven Romano-British period structures were convincingly identified, although this admittedly excludes evidence for two structures associated with wells. Additionally there were a large number of postholes that could not be assigned to recognisable features; whilst some of these may be the elements of structures others may genuinely represent isolated individual postholes or other types of features such as fence lines. The structures consist of five beamslot and post-built structures, a single six-post structure and a corn-drier. 110

9 The lack of structures is undoubtedly in part due to truncation, as the pattern of ditched enclosures in some parts of the sites strongly suggests the existence of more structures which have not left any archaeological trace. There does, however, appear to be a genuine lack of major structures of any kind. If aisled buildings for example had been present then these should have been clearly identifiable. Although building stone, tile, tesserae etc. was recovered all this material appears to have been brought to the site for secondary re-use and there is no evidence that any of it was employed on site in a primary role. It appears that all the buildings were probably constructed of timber and part of one such timber was found reused RB Well 12 (F.1402). More generally, the timber from the wells suggests that the standard of woodworking at the site was simple, but competent and effective and that structural timbers were probably exclusively axe cleft rather than sawn. The postholes from the site were typically circular in plan, although many were quite irregular, and were m in diameter with a mean diameter of c. 0.5m. The surviving depths ranged between m with a median surviving depth of c. 0.20m. This suggests that the postholes were originally c m deep, with a median depth of c m. Few of the postholes had identifiable post-pipes so nothing meaningful can be inferred about the actual posts. The beam-slots were m long excluding a single atypical example that is 16.9m long m wide and had surviving depths of m. Wells The wells are amongst the most distinctive features of the Romano-British period and - as water is such a fundamental human requirement also amongst the most informative in terms of understanding the site. Additionally, they provide most of the contexts with waterlogged survival, with preservation of wood/timber and leather plus environmental remains in the form of waterlogged seeds, pollen and insects. The waterlogged survival is variable; apparently because of a strongly fluctuating water table so only some wells have waterlogged survival. One wattle-lined and two box- lined wells were conclusively identified in several instances through the survival of actual linings and it is likely that all the wells on site were wattle or box lined, although the possibility of cask linings cannot be excluded. Based upon their form and profile it is possible to determine with some certainty if the examples where no lining survives were wattle-lined (eight) or box-lined (12 13). The wattle-lined wells typically have a funnel-shaped profile with a large upper cone leading to a vertically sided circular shaft in the base. The boxlined wells are large and deep steep-sided circular or oval features, whose form and size is largely determined by the necessity of creating a substantial working space around the shaft to allow the timber elements to be nailed together. There is sometimes, but not always, a slightly deeper square or rectangular vertically sided shaft in the base of the main pit where the lining was located. The precise number of wells is uncertain, as in a number of instances there are several phases of well in identical or near identical locations. In these cases the 111

10 features have been treated as a single well, but the multiple phases are distinguished. The provision of wells varied considerably, some settlements only ever had one well at any point in time whilst others had three or possibly four suggesting different levels of requirement. In a number of instances the fill sequence indicates that the well shafts were not deliberately backfilled and instead over time the surrounding fill and adjacent natural fissured and collapsed/slumped into the shafts. RB.1 has two probably wattle-lined wells in succession. RB.2A has three probable and one definite box-lined wells, with possibly up to three wells in operation at the same time. RB.2B has no associated wells; however, two probable box-lined wells located in Way 3 may well be linked to this area. These two wells probably operated in succession. RB.2C has six or seven probably box-lined wells and three probably and one definitely wattle-lined wells. There were certainly three and probably four wells sometimes in operation at the same time. RB.2D has one box-lined well and three probably wattle lined wells, with two and conceivably three wells sometimes in operation at the same time. Burials The Romano-British period burials at the site include both cremations and inhumations, in general terms cremation is typically superseded by inhumation as the preferred burial rite from the later 2nd century onwards becoming dominant by the mid-3rd century. The majority of burials occur in small cemeteries; these are generally not specifically enclosed but are sometimes defined by pre-existing boundaries. There are also some isolated burials. Unsurprisingly, given the degree of truncation there is no evidence where the actual cremation rite took place. In most instances there was no discernible cut for the cremations, indicating that the holes were not significantly larger than the vessels placed in them. Most cremations were within a coarseware jar, the cremated bone rarely if ever filled even half of this vessel and inverted bowls were sometimes placed over the cremated remains as lids. Accessory vessels were often placed next to the primary containers. The inhumations were typically of extended supine individuals, those for adults occurred in rectangular vertically sided flat-bottomed cuts m long by m wide with a surviving depth of m. The presence of nails indicates that a significant proportion of the burials were in coffins and hobnails indicating the present of footwear were also relatively common. Ditches Ditches represent the most prominent feature type during the Romano-British period and come in a wide variety of forms and sizes, indeed in several instances both the form and size of a ditch changes markedly along its length. Although they vary considerably in width ( m, mean width c. 1.05m) and surviving depth ( m, mean width c. 0.4m) there appears to be a continuum of size with no evidence for distinct size based groups. 112

11 The ditches relate primarily to enclosure systems, although many of them will have fulfilled a dual role related to water management. In most instances the evidence indicates that ditches were cur only once and then proceeded to silt up naturally. It is, however, possible that in some instances ditches were cleaned/re-cut so thoroughly than all evidence of earlier phases was removed. Additionally shallower re-cuts may have left no archaeological evidence. Despite these provisos the overall impression is that most ditches were relatively short-lived features, lasting for a few decades at most. In most instances the infilling material appears to represent natural silting up of the ditch, with occasional collapses of the edges of ditches and slumping of adjacent banks. There are a few instances where ditches appear to have been utilised for refuse disposal, although this is relatively infrequent. It is, however, probable that the associated banks - perhaps with small remnant hollows over the ditches outlived the main ditches by considerable periods. There is a single instance of a long-lived ditch alignment with evidence of up to eight or nine re-cuts. There are also a number of linear features that have been broadly characterised as ditches but that may well represent wheel ruts. Pits A large number of Romano-British period pits were identified (295). They come in a wide variety of forms and sizes, with most appearing to be hastily dug displaying little if any care in their creation and were rapidly backfilled. The bulk of these features may well have been quarry pits for gravel, sand and clay; although unlike at other sites, there is little specific evidence, such as the chasing of specific seams of material. Additionally, a number of morphologically similar pits cannot have been quarries as they were dug through earlier features. Some of the pits occur in distinct clusters of 8 15 inter-cutting pits, these all appear to represent relatively short-lived episodes indicating repeated activity in a favoured locale. There is no evidence that pits were dug specifically for refuse disposal or as cess-pits, although a few were used as such on an ad hoc basis. As a result the function of most pits remains obscure with quarry pits representing an unsatisfactory default option. A small minority of pits demonstrate a substantially greater degree of regularity in their form and in some instances these may have been lined in some manner. These appear to have been longer lived and probably fulfilled a range of particular individual functions. In addition there were a small number of pits with surviving linings. Tree-throws, Tree Boles and Hollows A range of rather irregular features have been interpreted as tree throws, tree boles and naturally formed hollows dating to the Romano-British period. None of these features appear to be particularly significant but they do shed some light on the landscape of the period. 113

12 19th Century Quarries Roman Roads and Trackways Double-ditched system Way 3 Way 1 0 metres 100 Way 2 Figure Plan highlighting Romano-British routeways and double-ditch boundary, with photograph of Way 3 and plan of Milneweye

13 Later Infilling Deposits In some instances it appears that after use and an initial period of infilling features such as ditches and wells stabilised their profiles and remained as relict hollows for a considerable period after they went out of use. These then appear to have often been deliberately infilled, often with midden-like deposits that cover a number of earlier features. It is unclear if this infilling relates principally to waste disposal or if it was intended to create a more level landscape, perhaps to facilitate agriculture. Routeways and Boundaries Three routeways were identified (Ways 1 3) plus a fourth putative example (Way 4) and some form of double-ditched boundary (Fig. 3.03). Way 1 - Way 1, located toward the eastern end of the investigated area, ran in a northeast-southwest direction from the gravel ridge down onto the clay. It was traced for a distance of c. 130m across the excavated area and was also identified in trenches to the southeast, giving a total length of c. 210m. The relatively straight edges of the road were delineated by well defined continuous ditches typically 11 28m apart; these ditches were c. 1.8 wide and 0.8m deep. There was no convincing evidence of any re-cuts associated with these ditches the fill sequences were generally simple. No traces of any metalling associated with Way 1 were identified. Way 1 passed through RB.1, stratigraphically it is clear that the enclosure ditches associated with RB.1 cut through the already backfilled ditches of Way 1, effectively narrowing it by c. 3.5m. The idea that the ditches of Way 1 were backfilled by this time is supported by the lack of material from these ditches, even when adjacent to RB.1. As RB.1 was established in the mid-1st century this indicates that Way 1 was also established in the mid-1st century and was extremely brief lived, at least in its ditched phase. It seems probable that Way 1 continued in use after this time, presumably defined by a pair of low banks rather than ditches. The only pottery associated with Way 1 dates to AD and AD, however this material derived from where Way 1 met Way 2 (F.2584) and it appears that the junction was defined differently from the rest of the ditches. Way 1 must have existed until at least the end of the main phase of occupation at Site RB.1 in the late-2nd century AD, whilst it may well have continued after this date there is no compelling evidence that it did. Phase Way 1.1: ditched routeway (mid-1st century AD) Phase Way 1.2: banked routeway (mid-1st to late-2nd century AD) Ditches: 2537, , 2750, 2752, 2759, 2584, 2760 Pit: 2590 Way 2 - Way 2 joins to Way 1 and runs perpendicular to it towards the southeast. Only a very short length of c. 8m of Way 2 could be traced before it ran into an area of intensive Post-Medieval/Modern quarrying. It is therefore impossible to describe Way 2 in detail, although the short exposures of ditch revealed were broadly similar to those of Way 1. As already mentioned, pottery from this junction dates to AD and AD. It is probable 115

14 that Way 2 led to the Romano-British settlement at Vicar s Farm, which was established c AD. If Way 2 is solely related to the Vicar s Farm settlement, which is unproven, then it may well have been created after Way 1. As the Vicar s Farm settlement outlived RB.1 then Way 2 probably continued until at least the end of the main phase of occupation at Site RB.1 in the late-2nd century AD. As with Way 1, the lack of re-cuts etc. suggests that an initial ditched phase was followed by a phase where the routeway was delineated solely by banks. Phase Way 2.1: ditched routeway (late-1st century or early-2nd century AD to mid-2nd century AD) Phase Way 2.2: banked routeway (mid/late-2nd century AD) Ditches: F.2584 Way 3 - Way 3 was a sinuous but broadly southeast to northwest aligned ditched routeway, whose line was intermittently traced across several excavated area for a distance of c. 480m. In general, Way 3 appears to meander probably following the topography of the ridge located around 100m or so back from the actual crest. Way 3 was in existence by the start of the Romano-British period and may well have substantially earlier origins. As Way 3 runs immediately to north of a Prehistoric ring-ditch (Ring-ditch 4) and slightly changes alignment at this point it is possible that a low central mound was still extant and influenced the line of the routeway. Indeed there are suggestions of two distinct shifts in the alignment of the routeway, perhaps indicating that it ran between a series of minor topographic points. To the southeast Way 3 presumably joined to Way 1 beyond the limit of excavation, potentially creating a crossroads as it presumably continued beyond this point. To the northwest Way 3 ran to the southwest of RB.2A and the northeast of RB.2B. It is likely that it continued beyond this point to the northwest into an area of intense Post-Medieval and Modern quarrying. Additionally, it probably met with Way 4, a putative routeway running to the northeast along the site of RB.2A. Way 3 was composed of a series of discontinuous and extremely irregular ditches that were typically c. 8m apart and up to 2.0m wide. In most instances it appears that these ditches are composed of elongated quarry pits that were dug along the sides of the routeway. There is no overall pattern to these and they were potentially dug in an ad hoc manner to provide metalling for the routeway in locations where the ground conditions had deteriorated. There is also a suggestion that some of these features were may have been deliberately linked together to improve drainage locally. Up to four phases of re-cuts of features were discernible in places along the route of Way 3 indicating repeated quarrying. In most areas no surface associated with the routeway survived. In several locations there were indications that the initial surface consisted simply of removing the topsoil to create a moderately firm and relatively flat surface. Gravel was then subsequently added locally as a metalling where required, although only one convincing area of this survived (F.2668). There is no evidence that Way 3 was a focus for activity, the only features associated with it are what appear to be wheel ruts (F.1721, 1781) at 116

15 one point suggesting that carts/wagons went from Way 3 to the adjacent fields at this point. Two wells (RB Wells 17 18) were located within the line of Way 3, at the eastern end of settlements RB.2A B (Table 3.5). No trace of any well lining survived, when allowance is made for the likely extent of the actual shafts of these wells the line of Way 3 would have been somewhat impinged upon but not seriously compromised as a routeway. Given the absence of wells in RB.2B and the earliest phases of RB.2A (Phases RB.2A.1 2) it is probable that the wells associated with Way 3 supplied these with water until the mid-2nd century. The location of the wells also suggests that they may also have helped supply the requirements of the users of Way 3 and agricultural needs relating to nearby fields to the east. RB Well Feature Phases Lining Construction date Backfilling date Box? AD AD Box? Unk. 150AD+ Table 3.5: Wells from Way 3. Little material culture was associated with Way 3 and the distribution of the material suggests that it derived principally from nearby settlements. This material is of limited value as dating evidence and any attempt to devise an overall phasing system would be misguided. Despite these provisos there is relatively strong evidence to suggest that some of the features associated with Way 3 were dug as early as the 1st century AD and that some were still open into the 4th century, including the deposition of a coin issued in AD. As was noted during the evaluation (Evans & Newman 2010, 36 37), parts of Way 3 correspond almost exactly with the documented route of Milnewaye (or Mill Way), a small Medieval road that was in existence by c and that led from the rear of Howes Crofts to the gravel pits in Grithow Field (Hall & Ravensdale 1976, 36). The evidence from the Romano-British settlements suggests that when these went out of use rather than being simply abandoned and the areas passed directly into agricultural usage. It is likely that the area continued to be farmed through the Early and Middle Saxon periods and presumably Way 3 remained as some form of routeway. The extensive survival of ridge and furrow plus headlands and banks in the area until the 20th century, which have been plotted from aerial photographic evidence (Palmer in Redfearn 2001), indicates that classic Medieval open-fieldsystems under arable cultivation as a series large, hedge-less open fields farmed in strips were laid out c (Oosthuizen 2006), although locally there is evidence for 8th 9th-century intensively cultivated proto-open fieldsystems (Oosthuizen 2005; 2006). This layout preserved the still extant Way 3, in effect fossilising it. Some supposedly St. Neots-type pottery associated with Way 3 recovered during the evaluation phase had been re-classified as Roman shelly ware. This means that no post-roman material can now be associated with Way 3. Although the nature of Way 3 makes it impossible to phase entirely convincingly it can be sub-divided to an extent: Phase Way 3.1: routeway with some metalling and flanking pits/ditches with wells located on routeway (mid-1st century BC/early-1st century AD to late-2nd century AD) 117

16 Phase Way 3.2: routeway with some metalling and flanking pits/ditches (late-2nd century to mid-4th century AD) Phase Way 3.3: continued existence of routeway after abandonment of settlement RB.2 (mid-4th century onwards) The ditches, many of which are likely to be elongated pits, are extremely variable but up to 3.8m wide with a maximum surviving depth of 0.7m: F.1176, , 1588, 1599, 1620, 1643, 1652, , 1671, 1687, 1691, , 1709, , 1726, 1736, 1760, 1763, 1771, 1792, 1797, 1812, , , , 2136, , , Features that are definitely pits rather than ditches: 1629, 1651, 1665, 1676, 1686, 1735, 1759, 1761, 1764, 1838, 1842, 1852, , Wheel ruts: F.1721, Metalling: F A few postholes were present: 1725, 1762, RB Well 17 (F.1821): a broadly circular feature 4.0m in diameter, with a maximum surviving depth of 1.1m. No trace of the lining survived, but this well was probably box-lined. Pottery which probably relates to the construction of the well dates to AD, whilst pottery from the backfilling dates to AD and AD. RB Well 18 (F.1841): a broadly circular feature 4.45m in diameter, with a maximum investigated surviving depth of 0.5m. No trace of the lining survived, but this well was probably box-lined. Pottery from the backfilling well can only be broadly dated to AD. Way 4 - The existence of Way 4 is speculative. Its southeastern side is defined by the enclosure ditches of the northwestern side of RB.2A (Phase RB.2A.2 3), whilst if the other side of it was originally ditched the evidence from this has been obliterated Post-Medieval and Modern quarrying. Given that in Phase RB.2A.3 the principal focus of this site (Enc. 2A.7) appears to have faced onto the putative Way 4 its existence is probable, but cannot be definitively identified. Due to its nature no features are assigned to Way 4. Double-Ditched Boundary - The double-ditched boundary runs northeast to southwest from RB.2C on the ridge down onto the clay. It is defined by a pair of straight parallel ditches set 31m apart that were 0.8m wide with a maximum surviving depth of 0.3m with no trace of any re-cuts. It was traced for a distance of 60m across the excavated area and the southern of the ditches was also identified in two trenches to the southwest one from the evaluation phase and one from the excavation phase giving a distance of 120m. Athough after this point it could not be identified in further trenches, it is likely to have continued as a boundary beyond this defined either by ditches that were too shallow too survive or in another archaeologically invisible manner. It is also worth noting that the southern ditch appeared to continue further than the northern ditch, although this may simply represent differential truncation. The double-ditched boundary was established before the main occupation of RB.2C (Phase RB.2C.2) in the mid-2nd century AD. When this took place the northeastern end of the double-ditched boundary went out of use, pottery 118

17 associated with this stretch (F.523) dated to AD. Pottery associated with the length to the southwest which continued in use contained some pottery of the same date (50 100AD) plus later material of AD indicating that it continued in use until the 3rd century AD. Although densities are not high the distribution of pre 150AD pottery in the area that became RB.2C does hint at some level of activity at the northeastern end of the double-ditched boundary. The lack of re-cuts and the absence of post-250ad coins and pottery from the area of the double-ditched boundary, whilst not conclusive, do suggest that it did not continue after the mid-3rd century AD. If the double-ditched boundary continued on the same alignment beyond the limits of where it could be traced then it would have passed immediately to the south of Site VI, where there is evidence for a Late Iron Age/Early Romano-British settlement and agricultural activity of the 2nd 3rd centuries AD. Phase D-D.1: double-ditched boundary (mid-1st to mid-2nd century AD) Phase D-D.2: double-ditched boundary continues to southwest of Site RB.2C (mid-2nd to early-3rd century AD) The ditches (F , 523, 606 plus 84 from the evaluation) show no signs of being re-cut and are up to 1.03m wide with a maximum surviving depth of 0.65m). They generally contained very little, material although there was some evidence for the deliberate dumping of discrete groups of animal bone (sf.49 and 50). Settlement RB.1 Settlement RB.1 is located on the southwest slope of the gravel ridge, lying at between m OD (Tables 3.6 9; Fig. 3.04). Geologically it sits at the junction between the gault clay proper and more mixed deposits, although little of the site is located on good quality gravels. RB.1 lies both sides of Way 1, whilst it is conceivable that the two sides of the routeway are separate entities it is more likely that they are elements of a single site. Way 1 was clearly in existence prior to the creation of RB.1; the site was therefore located on a routeway immediately adjacent to the flat claylands. The western, eastern and southern limits of RB.1 lay beyond the limits of investigation; additionally, a modern drainage ditch meant that a significant area of the settlement has not yet been investigated. Phase Length (m) Width (m) Extent (square m) Focus of occupation with wells etc Largely empty > Largely empty Largely empty, sub-divided with corn drier Total Table 3.6: RB.1 enclosures. Comments 119

18 C5 R-D3 Enc 1.3 Enc 1.2 Cremation Well 1 Well 2 Enc 1.1 Way 1 Way 2 Well 2 Str 1 Enc 1.4 Clay-lined pit RB1.1 RB metres Figure Site RB1: Phase plans for RB1.1-2

19 No. of enclosures Enclosed area (square m) Wells Burials Unk. 1 0 Table 3.7: Proxy measures for activity levels at RB.1. There are two phases of activity identifiable at RB.1: Phase RB.1.1: initial ditched enclosure system (mid-1st to late-2nd century) Phase RB.1.2: relict well continues in use (late-2nd to late-3rd century) Phase RB Initial Ditched Enclosure System (mid-1st to late-2nd century) RB.1 was established in the mid-1st century AD, the best dating evidence comes from a cremation cemetery (C5; Fig. 3.05) located just to the north of RB.1 and assumed to be associated with it. This has evidence for five burials of AD, with two dated more closely to 40 70AD. There is a small amount of other pottery of this date from the site, some of this appears to relate principally to Way 1 and a group from the extreme eastern end of the area investigated (F.2583) may conceivably relate to another settlement. The other location to produce early pottery was a well (RB Well 01), supporting the idea that RB.1 originated at this time. The area to the west of Way 1 consisted of three enclosures (Enc.1.1 3). The main focus of activity was Enc.1.1, which pre-dates the other enclosures, or at least was the first to be created. It also appears purely on the grounds of spatial logic that the creation of Enc.1.3 predates Enc.1.2. Enc.1.1 was sub-divided into at least three areas, with the northwestern sub-enclosure probably the location of a building that has left no surviving archaeological traces. Another subenclosure in the northeast corner may have contained some form of structure that has left no archaeological trace and the presence of some disarticulated human bone in its ditches may be significant. The main part of Enc.1.1 contained two successive wells (RB Wells 01 02). It is noteworthy that the eastern end of Enc.1.1 encroached upon the line of Way 1 to an extent. There was evidence for a narrow ditched routeway running along the southern side of Enc.1.1. To the north of Enc.1.1 there was another enclosure (Enc.1.2), which contained few traces of internal activity. To the west of Enc.1.1 there was a further enclosure (Enc.1.3). This also contained few traces of internal activity but was not fully investigated. These two enclosures (Enc.1.2 3) presumably relate to stock management and/or agricultural activities. In addition to the cremation cemetery to the north of the site that has already been mentioned there was an isolated cremation in a wooden box or casket (F.2606). 121

20 F.2702 F.2667 F.2665 F.2664 F.2663 F.2666 R-D3 0 metres 2 Archaeological feature Excavated slot Prehistoric feature Figure Site RB1: Plan of Cemetery 5

21 F metres 2 Figure Site RB1: Plan and photograph of Structure 1

22 To the east of Way 1 there was a single enclosure (Enc.1.4), although this was divided into at least three sub-enclosures. There were few internal features within these sub-enclosures, and these were largely confined to the westernmost sub-enclosure. The northwestern corner of this sub-enclosure had been further sub-divided and contained a structure identified as a corn drier (RB Str. 01; Fig. 3.07). The only other feature of note lay to the north of Enc.1.4; this was a clay-lined pit that presumably fulfilled a specialised function. The pottery evidence indicates that the main phase of occupation at RB.1 ended c AD, but cannot be dated more precisely. This means that Phase RB.1.1 lasted for c years, with little evidence for change during that period. Although burial ceased at C5 by c. 100AD and no replacement cemetery was identified as RB Well 02 was created after c. 150 AD and the corn-drier (RB Str. 01) was still in operation at this time RB.1 appears to have still been flourishing at this stage. This phase of RB.1 is best interpreted as a relatively simple rural farmstead. The end of Phase RB.1.1 appears to be associated with the deposition of a significant quantity of pottery, quern and animal bone in two ditches (F ) that represent the front of the subenclosure where any buildings were probably located. This presumably represents some form of clearance deposit. There is also a considerable quantity of quern in the construction deposits of RB Well 02, whilst it is probable that this well was constructed during Phase RB.1.1 the possibility that it relates to the transition to Phase RB.1.2 cannot be entirely dismissed. Enc.1.1- Broadly rectangular although its southeastern end is distorted due to the enclosure being on a different alignment to Way 1 with an internal extent of c m by 48m covering 3490 square metres. The main ditch for Enc.1.1 (F , 2516, 2519, , 2550, ) showed no sign of any re-cuts in most locations and was up to 2.7m wide, with a maximum surviving depth of 0.6m. On the southwestern side of Enc.1.1 there is apparently a c. 2.0m wide routeway of some type (F ); the nature of this is unclear due to the limited exposure achieved so far. There was evidence for at least two sub-enclosures. A discontinuous curving ditch (F ) in the northeast corner of Enc.1.1 enclosed an area of c. 10m by 9m. Although there were no features within this sub-enclosure it is likely that some form of structure existed there. The curving ditch of the sub-enclosure was 1.3m wide, with a maximum surviving depth of 0.63m and was unexceptional apart from the presence of some disarticulated human bone. The other sub-enclosure located in the northwestern corner of Enc.1.1 (F , 2522) was c. 30m square with ditches (F , 2522) up to 0.9m wide with a maximum surviving depth of 0.35m. There was a c. 7.5m wide entrance on the southeastern side of the sub-enclosure and a c. 4.5m wide entrance of the southwestern side. The fills of ditches F (also investigated in the evaluation as F.434) contained a considerable quantity of pottery, quern and animal bone although the latter was heavily decayed and poorly preserved, to the extent that most could not be recovered and existed only as stains. This material appears to represent a single rapid dumping event, with the greatest concentrations located either side of the entrance into the sub-enclosure. With the exception of sherds from a single amphorae plus some largely abraded material little of the material definitely dates to before 150AD. With the exception of a few probably intrusive sherds no material need be later than 200AD, indicating a date of c AD. Over 1,100 sherds of pottery weighing over 21kg were recovered, plus 13.3kg of quern with five different stone types represented. Although incomplete due to later truncation this represents one of the most significant Romano-British assemblages from the excavations. Within this sub-enclosure there were a few relatively nondescript features (F , 2534). In what appears to be the central area of Enc.1.1 there was a cluster of relatively nondescript inter-cutting pits (F , ). Located beside these was RB Well 01 (F.2756), a circular steep-sided flat-bottomed feature 2.9m in diameter with a maximum surviving depth of 0.57m. The original well was c. 2.0m in diameter; the additional width is due to post-use slumping and collapse of the surrounding natural. No trace of the well lining survived, but it 125

23 was probably wattle-lined. There were no finds from deposits that can be related to the construction of the well. The basal fill, which may be use related, contained pottery of AD and fragments of a nailed shoe bottom. The backfilling of the shaft contained pottery of AD, but this is probably residual. To the southwest of RB Well 01 was RB Well 02 (F.2768), a rather irregular roughly subrectangular feature 5.3m by 5.1m with irregular steep sides and a maximum surviving depth of 1.25m (Fig. 3.06). The cut for RB Well 02 splays noticeably outwards to the northwest; this appears to represent a crudely stepped ramp that formed the point of access to the well during the construction process. There was evidence for the deliberate spreading of layers of gravel over the exposed natural gault clay in this ramp area, providing a less treacherous footing. The profile of the base indicated that there was a slightly deeper c. 1.0m diameter circular shaft in the centre of the feature, suggesting that the well was wattle lined. Although the bulk of the fills relate to the construction of the well these are problematic to distinguish, as due to later collapse/slumping they could not be separated from the backfilling of the shaft. The bulk of the pottery (just over 100 sherds weighing 2.9kg) and quern (18.8kg, with four different stone types represented) relates to the initial construction of the well and includes some pottery that is no earlier than 150AD. There is no pottery that can be linked categorically to the use of the well, although some large unabraded sherds from two colour coated beakers of AD may have been deposited during the use-life of the well. The backfilling of the shaft contained a one-piece leather shoe for a child dated to c AD and pottery dated to and AD. Although no trace of the original well lining survived there was some waterlogged survival; the waterlogged plant remains are not numerous and are poorly preserved, but there was good pollen and insect preservation. Other features in Enc.1.1 included a posthole (F.2603) and several ditches (F.2604, 2755, ). Such short lengths of most of these ditches have been revealed so far that it is fruitless to speculate about them. One exception is F.2755; the alignment of this parallels those of Enc.1.3 indicating that it is probably linked to this. Enc Enc.1.2 is located to the northeast of Enc.1.1 and is broadly rectangular although with a distinct kink at its northwestern end and was up to c. 115m long and 28 37m wide covering 2810 square metres. Its southwestern boundary was created by the pre-existing northeastern boundary of Enc.1.1. Its other boundaries (F.2502, 2620, 2632, , 2659; pits F.2525, 2535) show no signs of any re-cuts and are up to 2.12m wide, with a maximum surviving depth of 0.69m. There is some form of sub-enclosure of unknown function at its southeastern end (F.2500, 2546). The only other features within Enc.1.2 are some nondescript postholes (3: F , 2515) and pits (8: F.2503, 2506, 2512, , , 2654) that lack direct dating evidence but are probably Roman. Enc Enc.1.3 is located to the west of Enc.1.1, albeit on a radically different alignment reflecting the curving topography of the slope, and it shares a short length of boundary with Enc.1.2. Only its eastern portion has been investigated and it is unclear how large it is, the portion revealed is c. 50m wide by >55m long covering 3440 square metres. There are indications from evaluation trenches that the enclosure may continue for some distance but this is uncertain. There is no evidence that its boundary ditches (F.2632, 2659, 2738) were ever re-cut; they are up to 1.2m wide with a maximum surviving depth of 0.6m. Ditch F.2755, located within Enc.1.1, shares an alignment with the ditches of Enc.1.3 and is probably related. There only features within Enc.1.3 were a nondescript pit (F.2626) and posthole (F.2650). Enc Enc.1.4, located to the east of Way 1, is probably broadly rectangular in form although its northwestern end is distorted due to it being on a different alignment to Way 1 but has not been revealed in its entirety. The investigated portion is c. 140m long by c. 30m wide, covering 4230 square metres. The boundary ditch of Enc.1.4 (F.2583, 2735, , 2801) shows evidence of at least one and probably two phases of re-cutting. The ditch is up to 2.0m wide with a maximum surviving depth of 0.85m. Some internal ditches (F.2586, 2728), that are up to 1.15m wide with a maximum surviving depth of 0.7m, serve to divide Enc.1.4 into three sub-enclosures, two of which are c. 41 and 57m long. 126

24 The northwestern corner of the westernmost sub-enclosure located closest to Way 1 was separated off by a curvilinear ditch (F.2638) that was up to 0.65m wide with a maximum surviving depth of 0.35m. This created a broadly sub-circular area c m in diameter, located within this was RB Str. 01 (F.2567; Fig. 3.07). This consists of a sunken T-shaped feature with ligatured arms 3.5m long by 2.85m wide. This structure appears to have acted as a flue, as traces of scorching were apparent in places. The likelihood is that there was an additional superstructure around this feature, but none of the elements of this were deep enough to leave archaeological traces. Based upon a range of parallels RB Str. 01 was probably a corn drier. There was a small quantity of charred remains from RB Str. 01, whilst not particularly plentiful when these are compared the general paucity of charred remains from RB.1 this indicates that these may well be meaningfully associated with the corn drier. The structure was deliberately backfilled with sterile clay, the apparent care taken over this suggests that the RB Str. 01 may have continued in use in some manner after this. RB Str. 01 itself produced no dateable material, while the associated gully contained pottery that is no earlier than 150AD. The only other features within the westernmost sub-enclosure of Enc.1.4 were a posthole (F.2588) and a pit (F.2589). In the central sub-enclosure there was a row of postholes (F , 2747) that probably represent a fence line. Lying just to the north of Enc.1.4 was a pit (F.2811), this was only partially located within the area of excavation but was rectangular in form with vertical sides, a flat base and traces of a clay lining. It was over 2.0m long by 1.77m long with a surviving depth of 0.4m. There is no sign that the clay lining had been heat affected and it is probable that this pit held liquid. This feature produced no dateable material but is probably Roman. Burials: - Cemetery C5 consists of a cluster of five urned cremations (F ; Table 3.8; Fig. 3.05) dating to the mid/late-1st century AD. These are presumed to relate to RB.1, c. 27m away, however their precise location probably owes more to Prehistoric ring-ditch 3, as they are located at one end of the its ditch and it presumably still existed as an identifiable relict landscape feature. The cremations were all relatively badly preserved due to later ploughing and no associated cuts could be identified. There was also one isolated cremation, which appears to have been interred within a wooden box or casket (F.2606) and was located c. 60m north of RB.1. Phase RB Relict Well Continues in Use (late-2nd to late-3rd century) Evidence for continued activity was limited to the central area Enc.1.1, although it is unclear whether after c AD any of the earlier enclosures survived as relict banks or not. The main feature that continued in use was a well (RB Well 02), which was presumably retained as a useful water supply in an agricultural landscape. Two features in the vicinity of the well (ditch F.2514, pit F.2561) contained pottery that is later than 200AD; it seems likely that this represents material that accumulated in the upper hollows of relict features rather than the features proper. The shaft of RB Well 01 contained some large unabraded sherds from two colour coated beakers of AD and a leather shoe of c AD. This indicates that the well went out of use in the late-3rd century, after existing as an isolated feature for c years. As the well appears to have been deliberately backfilled this indicates that the area was not abandoned at this point, but continued in agricultural usage. Some level of continued visitation is indicated by the recovery of a stray find of a coin minted in AD. 127

25 Feature Phase Cut Bone Wt (g) Cremation container Dish lid Secondary vessels Additional items Comments [8374] Not identified Not identified Not identified Not identified nails [3873] indicate some form of wooden box or casket [8743] sandy ware jar AD [8705] Coarse sandy greyware jar AD [8728] sandy ware jar 40 70AD [8713] Coarse sandy greyware jar AD Not [8742] Q identified jar 40 70AD Table 3.8: Cremations from RB.1. Unk. Unk. Unk. Unk. Unk. Unk. [8704.1] imitation terra nigra CAM12 platter AD No [8711] Coarse sandy greyware imitation terra nigra CAM12 platter 40 70AD [8704.2] buff sandy ware vessel AD [8729] Coarse sandy greyware beaker 40 70AD [8710] Buff sandy closed vessel AD Unk. Unk. Unk. Unk. Unk. Unk. Isolated burial, plough damaged Plough damaged Plough damaged Heavily plough damaged Plough damaged Plough damaged RB Well Feature Phases Lining Wattle? Construction date Mid 1st century Backfilling date Mid 2nd century Wattle? AD AD Table 3.9: Wells from RB.1. Comment Pollen, insects, leather shoe, pottery, quern Material Culture and Environmental Evidence In general only low densities of material culture were recovered from individual features at RB.1, producing low- to medium-sized overall assemblages. The exceptions to this are the substantial groups of material from ditches F and well F The pottery from the cremations is also inherently significant. The bulk of the environmental samples from RB.1 128

26 contained only low densities of charred material, samples from the well fills produced poorly preserved waterlogged plant remains. There was good survival of pollen and insects from one well. Summary The evidence from RB.1 indicates the existence of a relatively simple predominantly single-phase rural farming settlement plus associated cemetery, established in the mid-1st century AD ca 40 70AD on a routeway where it reached the flat claylands. These were probably largely utilised for arable production and RB.1 may have specialised in this form of agriculture given the discovery of a corn-drier and large quantities of quern. This occupation continued until c AD, at which point a significant amount of material was dumped and the only feature to continue was a well which went out of use c AD. Settlement RB.2 RB.2 is a rather sprawling entity; which can be split into four sub-sites (RB.2A D). The degree to which these four sub-sites functioned as independent entities is debateable; it seems likely that they are in fact interrelated elements of a single complex rather than an aggregated group of sites. The degree of inter-relationship is strongest between the pairings of RB.2A B and RB 2C-D. Overall, RB.2 is located on the western side of the gravel ridge, with its westernmost elements on a slope leading down to the southwest. Most of the area is relatively flat at m OD, but some of the enclosures run down the slope to 18.11m OD. It largely overlies gravels, although the westernmost portion extends onto the gault clay. The northwestern limit of the settlement has not yet been investigated (Site V) and dense Post- Medieval Modern quarrying to the north and northeast means that the limits of the RB.2 are unclear. As a whole, the site spans over 425m+ northwest to southeast and 180m+ southwest to northeast, covering square metres. RB.2 is partly divided by Way 3, with RB.2A located to the northeast and RB.2B to the southwest, there also appears to have been a routeway running along the northwestern side of RB.2A (Way 4). It seems likely that the routeway represented by Way 3 continued to the north-northwest running by Sites RB.2C D; however truncation means that direct evidence for this is lacking. A double-ditched feature interpreted as some form of landscape division overlapped with part of the site (RB.2C) and ran from there in a southwestwards direction. Whilst the spatial distinctions between the sub-sites are not absolute as some overlap is apparent at their extremities the distinctions appear to be largely valid. Each sub-site has been phased independently to provide greater rigour. Given the inherent stratigraphic and dating limitations absolute certainty is impossible; nonetheless it does appear that the transitions between phases on different sub-sites are often broadly contemporary. To 129

27 combine all the phases of all the elements of RB.2 into a single phasing system requires a minimum of six phases. This, however, partially masks a broader pattern whereby activity can be split into three stages; early, mid and late (Table 3.2). These can be broadly defined as a series of spatially distinct loci (early), which went through a process of expansion and agglutination (mid) and then contraction (late). Prior to the beginning of the Romano-British occupational sequence there is no evidence for activity in the immediate environs since the Early Iron Age ( BC), indicating that during the Middle Iron Age ( BC) and Late Iron Age (100 50AD) the area was utilised solely for agriculture. The Romano-British occupational sequence began in the Latest Iron Age (50BC 50AD), with the earliest activity at RB.2B c. 50BC 50AD when a settlement was established (Phase RB.2B.1). Some pottery of c. 50BC 50AD was recovered from parts of the site, but not in large enough quantities in any location to indicate other areas of activity. By the mid-1st century AD (c AD) this settlement was joined by evidence for cremation burials taking place in two other areas (Phase RB.2A.1 and RB.2D.1). This was also the point where a settlement was established c. 400m to the southeast (RB.1). In the mid-1st century (probably c AD) the enclosed area at RB.2B expanded, but this was apparently no longer a focus for occupation and instead seems to simply be a series of agricultural enclosures (Phase RB.2B.1). There is evidence for a series of enclosures and some occupation at RB.2A (Phase RB.2A.2) and it appears likely that the focus of occupation shifted from one area to the other. At the same time a series of enclosures with evidence for occupation was established at RB.2D (Phase RB.2D.2). In the mid-2nd century (c AD) there were major modifications to the enclosure systems and occupation foci at Sites RB.2A (Phase RB.2A.3) and RB.2D (Phase RB.2D.3), with RB.2B continuing unchanged. The area between RB.2A and RB.2D (RB.2C) was also enclosed with evidence for both occupation and agricultural enclosures (Phase RB.2C.2). This led to the creation of a nearly continuous L-shaped settlement area with flanking enclosures. Soon after this activity ceased at RB.2B (c AD), although as this area had ceased to be an occupational foci some time earlier this change may not have been particularly significant. This pattern then continued until the mid-late-3rd century AD when the areas of activity contracted, in contrast to the mid-2nd century transition this appears to be more prolonged process. Activity largely ceased at RB.2D in c AD with the exception of a single relict well (Phase RB.2D.4) and the occupied area at RB.2C contracted markedly and became noticeably different in character in c AD (Phase RB.2C.3). At a slightly later date c AD the area of activity at RB.2A contracted (Phase RB.2A.4), and it appears that it may now have effectively acted as an adjunct to RB.2C. Activity at RB.2A and RB.2C appears to have ended at broadly the same time (c and c AD respectively) with the relict well at RB.2D going out of use before this c AD. Although archaeologically discernible activity ceases in the mid-4th century AD it appears that the area was not 130

28 entirely abandoned, but rather passed directly into agricultural usage with evidence for the continuity of Way 3. Settlement RB.2A RB.2A is the most northeasterly of the four spatial elements that constitute RB.2 (Tables ; Fig. 3.08). It is located to the northeast of Way 3 and to the southeast of the putative Way 4. The main investigated area covered up to c. 115m northwest to southeast by c. 115m southwest to northeast covering 9700 square metres; however, the settlement continued northwards beyond the limit of excavation for a distance of at least another c. 50m into an area heavily disturbed by Post-Medieval/Modern quarrying, covering an additional 4900 square metres. Our understanding of this northern part of the settlement derives solely from trenches excavated during both the evaluation and excavation phases (Tr. 7, , 128, ). There is some Prehistoric activity in this area (PS3), that all substantially predates the Romano-British period and it is likely that by the time RB.2A originated no visible traces of prehistoric activity remained. The settlement was entirely located upon gravel geology, although this is locally quite variable, and the site is relatively flat lying at m OD. That said, it is possible that during the Romano- British period there were minor undulations that have been removed by later activities. Phase Length (m) Width (m) Extent (square m) Enc.2A.1 2A Largely empty Enc.2A.2 2A c Largely empty Enc.2A.3 2A.3 53 c Largely empty Enc.2A.4 2A Largely empty Enc.2A.5 2A Largely empty, inhumation F.1935 Enc.2A.6 2A Largely empty Enc.2A.7 2A.3 16 Focus for occupation with RB Wells 13 16, partly lined pit F.1997 Enc.2A.8 2A.3 Unk. Unk. Unk. Largely un-investigated, probably largely empty Table 3.10: RB.2A enclosures. Comments The sequence at RB.2A can be divided into four main phases, although in many respects the last three phases represent a broad continuum: Phase RB.2A.1: cremations of C3 (mid/late-1st century) Phase RB.2A.2: initial ditched enclosure system (late-1st to mid-2nd century AD) Phase RB.2A.3: modified and expanded ditched enclosure system (mid-2nd to late-3rd century AD) Phase RB.2A.4: contracted ditched enclosure system (late-3rd to mid- 4th century AD 131

29 Way 4 C3 Enc 2A.2 Enc 2A.1 RB2A.1 RB2A.2 Way 3 Tile-lined pit Well 13 Way 4 Enc 2A.7 Enc 2A.8 Well 16 Well 15 Enc 2A.5 Str 3 Enc 2A.6 Way 4 Enc 2A.7 Well 15 Well 14 Enc 2A.4 Enc 2A.3 RB2A.3 RB2A.4 Way 3 Way 3 0 metres 100 Figure Site RB2A. Phase plans for RB2A.1-4

30 F centimetres F.2089 F.2088 F.2109 F.2093 F.2084 F.2085 F.910 (2609) F th Century Quarries F.2036 SF769 Cemetery 3 F.1958 (5876) (5874) (5826) (5872) (5870) Archaeological feature Excavated slot F.1931 Settlement 2D 0 50 F.1935 metres 0 1 Figure Site RB.2A Cemetery 3 0 metres 10 centimetres

31 No. of enclosures Enclosed area (square m) Wells Burials 2A.1 4 2A.2 2+ c A A Table 3.11: Proxy measures for activity levels at RB.2A. Phase RB.2A.1 - Cremations of C3 (mid/late-1st century) The earliest phase of activity at RB.2A consists of a cremation cemetery (C3; Fig. 3.09). The primary phase of C3 represents a short-lived mid-1st century cremation cemetery although there is a suggestion of possible earlier activity with at least four burials (Table 3.12). Two of these cremations were contained within a ring-ditch and a penannular ditch respectively. Although certainty is impossible this cemetery may have been associated with the contemporary Site RB.2B, located just over 100m to the southwest. If this is the case then it is intriguing to note the presence of some disarticulated noncremated human bone in the ditches of RB.2B indicating the existence of two different rites for disposing of the dead, although as the disarticulated human bone is not closely dated these need not be contemporary. There is no stratigraphic or artefactual way to determine the relative sequence of burials in C3. The distribution of the burials suggests that they were arranged upon the same southwest to northeast axis as the later Romano-British settlement in the area (Phase RB.2A.2). Indeed, it is likely that the putative low mounds associated with the penannular ditch and more especially the ring-ditch continued to act as landscape features for a considerable period afterwards. These ditches also continued to act as a focus for later burials (Phase RB.2A.2). The duration of Phase RB.2A.1 is uncertain; it may have lasted only a few years, but, alternatively, could have spanned several decades. In broad terms all the elements of C3 fit within the broad regional tradition of Aylesford- Swarling burial (Hill et al. 1999), and it is a relatively unexceptional example of the type. The best-preserved element of C3 was a circular ring-ditch (F.2025); this was 100% excavated and was 4.9m in diameter and up to 0.82m wide, with a maximum surviving depth of 0.28m. There is some ambiguity about the origins of this feature; the fill of the ring-ditch contained a Trinovantian/Catuvellani coin of c. 20BC 10AD (sf.769) which was recovered during preexcavation metal-detecting. Whilst this cannot be directly associated with any burials and is earlier than the apparent primary burial it seems improbable that its location is purely coincidental. Located centrally within the ring-ditch was a heavily truncated cremation (F.910). The ring-ditch contained some pottery dating to later than 100AD indicating that it was still partly open in the 2nd century AD, although a low central mound probably survived for considerably longer and the silted up ditch continued as a focus for burial with a later cremation (F.824) and inhumation (F.2036). 134

32 Feature Phase Cut Bone Wt (g) Crem. container Dish lid Secondary vessels Additional items Comment 824 2A A A A A.1 Not identifiable Not identifiable Circular, c m, 0.35m+ deep Not identifiable Ovoid, m, 0.23m+ deep Table 3.12: Cremations from RB.2A. [2295] Whiteslipped closed vessel AD [2609] Blackslipped medium sized jar, AD [5870] Grogtempered beaker 40 70AD [5872] QG1 beaker AD [5874] QG1 beaker 40 70AD [5876] QG1 beaker 40 70AD [5988] Q4 butt-beaker 30 70AD [7356] Buff sandy butt beaker 30 70AD Unk. Unk. Unk. Unk. Unk. [5870] CSGW beaker 40 70AD Copperalloy rosette brooch [2623] Iron disc [5868], sheep ribs [5426] Unk. Unk. Unk. [7357] Q1 small jar AD [7357] Coarse sandy greyware flagon AD [7358] Coarse sandy greyware closed vessel AD [7359] Fine sandy micaceous greyware butt beaker 40 80AD Single hobnail, possibly an incidental inclusion Secondary burial in ring ditch F Heavily truncated. Ring ditch indicated sate of 100AD+ Heavily truncated, in centre of ring ditch Largely intact Heavily truncated Truncated, in centre of penannular ditch 135

33 Located c. 3.7m to the northwest of the ring-ditch was a pennaular ditch (F.2088), this was 100% excavated although its southeastern portion had been removed by later features, and was c. 3.6m in diameter with the opening facing to the northeast. It was less substantial than the ring-ditch; at up to 0.29m wide, with a maximum surviving depth of 0.11m. The terminals of the pennaular ditch appear to have been elaborated with two postholes (F.2085 and F.2089), located outside the entrance. Inside the penannular ditch there was an undated posthole (F.2084) and a cremation (F.2109). There was also a single hobnail in the general fill. The penannular ditch probably silted up even more rapidly than the ring-ditch, although again a low central mound may have survived. In common with the ring-ditch the penannular ditch was also a focus for later burial, with an inhumation present (F.2093). Located c. 7.5m southwest of the ring-ditch was another cremation pit (F.1958); this was broadly circular with a diameter of c m and a maximum surviving depth of 0.35m. Although no trace of a surrounding ditch survived it is conceivable that an insubstantial example could have been entirely removed. The first depositional act was the placement of some articulated sheep ribs ([5426]), which lay under the pots. Then four pots were all broadly similar in character and identifying the cremation container is problematic. Only 109g of cremated bone was recovered; of this 38g came from the general fill of the pit and no individual pot contained more than this (weights 1g, 17g, 19g, 34g). Osteological analysis will be needed to clarify the nature and structuring of this burial. The final burial associated with the initial phase of C3 was a heavily truncated cremation (F.2005) located 15m to the northeast of the ring-ditch. Again, although no trace of a surrounding ditch survived it is conceivable that an insubstantial example could have been entirely removed. Given the scale of later features in this area and the extent that ploughing had truncated some of the cremations it is eminently plausible that further examples once existed and have left no archaeological traces. It is possible that a small number of isolated pits and postholes that are either unphased or assigned to later phases may be contemporary with this phase; however, there are no specific instances where a compelling case can be made for this. Phase RB.2A.2 - Initial Ditched Enclosure System (mid/late-1st to mid-2nd century AD) Probably following on immediately from an earlier cemetery (Phase RB.2A.1) the area was sub-divided by a series of ditched enclosures. This system was created in the period c AD; as a significant transition occurred at RB.2B soon after c. 70AD, it is plausible that the two events are contemporary. Whilst the most substantial of the enclosures of this phase (Enc.2A.1) is relatively well defined the others (Enc.2A.2) are less clear, in large part because of their overlap with the subsequent enclosure system (Phase RB.2A.3). Enc.2A.1 runs along the northern side of Way 3 and had an entranceway connecting it to Way 3. There appears to have been another enclosed area to the northwest of Enc.2A.1, this in turn had one or more enclosed areas to its northeast effectively creating an L-shaped pattern (Enc.2A.2). This form provides support for the idea that a routeway (Way 4) existed to the northwest. Some rather ephemeral ditches and other features to the northeast of Enc.2A.1 suggest the possibility of some form of trackway running beside it. Although a number of internal features were identified within the enclosures these do not form a coherent pattern and there is no clear evidence for occupation or particular activities taking place. No wells were present within the enclosures, indicating that water was probably obtained from a well located within Way 3. Although certainty is impossible it is likely that some burials in the ditches of the earlier cemetery (C3) date to this phase, although it is possible that they 136

34 are later (Phase RB.2A.3). The circular ring-ditch (F.2025) contained a cremation (F.824) and the inhumation of a child (F.2036), while the penannular ring-ditch (F.2088) contained another inhumation of a child (F.2093). The ditched enclosure system of Phase RB.2A.2 was substantially modified (Phase RB.2A.3) at some point c AD, having lasted for c years. Given the relative paucity of evidence it is unlikely to have functioned as an independent settlement; one possibility is that it operated in conjunction with RB.2B, although even in conjunction they lack a clear focus of occupation/activity. One possibility is that the most substantial enclosure (Enc.2A.1) acted as a stockyard with the other less substantial enclosures fulfilling other agricultural roles. Enc.2A.1 - Enc.2A.1 is broadly rectangular with an internal extent of c. 57m by 35m covering 1500 square metres. The main ditch for Enc.2A.1 (F.1855, 1873, 1883, 1901, 1953, 1984, 2016, 2133) showed no sign of any re-cuts and was up to 2.3m wide, with a maximum surviving depth of 0.8m. There were also several features interpreted as earlier quarry pits on the line of the enclosure ditch (F.1871, 1874, 1987), these probably represent elements of the initial construction of the enclosure. An entranceway on the southeastern side of the enclosure consisted of a series of inter-cutting pits and short lengths of ditch (F , , 1983). The entrance was c. 4.2m wide; despite the associated features there is no evidence to indicate the existence of an above ground structure. There were a number of internal features within Enc.2A.1, these were all relatively shallow and none displayed any particularly interesting characteristics. These consisted of a ditch that may represent some form of sub-division (F.1880), four postholes (F.1857, 1865, 1866, 1892), nine pits (F.1889, 1891, , 1925, ) and some curving irregular linear features that are probably wagon/cart ruts (F , 1926). A possible trackway along the northeastern side of Enc.2A.1 is indicated by some shallow ditches (F.1882, 1916, ), plus some pits (F ) and a posthole (F.1870). Enc.2A.2 - The other less well-preserved ditches probably represent two to four enclosures extending for c. 93m by 48m covering c square metres. The ditches for these are generally much narrower and shallower than those of Enc.2A.1; they are up to 0.9m wide with a maximum surviving depth of 0.6m (ditches F.1534, 1549, 1756, 1951, , 2002, 2006, 2010, 2013, 2024, 2027, , 2040, 2052, 2055, , 2072, 2076). The ditches of these enclosures were relatively insubstantial at up to 0.98m wide with a maximum surviving depth of 0.60m. There were a number of internal features within these enclosures, as with Enc.2A.1 these were all relatively shallow and none displayed any particularly interesting characteristics. These consisted of three postholes (F , 1905) and nine pits (F.1854, 1893, 1896, 1904, 1914, , 1952, 2031). Several secondary burials located within Enc.2A.2 and closely associated with the earlier cemetery (Phase RB.2A.1) probably relate to this phase, although there is no dateable material associated with them and they could conceivably be later. The burials consist of one cremation (F.824) and two inhumations of young individuals (F.2036, 2093; Tables ). Phase RB.2A.3: Modified and Expanded Ditched Enclosure System (mid-2nd to late-3rd century AD) At some point, c AD the ditched enclosure system of Phase RB.2A.2 was substantially altered and expanded. Although the alignment of the enclosure system and some of the boundaries remained the same the overall pattern was markedly different. The enclosure system covered c. 93m by at least 155m covering square metres, although a substantial proportion of 137

35 this lay outside the area of excavation. This was the most archaeologically visible phase of activity at this site and it was part of this that was recognised in the evaluation phase as a rectilinear network of ditches with four enclosures (Evans & Newman 2010). The overall enclosure can be divided into four zones; a large enclosure to the east (Enc.2A.3), a well defined central strip of three square enclosures (Enc.2A.4 6) and a less well defined western strip of enclosure(s; Enc 2A.7) and an area to the north (Enc.2A.8). The large eastern enclosure (Enc.2A.3) was accessed by an entrance on its eastern side and was largely devoid of internal features, suggesting that it probably functioned as a stockyard. The three smaller enclosures immediately to the west (Enc.2A.4 6) are likely to have fulfilled specialised agricultural roles, although the nature of these is uncertain. The western enclosure (Enc.2A.7) was probably a farm of some kind, and it is notable that the ditch(es) dividing it from the enclosures to the east is markedly wider and deeper than the others of Phase RB.2A.3. The lack of a pronounced slope in the base of the ditch makes it unlikely that this greater scale relates principally to drainage, but it would have made it a much more formidable barrier to livestock. The internal sub-divisions within Enc.2A.7 suggest the presence of buildings, although direct evidence for these is lacking. Although four wells are present (RB Wells 13 16; Table 3.14; Fig. 3.10) they are unlikely to have all been created at the same point in time and two went out of use before the end of this phase (RB Wells 13, 16). Despite this it is difficult to conceive of credible scenarios where less than three wells were in existence at some points in time, suggesting a considerable level of water provision. A partially lined pit (F.1997) indicates that some of this water was for specialised non-domestic purposes (Fig. 3.11). Although this enclosure system maintained the same alignments as the earlier system (Phase RB.2A.2) there appears to have been a shift in emphasis. Whereas the most significant of the earlier enclosures (Enc.2A.1) had been accessed from Way 3 to the southwest, the most significant or at least utilised enclosure of this phase (Enc.2A.7) fronted onto Way 4 to the northwest and the largest enclosure (Enc.2A.3) was accessed from what appear to have been open fields to the southeast. The re-use of building stone in the construction of RB Well 16 and tesserae plus tile in pit F.1997, combined with a general scatter of tile and other material in a range of features, indicates the demolition or substantial modification of a major Romano-British building in the general vicinity c AD. This enclosure system certainly survived until after 250AD, and probably until c AD with no major changes or modifications indicating that it was in existence for c years. This demonstrates a considerable level of continuity, spanning several generations. Enc.2A.3 - The boundaries that relate to Enc.2A.3 include those that relate solely to this enclosure (F.1856, 1861, , 1881, 1884, plus F.2125 and eleven in the trenches beyond the limit of excavation), plus those it shares with other enclosures (F.1932, 1986, 1990, 2009, , 2030, 2032, 2054, plus pits F which appear to be related to the enclosure ditch). The ditches of this enclosure varied considerable, with its northwestern and southeastern sides being markedly wider than its southwestern side. The ditches were up to 138

36 2.8m wide with a maximum surviving depth of 0.45m. Enc.2A.3 was 53 67m wide (the variation due largely to it being stepped on its western side) and 100m+ long, with the evidence from the trenches to the north of the area indicating that it is 140m+ long. It covers at least 5100 square metres, but its true size is likely to be over square metres. There was a 4.6m wide entrance on its eastern side. Internal features within Enc.2A.3 were sparse and consist of a single nondescript pit (F.1879) and a tree throw (F.1913). Even if the un-phased features belong to this phase and some assigned to Phase RB.2A.2 were re-dated the enclosure would still remain largely empty. Enc.2A Ditches that relate to Enc.2A.4 6 include those that relate solely to these enclosures (F , , , , 2049, 2051, 2053, 2061, plus pit F.2050 marking a corner), those that are shared with Enc.2A.3 to the east (F.1932, 1986, 2009, 2030, 2032, 2054) and those that are shared with Enc.2A.7 to the west (F , 2017, , 2043, 2057, , , 2094, 2110, plus pits F.2039, marking junctions). The ditches varied considerably in size and were up to 2.6m wide with a maximum surviving depth of 0.74m. Enc.2A.4 had internal dimensions of 33m by 26m and an extent of 790 square metres; there were no internal features that could definitely be assigned to this enclosure. Enc.2A.5 had internal dimensions of 31m by 24m 660 square metres, features within the enclosure consisted of seven pits which form a cluster that was partially investigated during the evaluation (F.1920, , 1936, 1941, ), a ditch (F.1940) and what is probably a wagon/cart rut (F.1937). There was also an inhumation (F.1935) and a disturbed area with human bone (F.1931), which may indicate the location of a second inhumation. Enc.2A.6 had internal dimensions of 29m by 22m and an extent of 490 square metres; the only associated feature was a pit (F.2081). Feature Phase Cut dimensions Skeleton Orientation Coffin Hobnails Grave goods Dating evidence from grave fill Comment A.3 Unk Unk. Unk. Unk. Unk. 150AD A A A.3 1.9m by 0.6m, 0.25m+ deep 0.61m by 0.35m, 0.20m+ deep 0.71m by 0.41m, 0.05m+ deep NW SE SW NE NE SW Table 3.13: Inhumations from RB.2A AD+ 100AD+ (from ring ditch) Heavily disturbed, status uncertain Aligned to Enc.2A.5, skull removed by ploughing, possible pillow stone Secondary burial in ring ditch F.2025 Secondary burial in penannular ditch F Heavily truncated Enc.2A.7 - The boundaries that relate to Enc.2A.7 include those that relate solely to this enclosure (F.1521, 1566, 1579, 1650, , 2064) and those that are shared with Enc.2A.3 6 to the east (F.1990, , 2017, , , 2043, 2057, , , 2094, 2110, plus pits F.2039, , ). The ditches varied considerably in size and were up to 2.6m wide with a maximum surviving depth of 0.74m. The enclosure is 70m+ long and 16 34m wide, covering 1775 square metres. 139

37 F metres centimetres Figure Site RB.2A Romano-British well 15

38 Figure Site RB.2A partly-lined pit F.1997

39 There were numerous internal features within Enc.2A.7. These consisted of a number of ditches within the enclosure (F.1560, , 1675, 1679, 1717, 1753, , 1945, 1965, 1995, 2003, , 2071, ). It is conceivable that Enc.2A.7 was in fact several separate enclosures, but the pattern is not coherent to enough to convincingly identify these. There were also a number of isolated features, including five pits (F.1602, 1695, 1715, 2004, 2080) and four postholes (F.1716, , 2075). In the southeast corner of Enc.2A.7 there was a distinct cluster of features; these appear to be associated with RB Well 13 and are within a small subenclosure defined by a ditch (F.2011). In addition to six pits that lacked any particular distinguishing features (F.1996, , 2012, 2026) there was a partially lined pit (F.1997). The partly lined pit is a steep almost vertical sided sub-rectangular feature 0.90m by over 0.84m in extent with a surviving depth of 1.2m. There was c. 0.08m thick clay lining across the base of the pit that extended up the lower c. 0.15m of the sides. Pressed into the clay lining the base were a number of tile fragments (26 pieces weighing c. 9kg) and tesserae (34 weighing 546g). The tesserae were predominantly of limestone and had evidence of two earlier phases of use prior to their employment in the pit. The function of this partially lined pit is unclear, given its proximity to RB Well 13 and its clay lining then it is likely that it held water. The tile and tesserae lining would have provided a markedly less easily damaged/disturbed base than clay alone. Of the four wells belonging to this phase, three are located on the boundary between Enc.2A.3 6 and Enc.2A.7 (RB Wells 13 14, 16), with a fourth (RB Well 15) located just to the west of this line within Enc.2A.7. There are indications that all four wells were principally orientated towards Enc.2A.7, but they may well also have been accessible from Enc.2A.3 6. RB Well 13 (F.2058) is a broadly circular feature c. 5.1m in diameter, with a surviving depth of 2.16m. Although indications of the shaft survived no trace of the actual lining was present. The evidence is slightly ambivalent, but on balance the well was probably box-lined. There is no dateable material directly associated with the construction of the well, but on stratigraphic grounds it cannot be earlier than c. 120AD. Pottery probably associated with the use or initial backfilling of the well dates to AD, but this may well be residual. There was a notable concentration of pits and other features immediately to the northwest of this RB Well 13; these appear to be located within a small sub-enclosure defined by a ditch (F.2011). It is likely that some of these are functionally linked to the well. The most notable of these was a subcircular pit with a lining of tile fragments and tesserae in clay (F.1997). There was good insect preservation from the backfilling of the well shaft. RB Well 14 (F.2023) is a large broadly circular pit c. 4.4m in diameter, with a maximum surviving depth of 1.67m. Although indications of the shaft survived no trace of the actual lining was present. This well was probably box-lined. No dateable material could be directly associated with the construction of this well; however, on stratigraphic grounds it must be later that c. 150AD. Some pottery that may be associated with the use of the well is no earlier than 250AD; material from the backfilling of the well shaft includes pottery that was no earlier than 300AD, two coins issued in AD and one issued in AD. RB Well 15 (F.2056) lay within a sub-circular construction cut 4.2 by 4.0m in extent, with a maximum surviving depth of 1.7m. Located in the northwestern part of the well was the lower portion of a square box-lined well that was 1.25m in extent. This was constructed from four oak uprights with on-edge horizontal oak boards running between them and horizontal ash poles that held the uprights in position during construction. This lining was constructed from a single tree and it may prove possible to date its felling through dendrochronology. There was no pottery associated with the construction of this well, but startigraphically it cannot be earlier than c. 200AD. Deposits associated with the use of the well contained pottery of , and AD. The initial backfilling of the well contained pottery of , , and AD, plus leather of the later 3rd 4th centuries AD. This indicates that the well was in use in the 3rd century and backfilled in the 4th century. Two leather adult shoes, one of one piece construction the other of nailed construction, were recovered from the backfilling of the well shaft. Also in the backfilling of the well shaft there was good waterlogged preservation of plant remains, pollen and insects. RB Well 16 (F.2044) is a roughly circular feature 5.4m by 5.3m in extent, with a maximum surviving depth of 1.96m. Although no trace of a lining proper survived the well was probably box-lined. There were some large pieces of stone in the base of the well; these appear to have been placed there solely to provide firm/dry footing during construction. 142

40 Whilst some of the stone is simply large un-worked cobbles, which could have been obtained from the local gravels, others are faced and shaped blocks that must be reused from a substantial stone structure. There were also two in situ upright round-wood stakes; these appear to be linked to the placement of the stones in the base of the well. Pottery associated with the construction of the well is dated to AD, whilst pottery that relates to the initial backfilling of the well dates to AD. Some material that probably represents the later infilling of a relict hollow after the well had gone out of use dates to AD. A good assemblage of insect remains was recovered from the backfilling. Enc.2A.8: A number of ditches from trenches to the north of the excavated areas appear to relate to the Phase RB.2A.3 network, but cannot be linked to particular enclosures. These have been designated Enc.2A.8; however they may represent more than one enclosure. Associated features include ditches F.128, , 2112, , 2123 plus pit F RB Well Feature Phases Lining Construction date Backfilling date A.3 Box? AD AD A.3 4 Box? 150AD AD A.3 4 Box 200AD AD A.3 Box? 200AD AD Table 3.14: Wells from RB.2A. Comment Associated with partly lined pit. Insect remains. Three coins from backfilling of shaft. Lining probably constructed by sites inhabitants form single oak tree. Insect, pollen and waterlogged plant remains. Potential dendrochronology. Stones in base for firm footing. Insect remains. Phase RB.2A.4 - Contracted Ditched Enclosure System (late-3rd to mid-4th century AD) In the late-3rd century the focus of activity contracted solely to Enc.2A.7. Although a small amount of 4th century pottery and a coin were found elsewhere this appears to represent material trapped in relict hollows and tree throws. The ditches of Enc.2A.7 were still extant, but much reduced in both width and depth compared to Phase RB.2A.3. Two wells continued into this phase (RB Wells 14 15), indicating a considerable water supply given the reduced scale of activity. One possibility is that the role of Enc.2A.7 and its associated wells was to provide water for use at RB.2C. The final cessation of activity probably dates to the mid-4th century AD, none of the pottery need be later than this date and the latest coin was issued in AD. Overall, a date for the end of activity of c AD appears likely. As the wells both appear to have been deliberately backfilled this indicates a degree of continuing care rather than straightforward abandonment, perhaps suggesting that the area of the enclosure passed directly into agricultural usage. The environmental evidence from the backfilling of RB Well 15 also presents a picture of broad landscape continuity, rather than decline and abandonment. Features which continue from Phase RB.2A.3: RB Wells 14 15, ditches F.1600, New features created in the 4th century: pits (F.1999, 2012), tree throw (F.2047). 143

41 Features which appear to represent relict material trapped in hollows: F.1877, 2032, RB Well 15. Unphased A six poster building (RB Str. 03) and a number of isolated discrete features including pits (12; F.1872, 1890, , 1924, 1942, 1944, 1946, 1985, 1988, ) and postholes (13; F.1885, 1902, 1915, 1919, 1927, 1930, , 1943, 1947, 1959, 2048, 2077) cannot be assigned to a particular phase. RB Str. 03 (Fig. 3.12) is probably a six poster with one posthole not surviving 5.95m by 2.7m in extent consisting of postholes F , which are m in diameter with surviving depths of m. There was no dateable material from any of the postholes and this structure is assigned to the Romano-British period purely because of its proximity to definitely Romano-British features. There is no nearby Prehistoric activity; the closest identified being PS3 c. 62m away. Material Culture and Environmental Evidence In general only low densities of material culture were recovered from individual features at RB.2A, producing low- to medium-sized overall assemblages. This was the only Romano-British settlement where no dense pottery deposits were identified. The pottery and other material associated with cremations is important, apart from that there are few individually noteworthy finds although a copper alloy disc brooch (F.2043) and a lead steelyard weight (F.1856) were recovered. The building materials recovered (tile, tesserae, stone) relate solely to re-use for a range of secondary functions and are therefore of limited interpretive value at a site-specific level. Similarly some of the material, such as the iron slag and coins that were used to manufacture copies, appear to represent overspill of material whose principal concentration is found at RB.2C. Although the bulk of the environmental samples from RB.2A contained only low densities of charred material, samples from several well fills produced well-preserved waterlogged plant remains, pollen and insects that permit detailed environmental reconstruction. Additionally, the preserved timbers from one of the wells provide valuable information about woodworking etc. Summary RB.2A consists of a short-lived mid-1st century AD cremation cemetery (Phase RB.2A.1), which was succeeded by a ditched enclosure system that was probably utilised as stockyards and for other agricultural roles (Phase RB.2A.2). This enclosure system was modified and expanded in the mid-2nd century AD (Phase RB.2A.3), although still largely utilised as stockyards and for other agricultural roles the presence of four wells and a partly lined pit indicates that other activities were taking place. In the late-3rd century the enclosed area contracted, although the majority of the wells continued in use until the mid-4th century (Phase RB.2A.4). As enclosures were abandoned during the late-3rd and 4th centuries the areas they formerly occupied appear to have passed directly into agricultural usage. 144

42 Structure 3 0 metres 10 F.1963 F.1964 F.1962 F.1961 Archaeological feature Excavated slot F metres Figure Site RB.2A plan of Romano-British structure 3

43 RB2B.1 RB2B.2 Enc 2B.1 Human bone Enc 2B.2 Well 17 Well 18 Well 17 Way 3 Way 3 Enc 2B.3 Enc 2B metres Figure Site RB.2B phase plans for RB2B.1-2

44 Settlement RB.2B RB.2B is the most southeasterly of the four spatial elements that constitute RB.2, located to the southwest of Way 3 (Tables ; Fig. 3.13). The area of the settlement was completely investigated, although its northeastern portion had been severely affected by Post-Medieval/Modern quarrying. The area was broadly triangular in form and covered up to c. 147m northwest to southeast and 112m southwest to northeast, covering square metres. There is some Prehistoric activity in this area (Ring-ditches 1 2 & C4); whilst this substantially predates the Romano-British period it is possible that low mounds associated with the ring-ditches still survived as features in the landscape. The settlement was located primarily upon gravel geology, with clay along its western side. The site is largely relatively flat lying at m OD, with its western edge located on edge of a slope leading downwards to the west. Enclosure Phase Length (m) Width (m) Extent (square m) 2B.1 2B Largely empty, but probably focus of occupation 2B.2 2B Largely empty 2B.3 2B (6470 after Largely empty extension) 2B.4 2B (1020 after truncation) Largely empty Table 3.15: RB.2B enclosures. Comments No. of enclosures Enclosed area (square m) Wells Burials 2B ?1 2B ,770 (rising to 11,570)*?1 Table 3.16: Proxy measures for activity levels at RB.2B. * - figures include areas of Enc.2B.1 2. The sequence at RB.2B can be divided into two phases: Phase RB.2B.1: initial ditched enclosure system (mid-1st century BC mid-1st century AD) 147

45 Phase RB.2B.2: modified and expanded ditched enclosure system (mid- 1st century late-2nd century AD) Phase RB.2B.1 - Initial Ditched Enclosure System (mid-1st century BC to mid-1st century AD) The earliest phase of activity at RB.2B covers a broadly triangular area c. 73m northwest to southeast and 98m southwest to northeast, covering 4100 square metres. There were two enclosures; a triangular inner enclosure at the northern apex (Enc.2B.1) and an outer enclosure to the south and east of this covering the rest of the area (Enc.2B.2). There are stratigraphic hints that Enc.2B.1 may have been established prior to Enc.2B.2, this is also supported somewhat by the distribution of pottery dated to c. 50BC 50AD but is nonetheless uncertain and may simply reflect the nature of later re-cutting episodes. The origins of this settlement is unclear, although there are no features that can be dated to earlier than the mid-1st century AD there is a quantity of pottery of c. 50BC 50AD, indicating that the ditched enclosure system originated at some point during this period. The overall pattern suggests the presence of at least two roundhouses, which have left no direct archaeological evidence. There were no wells located within the site; however, water was probably obtained from RB Well 17 located nearby but assigned to Way 3. The ditches on the western side of the site were markedly deeper and are likely to have usually been filled with a considerable depth of standing water. The presence of disarticulated human bone in the enclosure ditches indicates that some form of disposal of the dead, such as excarnation, took place. It is also possible that the mid-1st century cremation cemetery to the northeast (phase RB.2A.1) may be linked to the inhabitants of RB.2B. The ceramic evidence plus the complete lack of coins suggests that the initial ditched enclosure system went out of use some time soon after c. 70AD, having been in existence for c years. This transition appears to have involved the deposition of several moderately sized groups of pottery in the enclosure ditches (F.1809/1810 plus 1731, 1688, F.1765). As the establishment of the ditched enclosure at RB.2A (Phase RB.2A.2) dates to c AD it is conceivable that the two events are contemporary. Enc.2B.1 - Enc.2B.1 is broadly triangular, with a maximum internal extent of c. 57m by 35m covering 1500 square metres. The main ditch for Enc.2B.1 appeared to have two distinct phases, but as the later re-cut was more substantial than the primary phase traces of the earlier phase were highly fragmentary (F.1697, 1706, , 1729, 1748, , 1818, 1823, ). The later, better preserved, phase was up to 2.85m wide, with a maximum surviving depth of 1.2m. There were also some pits that are probably elements of the digging of the ditches (F.1858, 1886). To the west and north of the main enclosure ditch there was a smaller outer ditch located c. 2.7m away (F.1744, 1765, 1832), this was up to 1.35m wide, with a maximum surviving depth of 0.76m. It appears likely that Enc.2B.1 contained a roundhouse, direct evidence for this is lacking and there were few surviving internal features within Enc.2B.1. These included some relatively slight ditches (F.1688, 1696, 1714, 1737), plus a single pit (F.1738) and a posthole (F.1705). Enc.2B.2 - Enc.2B.2 is broadly triangular with its northern apex effectively removed by Enc.2B.1 with a maximum internal extent of c. 98m by 73m, covering c square metres. The ditches (F.1142, 1574, 1593, 1598, 1645, 1731, , 1818, 1853; plus related pit F.1646) are up to 3.15m wide, with a maximum surviving depth of 1.35m. Lying within Enc.2B.2 was 148

46 a broadly oval sub-enclosure c. 30m by 18m in extent, this clearly had at least two phases and the ditches were up to 1.7m wide, with a maximum surviving depth of 0.6m (F , , , 1682, , 1692, 1698; plus related pits F ). This sub-enclosure probably contained a roundhouse, direct evidence for this is lacking and the only surviving internal features within the sub-enclosure were three relatively nondescript pits (F.1664, 1693, 1699). Internal features within the rest of Enc.2B.2 include a number of narrow and shallow ditches that do not form any coherent pattern (F.1594, 1641, 1672, 1681, 1727, 1730, 1732, 1741), 20 relatively nondescript pits (F.1621, , 1635, 1640, 1642, 1647, , 1707, 1719, 1747, 1808) and three postholes (F.1728, ). Phase RB.2B.2 - Modified and Expanded Ditched Enclosure System (mid-1st century to late-2nd century) Some time soon after c. 70AD the ditched enclosure system went through a major modification. The area that had been enclosed previously (Phase RB.2B.1) shows little further evidence of activity, although it is likely that relict banks associated with at least some of the earlier ditches survived meaning that the two earlier enclosures (Enc.2A.1 2) continued to exist in some manner. New rectangular ditched enclosures were established to the southeast of the pervious area running up to Way 3; the larger (Enc.2B.3) immediately adjacent to the earlier area of activity was c. 49m wide by up to 15m long with minor internal ditches indicating that it was accessed from the previous area of enclosure to the north (Enc.2A.2). Attached to the southeastern side of Enc.2B.3 was a smaller enclosure (Enc.2B.4), this was up to c. 59m long by 15m wide and was accessed from the southeast. Although Enc.2B.4 encroached upon the ditches of Prehistoric ring-ditches 1 2 it is plausible that there may still have been low remnant mounds that were respected. As in the preceding phase there was no well present within the site, again was probably obtained from RB Well 17 and subsequently RB Well 18 located nearby but assigned to Way 3. At some point Enc.2B.3 was expanded to c. 59m, this involved incorporating what had previously been part of Enc.2B.4. It appears that Enc.2B.4 continued to exist, but was reduced to a width of only c. 13.5m. This would still have potentially respected any low remnant mounds of R-D1 2. There were few internal features within the enclosures and only low quantities of pottery and other material were recovered. The evidence of Phase RB.2B.2 could well relate solely to stockyards and other agriculture enclosures, potentially being farmed from RB.2B.A. The Phase RB.2B.2 enclosure system is poorly dated due to a low density of material; there is some pottery of 150AD+ but no categorically 3rd century material, and no coins were recovered. This suggests that the ditched enclosure system ended c AD, having been in existence for c years. In the mid-2nd century AD the northern c. 18m of the apex of Enc.2A.1 was truncated, becoming part of RB.2C. It is unclear whether after c. 200AD there were any surviving relict banks within RB.2B, or if these had been entirely removed by a combination of natural processes and agricultural activities. Enc.2B None of the major cut features associated with Enc.2B.1 2 appears to have continued into Phase RB.2B.2. It is, however, likely that some of the banks associated with the 149

47 more substantial ditches were still extant. Determining which ditches this applies to is problematic, but a tentative list can be suggested (F.1574, 1645, 1697, 1720, 1729, 1731, 1748, , 1818, 1823, 1887). Possibly related to these relict banks some pits were dug near the northern apex of Enc.2B.1 (F ), perhaps marking it in some manner. There was also a pit at the southern edge of Enc.2B.2 (F.1634). Enc.2B.3 - Enc.2B.3 was initially c. 49m wide by up to 115m long, the ditches are up to 2.35m wide with a maximum surviving depth of 0.76m with no evidence for any re-cuts (F , 1398, , 1488, 1639, 1791, ). There were also some pits along the line of the ditches that appear to relate to their construction (F.1411, 1487, , ). Within the enclosure some small ditches appear to delineate a c. 26m wide routeway within the enclosure (F.1545, 1586). Enc.2B.3 was subsequently extended becoming c. 59m wide (ditches F.1388, 1394, 1420, 1426). The initial phase of Enc.2B.3 covered 4600 square metres rising to 6500 square metres when it was extended. The internal features were relatively nondescript and consisted of seventeen pits some in clusters plus three postholes (pits F.1399, 1547, , 1573, 1587, , 1597, , 1617, 1623, 1644: postholes F.1422, 1565, 1749). Enc.2B.4 - Enc.RB.2B.4 was c. 59m long by 15m wide and was accessed through a c. 45m wide gap to the southeast. The ditches for these are up to 1.0m wide with a maximum surviving depth of 0.35m with no evidence for any re-cuts (F.1384, 1389, 1392, 1423, 1427, 1535). No internal features were present within Enc RB.2B.4, although it is conceivable that low remnant mounts from Prehistoric ring-ditches 1 2 still survived. The initial phase of Enc.2B.4 covered 2050 square metres falling to 1000 square metres when Enc.2B.3 was extended. Unphased The only unphased features present within RB.2B were two postholes (F.1571, 1589). Material Culture and Environmental Evidence In general, only low densities of material culture were recovered from individual features, although several moderately sized assemblage of pottery were recovered, with RB.2B producing low to medium overall assemblages. The bulk of the environmental samples from RB.2B contained only low densities of charred material samples and no deposits with good waterlogged preservation were encountered. Summary RB.2B is the only site investigated that indicates Latest Iron Age (50BC 50AD) settlement securely pre-dating the Roman conquest of the mid-1st century AD. There appears to have been a small farmstead with a ditched enclosure system (Phase RB.2B.1) that lasted for c years which ended soon after c. 70AD. The ditched enclosure system was then modified and expanded (Phase RB.2B.2), this transition appears to have been marked by the discard of several groups of ceramics and it appears that the site was entirely given over to stockyards and other agriculture enclosures with no occupation proper. This transition is also noticeable for the shift from curvilinear (Phase RB.2B.1) to rectilinear (Phase RB.2B.2) enclosure forms. This enclosure system was modified once, but went out of use after c years in the mid/late-2nd 150

48 century AD. As activity in the vicinity continued until the mid-4th century (at RB.2A and RB.2C) this means that for the final c years of the overall sequence of Romano-British activity RB.2B had ceased to be a distinct place. Settlement RB.2C RB.2C is the central portion of the western side of the four spatial elements that constitute RB.2 (Tables ; Fig. 3.14). The main investigated area spanned up to c. 193m northwest to southeast by c. 107m southwest to northeast, covering square metres. The edges of the settlement were well defined, apart from on its northeastern side where there was an area of intensive Post-Medieval/Modern quarrying, The settlement was located upon both gravel and clay geology, with a distinct downwards slope from eastnortheast to west-southwest lying at m OD. Enclosure Phase Length (m) Width (m) Extent (square m) 2C.1 2C Largely empty 2C.2 2C Focus of activity with sub- enclosures, probable buildings, inhumation cemetery 2C.3 2C Largely empty, single inhumation 2C.4 2C Focus of activity with zoning, buildings, three wells 2C.5 2C Focus of activity with zoning, markedly curvilinear. Quasi-midden, three wells, lots of material Table 3.17: RB.2C enclosures. Comments Enclosure No. of enclosures Enclosed area (square m) Wells Burials 2C.1 2C C Table 3.18: Proxy measures for activity levels at RB.2C. 151

49 Enc 2C.1 Enc 2C.2 C2 Str 4 Str 7 Well 7 Well 9 Well 10 Enc 2C.4 Well 8 Enc 2C.5 Well 11 Well 12 Double-ditched system Double-ditched system Enc 2C.3 RB2C.1 RB2C.2 RB2C.3 0 metres 100 Figure Site RB.2C phase plans RB2C.1-3

50 Feature Phase Cut dimensions Skeleton Orientation Coffin nails Hobnails Grave goods Dating evidence from grave fill Comment 076 2C C C C C C C C C C C C C m by 0.52m 2.26m by 1.02m 1.96m by 0.60m 2.15m by 0.80m 2.43m by 0.75m 2.35m by 0.55m 1.99m by 0.97m 0.82m by 0.40m 2.10m by 0.90m 1.80m by 0.80m 1.60m by 0.60m 2.15m by 1.00m 2.31m by 1.30m None SSE NNW SSE NNW SSE NNW WSW ENE SSE NNW SSE NNW SSE NNW SSE NNW SSE NNW SSE NNW SSE NNW SSE NNW SSE NNW 8 23 From evaluation [1206.1] copper alloy finger ring Pottery 150AD+ Pottery 150AD Nene Valley colourcoated beaker [1642] between feet, AD Colour-coated beaker [1690] AD Colchester colourcoated beaker [1748], AD Pottery 100AD+ Pottery 50AD+ Pottery AD Pottery 50BC 50AD 9 20? piece of binding 2 7 c , probably re-used timber Table 3.19: Inhumations from RB.2C. 54 Iron lumps and bar near waist Colchester colourcoated mini bagshaped beaker beside right tibia [2494] AD, wooden box to right of head or shoulder, linch pin and knife blade by waist Colour-coated beaker between feet [2423], AD Nene Valley colourcoated beaker [2500] by right hand/pelvis AD Pottery AD Pottery 50AD+ Pottery 150AD+ In different enclosure from others Skull placed between feet, probably decapitated but head area removed by land drain Small cut with no surviving bone, could be infant or child burial with no survival. Frag of shale bracelet may be incidental inclusion 153

51 RB Well Feature Phases Lining Construction date C.2 Box? 150+ Backfilling date 3rd 4th century C.2 Box? th century C.2 Box? rd century , , , C.2 Box? Mid 2nd century Late 3rd century 2C.2 Wattle th century , 1220, C.2 Wattle? th century Table 3.20: Wells from RB.2C. Comment Possible associated structure Coins gilded copper alloy disc brooch, a copper alloy steelyard Possible associated structure Probably 3 4 wells 2 distinct phases of well with separate shafts. 1st phase has stones in base for firm footing. 2nd phase has preserved lining. Iron pilum, copper-alloy bracelet, iron gouge, antler working. Good waterlogged plant remains 2 phases of well, 2nd phase had associated temporary revetments The sequence at RB.2C consists of three phases: Phase RB.2C.1: limited pre-settlement activity (mid-1 st to mid-2nd century) Phase RB.2C.2: initial ditched enclosure system (mid-2nd to mid-3rd centuries) Phase RB.2C.3: modified and contracted ditched enclosure system (mid-3rd to mid-4th centuries) Phase RB.2C.1 - Limited Pre-Settlement Activity (mid-1st to mid-2nd century) By the mid-1st century AD RB.2C had been effectively defined as an area by default; with RB.2D to the north, RB.2B to the southeast and the doubleditched boundary to the west. The area is largely devoid of archaeological features, with just a single pit (F.1082) and two tree throws (F.976, 1496). A possible potin coin from the area (sf.121) appears to be an isolated find. There is some pottery that predates 150AD from this area, the only apparent concentration of material albeit a relatively low one appears to relate to the double-ditched boundary. 154

52 Phase RB.2C.2 - Initial Ditched Enclosure System (mid-2 nd to mid-3rd centuries) The area was enclosed in the mid-2nd century, although this event cannot be closely dated directly by evidence from RB.2C there is strong evidence that it corresponds to or slightly post-dates the commencement of Phase RB.2D.3 in c AD. Overall RB.2C was a large slightly irregular rectangular area with ditches on its southeast, southwest and northwest sides. Due to Post- Medieval/Modern quarrying there is no direct evidence regarding the northeastern boundary. Of particular note is the kinked and stepped form of the northwestern boundary, which indicates a form of primacy to RB.2D. It is also noteworthy that the southern boundary of RB.2C incorporated what had previously been the northernmost area of RB.2B. This overall area was divided into southwestern and northeastern halves by a multi-phase curving ditch, which is effectively a continuation of the eastern boundary of Enc.2D.3 to the north. In the southwestern half of RB.2C there were three well defined enclosures (Enc.2C.1 3), the situation in the northeastern half is less clear. Although zones are apparent broadly mirroring the threefold division of the southwestern half these lack clear delineating features and are therefore treated as a single enclosure (Enc.2C.4). Of the three southwestern enclosures the northern (Enc.2C.1) and southern (Enc.2C.3) have little or no evidence for internal features although the southern contained a single inhumation (F.527) and appear to be open spaces, perhaps utilised as stockyards or for other agricultural purposes. The central enclosure (Enc.2C.2) which is the smallest of the three and also incorporates what had previously been the end of the double-ditched boundary has much more evidence for usage. Its southwestern side appears to have been re-established in a slightly different location on two occasions and there is evidence for the creation of two successive sub-enclosures, plus evidence for two less substantial sub-enclosures or possibly structures. There was also an inhumation cemetery in its southwest corner (C2; Fig. 3.15). The northeastern enclosure (Enc.2C.4) had three distinct zones of activity. The northern zone opposite Enc.2C.1 was intensively sub-divided by a series of small ditches, indicating intensive utilisation, and had evidence for a building (RB Str. 04; Fig. 3.16). The central zone opposite Enc.2C.2 had evidence for three wells (RB Wells 07, 09 10) and a building (RB Str. 07; Fig. 3.16). Finally, the southern zone opposite Enc.2C.3 was largely devoid of activity. Many features on the eastern side of Enc.2D.4 were truncated by Post- Medieval/Modern quarrying and it is likely that the site continued to the east. Indeed it is possible that the primary focus of activity lay to the northeast. The evidence of Phase RB.2C.2 indicates the creation of a rural agricultural settlement of some kind, with evidence for zoning and a significant water supply. It is unclear to what extent RB.2C functioned independently or if it was simply a spatial element within a larger overall settlement complex. The date for the transition to Phase RB.2C.3 is somewhat uncertain, as pottery that definitely dates to after 250AD and coins of the 260 s AD and later are also not present a date of c AD appears most likely. There is a suggestion that the very end of this period may be the most likely (see Phase RB.2D.3). This would indicate that Phase RB.2C.2 spanned c years. 155

53 19th Century Quarries Cemetery 2 Settlement 2C F.631 F.658 F.629 Archaeological feature Excavated slot F.500 F.666 F.630 F.848 F.501 F.856 F.849 F.510 F metres Figure Site RB.2C Cemetery 2

54 19th Century Quarries Structure 4 Structure 7 0 metres 25 F.1112 F.1111 F.1110 F.899 F.1055 F.1056 F.1015 F.920 F.923 F.922 F.921 F.1073 F.900 F.1058 F.927 F.898 Structure 4 Structure 7 F.924 F.1033 Archaeological feature Excavated slot 0 metres 5 Figure Site RB.2C: Plans of Romano-British structures 4 and 7

55 Enc.2C.1 - Enc.2C.1 is rather irregular but broadly square enclosure c. 56m northwest to southeast by c m southwest to northeast covering 3800 square metres. Its western boundary (F.807, ) appears to have been re-cut up to twice along some of its length, while its northern side (F.696, 807) shows no evidence of any re-cuts. Its eastern side which it shares with Enc.2C.4 shows evidence for numerous re-cuts (F.686, , 876, 882, 941, , , , 1011, ). The southern boundary of Enc.2C.1 appears to be primarily linked to Enc.2C.2 and shows evidence for several phases of re-cutting (F.822, 851, 882, 953). The only identifiable internal feature within Enc.2C.1 was a group of relatively slight ditches on its eastern side (F , 1104, 1129) that effectively create a small semienclosed area c. 13.5m by 5m in extent. Enc.2C.2 - Enc.2C.2 is rather irregular but broadly square enclosure c m northwest to southeast by c. 50m southwest to northeast covering 2600 square metres. It shares its northern boundary (F.822, 846, 851, 882, 953) with Enc.2C.1. Its main western boundary (F.518, 525, 957) shows no sign of being re-cut, but appears to have been re-established in a slightly different location on two occasions (F ). Its southern boundary which it shares with Enc.2C.3 shows no sign of any re-cuts (F.517, 957, 1151). Its eastern side which it shares with Enc.2C.4 shows evidence for numerous re-cuts (F.866, 888, , 969, 1003, 1063, 1416, 1469, , 1548). On the eastern side of Enc.2C.2 there is evidence for two successive rectangular subenclosures. The first (F.857, 864, 950, 1076, 1102, 1134, 1136) was c. 30m by 12m in extent. The second (F.836, 851, 881) was c. 32m by 22m in extent. There were also two potential subenclosures that were of lesser extent and were much slighter in form. The form of these suggests that they may relate to buildings. One was c. 11m by over 2.4m in extent (F ), while the other (F.821) was c. 20m by 20m in extent. In the southwestern quarter of Enc.2C.2 there was an inhumation cemetery, this consisted of eleven inhumations plus one possible example (Table 3.19). The graves are all aligned south-southeast to north-northwest, and the skeletons are extended and supine with the head to the south-southeast. Although there are what might be identified as rows of graves these are not particularly well defined. There is evidence that coffins were used in some, but not all instances and the presence of hobnails indicates that some but not all the individuals were buried with shoes. Pots were relatively common as grave goods and there were also a range of iron objects. The cemetery appears to span the period c although it may have continued slightly after that time. Other relatively nondescript minor internal features located within Enc.2C.2 included a few pits (F.826, 829, 954), a tree throw (F.1137) and a short length of ditch (F.825). Enc.2C.3 - Enc.2C.3 is rather irregular but broadly rectangular enclosure c. 80m northwest to southeast by c. 44m southwest to northeast covering 4000 square metres. It shares its northern boundary (F.517, 957, 1151) with Enc.2C.2. Its western boundary (F.511, 956) shows no signs of being re-cut, whereas its southern boundary (F , 1739, 2128) was re-cut at least once. Its eastern side which it shares with Enc.2C.4 shows evidence for numerous re-cuts (F , 995, ). There are relatively few internal features within Enc.2C.3. On its northern side there is a westsouthwest to east-northeast aligned inhumation (F.510) and associated pit (F.527). This burial shares an alignment with the adjacent boundary ditch, which was only c. 1.35m away. The burial is located only c. 12m from the cemetery in Enc.2D.2, it may well simply be an outlier of this cemetery. The only other feature was a pit on the eastern side of the enclosure (F.994). Enc.2C.4 - Enc.2C.4 is a rather irregular rectangular enclosure c. 183m northwest to southeast by c. 45m+ southwest to northeast covering 7800 square metres. Its western boundary is shared with Enc.2C.1 3 and displays evidence of multiple re-cuts (for feature listings see above). Its northern boundary (F.697, 703, 937) shows sign of a single re-cut plus an associated posthole (F.704), whilst its southern boundary (F.1704, 1708, 1766) has no traces of any re-cuts. No evidence for an eastern boundary was identified, this side may have been open or alternatively the evidence may have been removed by extensive Post-Medieval/Modern quarrying. 158

56 Although there were no ditched boundaries Enc.2C.4 can be divided into three zones. The northern zone opposite Enc.2C.1 was intensively sub-divided by a series of relatively slight ditches into small square and rectangular sub-enclosures indicating intensive utilisation (F.761, , 901, , 921, 923, , 963, , , 1112, , 1032, 1105, 1113, 1128, 1133). There was also evidence for a building (RB Str. 04), this consists of two sides of a beam-slot structure 10.3m by 8.5m in extent (F , 920, 922, 924, 927, 1033, ; Fig. 3.16). The beam-slots are m wide with a maximum surviving depth of m. There was pottery in several of the beam-slots, this material appears to relate to the point after the beam-slots had either been removed or rotted and dates to after 200AD, suggesting that the building went out of use in the 3rd century. There were also a few relatively nondescript pits (F.940, ) and postholes (F.1015, 1025, 1035) The central zone opposite Enc.2C.2 had evidence for a building (RB Str. 07) and several wells (RB Wells 07, 09 10). It is unclear to what extent all three wells are contemporary, but there must almost certainly have been points in time when two wells were in existence. There were also some relatively slight ditches indicating that the area was sub-divided (F.1017, 1057, 1416) and three nondescript pits (F.1107, 1218, 1435). RB Str. 07 consists of two sides of a beam-slot structure 6.7m by 5.9m in extent with a posthole at the junction of the beam-slots (F.1056, 1058, 1073; Fig. 3.16). The beam-slots are m wide with a maximum surviving depth of m, whilst the posthole was 0.15m in diameter with a maximum surviving depth of 0.25m. There was pottery in both the beam-slots, this material appears to relate to the point after the beam-slots had either been removed or rotted dates to after 200AD, suggesting that the building went out of use in the 3rdcentury. RB Well 07 (F.1364, , ) was broadly oval 3.0m by >2.0m in extent with a maximum surviving depth of 0.85m. No trace of the original lining survives, but the well was probably box-lined. Pottery from the uppermost surviving fill was dated to AD, although this could relate to either the construction of the well or its backfilling. No material of note was recovered from this feature. A rather irregular vertically sided 4.5m long linear feature on the northern side of RB Well 07 (F ) may well be associated with it and could be a structural beam-slot. The pottery from this feature is not closely dateable, but some is no earlier than 150AD. Also apparently contemporary and related to the well are three features to the north of the beam-slot interpreted as root-boles and suggestive of a short length of hedge or similar feature (F ). RB Well 09 (F.996 plus F and F.1052) was broadly oval 7.55m by 5.95m in extent with a maximum surviving depth of 0.80m. No trace of the original lining survives, but the well was probably box-lined. The pottery from this feature is not closely dateable, although some is no earlier than 150AD, and the well does not appear to have continued in use beyond the 3rd century. A series of postholes around the edge of RB Well 09 are probably part of a structure around the shaft (F , F.1052). A shallow >3.5m linear feature to the north of the well (F.1042) may be a beam-slot of an associated structure. RB Well 10 (F.1028 and F ) represents a complex sequence of inter-cutting features. Although excavated and recorded as seven separate features it probably represents a smaller number of actual wells perhaps three or four although it proved impossible to articulate these well enough for certainty. No traces of the linings of these wells survived, it is however likely that they were box-lined. As far as they can be individuated the wells appears to be broadly circular or oval and m in extent, with maximum surviving depths of up to 0.85m. None of these features contained any dateable material; howeverm based on stratigraphy and other evidence the sequence of wells begins no earlier than c. 150AD and does not continue beyond c. 250AD. Finally, the southern zone opposite Enc.2C.3 was largely devoid of activity, with no features that can conclusively be assigned to this phase apart from a general hollow (F.1072) 159

57 Phase RB.2C.3 - Modified and Contracted Ditched Enclosure System (mid-3 rd to mid-4th centuries) Around c AD the nature of the ditched enclosure system at RB.2C altered markedly. It is possible that this transition included the construction of RB Well 08, if this is the case then it occurred no earlier than a coin issued in AD. The enclosed area contracted noticeably and became markedly more curvilinear. There was also a marked change in the deposits associated with the settlement, these are much darker and contained greater densities of material culture than in earlier phases. As well as filling discrete features this darker material was identifiable as a localised soil horizon (F.584) and was present as slumping in the upper portions of many earlier features. It appears if much of the site was effectively covered by a midden, or at extremely dirty soil layer. Although some spatial distinctions are apparent there are no neat divisions into identifiable separate enclosures, as a result this is all defined as Enc.2C.5. Within Enc.2C.5 up to five zones are apparent, although delineations are not clear enough for these to be independently defined. Three wells were present within Enc.2C.5 (RB Wells 08, 11 12; Figs ), all of which were probably contemporary with one another. Away from Enc.2C.5 there is some evidence for limited activity elsewhere, or at least material becoming trapped in hollows created by earlier features. The nature of the activity at Enc.2C.5 is rather uncertain; indeed it represents the most intractable archaeology of the entire site and cannot be disentangled to the same degree. Whilst this can in part be interpreted as a failure of analysis, it in part reflects a genuine difference in the nature of the archaeology. The failure to neatly organise the material is because it was in a sense messier than the rest of the archaeology. The existence of three contemporary wells indicates a considerable provision of water. The marked change in deposits, whilst primarily due to changes in refuse disposal strategies, nonetheless indicates considerable levels of activity. Some of the material in these deposits especially the considerable quantity of iron slag present may hint at a specialised industrial role. The key issue with regard to this, however, is to what extent this material was generated at RB.2C and how much was transported from elsewhere to be deposited there. One potentially relevant factor is that it appears that the inhumation cemetery (C2) ceased to be used at the start of Phase RB.2C.3, indicating either that burial shifted to some unidentified location or a change in the nature of the site. A minimum of 28 4th century AD coins were recovered from RB.2C, the latest are two SECVRITAS REIPVBLICAE issues dating to AD, and there is also no pottery from the site that need post date 350. Overall, a date for the end of activity of c AD appears likely. Enc.2C.5 - Enc.2C.5 is extremely irregular; its maximum dimensions are c. 183m northwest to southeast by c. 74m southwest to northeast. It covered an overall area of 4500 square metres. The northernmost identifiable zone of Enc.2C.5 spans the earlier multi-phase curving ditch (Phase RB.2D.2). To the east one of the relatively square sub-enclosures of Enc.2C.2 was re-cut continued in use (F.845, 847, 851, 881). To the east the southern boundary continued as a curving sinuous boundary (F.1103, 1138, 1278, 1442, 1444). The northern boundary was stepped on the line of the earlier multi-phase curving ditch to widen the enclosed area (F.962, , 972, 1030, 1109). There was then evidence for subsequent multiple re-cutting of ditches, which appear to have contracted the enclosed area (F.939, 968, , , 991, , , , 1149, 1154, , 1443). 160

58 (3640) (3641) (3643) (3644) F.1148 (3639) (3633) (3642) (3634) (3645) 0 metres (3638) (3637) (3636) (3635) centimetres Figure Site RB.2C Romano-British well 11

59 1 1 Settlement 2C 2 (3544) F centimetres (3544) F.1168 F.1173 F.1168 F Quernstone (4220) (4219) (4217) (4221) (4097) 0 F.1402 metres (4222) (4098) Quernstone (4247) (4099) (4251)(4250) (4246) (4249) (4248) 2 (4099) centimetres (4249) (4097) (4098) Figure Site RB.2C: Romano-British well 12

60 The area immediately to the south of this appears not to have possessed its own boundaries, but to be bounded to the north and south by the parallel curving sinuous ditches of other zones. This was the area of Enc.2D.4 where wells had been located in Phase RB.2D.2 and the presence of a well (RB Well 08) indicates a level of continuity. There were also some relatively slight ditches that crossed the area creating minor boundaries (F.1010, 1068, 1089, 1225), twelve relatively nondescript pits (F.1002, 1065, 1067, 1088, 1101, 1224, , , 1520) and a single posthole (F.1401). RB Well 08 (F.1236) was broadly oval 6.1m by >3.8m in extent, with a maximum surviving depth of 1.14m. No trace of the original lining survives, but the well was probably box-lined. A considerable quantity of material was recovered from RB Well 08. Although stratigraphic differentiation was poor it appears that the bulk of the material was deposited during the construction of RB Well 08. There was a considerable quantity of pottery (>6kg) little of which was demonstrably residual some of which dated to after 250AD and two coins issued in AD and AD respectively. There were also considerable quantities of animal bone, iron slag and metalwork. Notable items include a gilded copper alloy disc brooch, a copper alloy steelyard and a probable iron awl. The backfilling of RB Well 08 dates to after 300AD as a small quantity of pottery of this date was recovered and a layer that seals it contained a coin issued in AD. To the south of this there was an area that may represent between one and three zones; however, it is impossible to coherently resolve them into a pattern. The area was dominated by a large number of re-cutting ditches (F.529, , 944, , 949, 961, , 1012, , 1022, 1031, 1036, , 1060, , , 1077, 1090, , 1108, , 1126, 1140, 1143, 1152, 1160, , 1214, 1391, 1795, , , , ). There were also some relatively nondescript pits (F , 928, 967, 1013, 1037, 1059, 1100, , , 1468, 2122), plus some postholes (F , 945, 948, 1219) and an infilled hollow (F.1139). A number of deposits represented general infilling over earlier features (F , 1051, 1066, 1071, 1155, 1167); this appears to represent localised elements of a general occupation soil (F.584). There were also two well complexes (RB Wells 11 12). RB Well 11 (F.1020, , 1148) was broadly circular and 3.2m in diameter, with a maximum surviving depth of 1.42m (Fig. 3.17). No trace of the original well lining of this phase survived, although the shape of the shaft indicates that it was made from wattle. A group of reused stone blocks on the south side of the base of the well appear to have been placed there to provide firm/dry footing during construction. The pottery from this well indicates a date for its construction no earlier than 250AD; this is supported by a leather shoe of c AD. Also associated with the construction were a fragment of a 3rd 4th century copper-alloy bracelet, an iron gouge and evidence for antler working. The shaft of the first phase of RB Well 11 was ultimately capped with sterile gault clay. The well shaft was then re-cut, moving c. 0.8m to the northeast. This re-cut consisted of a 0.9m diameter wattle-lined shaft constructed from a variety of roundwood, some of which had become heavily distorted by post-depositional pressure. The transition between the two shafts of RB Well 11 has some associated pottery that is no earlier than 300AD. How long the second shaft of RB Well 11 continued in use is unclear, its fill was relatively sterile and contained no closely dateable pottery. There was, however, an iron object identified as the shank and head of pilum or javelin that was nearly vertical as if it had been thrust into the already partially backfilled well. There was good preservation of waterlogged seeds from both phases of RB Well 11. RB Well 12 (F.1168, F.1220 and F.1402) was a complex entity that may represent two successive adjacent wells; unfortunately these cannot be fully individuated (Fig. 3.18). No traces of the lining of the well shaft(s) survived but it is likely that they were of wattle. Overall, the well consisted of a large irregular oval cut 3.3m by 3.2m in extent with a maximum surviving depth of 1.6m. No trace of any lining or other elements associated with the first phase survives. During the construction of the second phase of the well the sides of the construction cut became unstable, necessitating some temporary ad hoc revetment involving boards set on edge and held in place by stakes. Additionally, some stone and timber was piled in front of one of these revetments to hold it in place. One of the boards set on edge was part of the seat of a stool and another of the timbers was a reused structural element. One piece of stone may be a crude column base; alternatively, it was used as a post pad or a pedestal for a wooden column. No trace of the actual well lining survived, the nature of the temporary revetments makes it likely that the shaft was wattle lined. The pottery from 163

61 Well RB 12 spans the mid-2nd 4th centuries AD and there is a leather shoe of c AD; it seems likely that the first phase of well dates to the mid/late-2nd century AD, whilst a coin issued after 260AD may relate to the backfilling of the first phase. The final backfilling of the second phase took place after 300AD. Finally, to the south of Enc.2C.5 the existence of a boundary ditch (F , 2124) indicates an enclosed but open area of some type. There is also some evidence for material becoming trapped in hollows created by earlier features outside Enc.2C.5 (F.919) Unphased A range of features cannot be assigned to a particular phase of the sequence. The majority are isolated relatively nondescript pits (19) or postholes (12), to which no particularly meaningful interpretation can be attached (pits F.837, 875, 1014, 1016, 1078, 1081, , 1094, 1098, , 1175, , 1419, 1450, 1470, 2100: postholes F , 1029, 1034, 1038, 1083, 1099, 1132, 1141, 1410, 1486, 1774). Although it is conceivable that some of these features belong to Phase RB.2C.1, it is likely that the majority relate to Phases RB.2C.2 3. Material Culture and Environmental Evidence In general only low densities of material culture were recovered from individual features of the earlier phases at RB.2C. This changed markedly in Phase RB.2C.3, with large quantities of several types of material being deposited. Several types of material were particularly prominent, including iron slag, although there are significant issues in determining what if any relationship this material has to in situ activities. The ceramics associated with the inhumation cemetery are of some importance, particularly with regard to comparing them to those associated with cremations. Individually important items include a copper-alloy disc brooch (sf.212) and bracelet (RB Well 11), an iron pilum head (RB Well 11) and ballista bolt-head (F.1278), and part of a wooden seat (RB Well 12). Although the bulk of the environmental samples from RB.2D contained only low densities of charred material, samples from several well fills produced well-preserved waterlogged plant remains. Additionally the preserved wood from the wattle-lined well (RB Well 11) is significant. Summary Until c AD RB.2C existed principally as an empty area defined by more intensively utilised surrounding areas (Phase RB.2D.1). At that point it was divided into four enclosures (Phase RB.2C.2); this effectively linked the four areas of RB.2 into a single contiguous if not necessarily coherent entity. The evidence indicates that Phase RB.2C.2 relates to a zoned rural agricultural settlement with wells, buildings and an inhumation cemetery. In the mid-3rd century AD, probably in the late-260 s AD, the area of enclosures contracted markedly (Phase RB.2C.3). The character of activity changed markedly; the layout became curvilinear rather than rectangular, deposits became markedly darker with the core of the occupation effectively becoming a quasi-midden and large quantities of material particularly iron slag were deposited. This continued until c AD. 164

62 Settlement RB.2D Settlemnt RB.2D is the most northwesterly of the four spatial elements that constitute RB.2 (Tables ; Fig. 3.19). The main investigated area covered up to c. 125m northwest to southeast by c. 130m southwest to northeast, covering c square metres. On its eastern side there was an area of intensive Post-Medieval/Modern quarrying, additionally the site has not been investigated in its entirety and continues to the northwest, this area will be investigated in 2013 (Site V). The settlement was located upon both gravel and clay geology, with a distinct downwards slope from east-northeast to west-southwest lying at m OD. Enclosure Phase Length (m) Width (m) Extent (square m) 2D.1 2D Focus of activity with cremation cemetery, well 2D.2 2D Largely empty 2D.3 2D c Focus of activity with cremation/inhumation cemetery, four wells Table 3.21: RB.2D enclosures. Comments No. of enclosures Enclosed area (square m) Wells Burials 2D.1 2 2D D D.4 1 Table 3.22: Proxy measures for activity levels at RB.2D. The sequence at RB.2D consists of four main phases: Phase RB.2D.1: isolated small enclosures and cremations (early-1st mid-1st century) Phase RB.2D.2: initial ditched enclosure system (late-1st mid-2nd century) Phase RB.2D.3: modified and expanded ditched enclosure system (mid-2nd late-3rd century) Phase RB.2D.4: relict well continues in use (late-3rd early/mid-4th century) 165

63 Well 4 Well 3 Well 4 C1 Well 5 Enc 2D.2 C1 Str 2 Cremations F.683 Well 6 F.699 Enc 2D.1 Enc 2D.3 Double-ditched entranceway RB2D.1 RB2D.2 RB2D.3 RB2D.4 Str 5 Str 6 0 metres 100 Figure Site RB.2D phase plans for RB2D.1-4

64 Feature Phase Cut Bone Wt (g) Cremation container Dish lid Secondary vessels Additional items Comments 502 2D D D D D D D D.3 Not identifiable Not identifiable Not identifiable Not identifiable c. 0.6m diameter circle, steep sides and rounded base 0.25m+ deep Not identifiable Not identifiable 0.59m by 0.54m rectangle, 0.16m [1211] Q3 large jar AD [1220] Q3 medium sized jar AD [1227] Coarse, sandy greyware jar AD [1224] Q2 medium sized jar AD [1341] Q1 large necked jar AD [1270] Q1 large necked jar AD [1323] Q3 large necked jar AD [1334] Coarse sandy oxidised [1212] Central Gaulish Samian dish AD [1210] Fine sandy buff ware closed vessel AD None identified Unk. Unk. Unk. None apparent None apparent [1320] South Gaulish Samian dish AD [1271] South Gaulish Samian dish AD [1276] Central Gaulish Samian dish AD None Unk. [1225] Buff sandy flagon AD [1321] Buff sandy small globular flagon AD [1272] Whiteware flagon AD [1273] Q1 medium sized jar AD [1277] Oxidised sandy flagon AD [1278] Oxidised sandy indented beaker AD [1336] Colourcoated globular flagon AD Unk. Unk. [1267] 20 hobnails and a frag of a knife or tool blade, [1326] four rounded pebbles possibly gaming pieces Two nails [1329] and [1343], probably incidental inclusions Nail and a small, thin corroded disc [1328], probably incidental inclusions None identified Substantively complete, upper portion truncated Very heavily plough damaged Extremely heavily plough damaged Truncated by land drain and plough damaged Substantively complete, upper portion truncated Substantively complete, upper portion truncated Substantively complete, upper portion truncated Southwestern outlier. Substantively 167

65 542 2D D D D D D.1 deep with vertical sides and flat base Not identifiable Oval, 0.55m by 0.37m, with steep sides and flat base 0.40m+ deep Circular 1.3m diameter, 0.3m deep Circular c. 0.45m diameter, steep sides and concave base 0.18m+ deep Not identifiable Sub-square 1.1m by 0.9m, 0.3m deep Table 3.23: Cremations from RB.2D. medium jar AD [1361] Coarse sandy greyware closed vessel AD [1408] Fine sandy micaceous oxidised ware beaker AD [1669] Buff sandy flagon AD [2304] Horningsea greyware necked jar AD [2465] AD [2670] Buff sandy flagon 30 70AD Unk. None [1671] South Gaulish Samian dish with part of rim deliberately clipped AD [2302] East Gaulish Samian dish AD Unk. [2706] Fine sandy micaceous greywares platter 30 70AD [1337] Blackslipped plain rim beaker AD [1361] Colourcoated beaker AD [1409] Fine sandy micaceous greyware closed vessel AD [1667] Coarse sandy oxidised flagon AD [2305] Oxidised sandy ware small flagon AD Unk. [2670] buff sandy closed vessel AD None identified Possible spike or nail sf. 71, probably incidental inclusion None None identified None identified None complete, upper portion truncated. Nene Valley colourcoated flagon rim [1339] AD might be deliberate grave good. Southwestern outlier. Extremely heavily plough damaged Northern outlier. Substantively complete, upper portion truncated Substantively complete, upper portion truncated Substantively complete, upper portion truncated Extremely heavily plough damaged Complete 168

66 Figure Site RB.2D phase RB2D.1

67 Feature Phase Cut dimensions Skeleton Orientation Coffin nails Hobnails Grave goods Dating evidence from grave fill Comment 850 2D D D m by 0.78m, 0.25m deep 1.80m by 0.73m, 0.25m+ deep 0.98m+ long by 0.54m, 0.15m+ deep Table 3.24: Inhumations from RB.2D N S 2426 ENE WSW AD 7 150AD+ None N S Poor skeletal preservation Poor skeletal preservation Possible child burial, no bone survives RB Well Feature Phases Lining Construction date Backfilling date D.2 3 Box AD AD D.2 3 Wattle? AD AD D.1 3 Wattle? 50 70AD AD D.3 Wattle? AD AD Table 3.25: Wells from RB.2D. Comment Preserved box-lining, coins in construction, pot disc and dice in backfill Lots of pottery and quern from construction Waterlogged and substantial but heterogeneous assemblage of material Phase RB.2D.1 - Isolated Small Enclosures and Cremations (early-1st to mid-1st century) The earliest activity in the area of RB.2D consists of two small square enclosures and two cremation burials (Fig. 3.20). The two enclosures (F.683, 699) appear to be the earliest elements and it is unclear whether they were still in existence when the cremations occurred. Although these two morphologically very similar small square ditched enclosures predate the first phase of Romano-British ditched enclosure they are on the same alignment as the enclosure system and occur on the edges of subsequent enclosures, indicating a strong degree of continuity. The relatively small size of these enclosures precludes a substantial number of possibilities and based upon regional parallels they are probably Late Iron Age burial enclosures, with the central cremations entirely removed by ploughing. 170

68 The two cremation burials (F.638, 936; Table 3.23; Fig. 3.20) appear to be relatively early, based upon the dates of the associated pottery. One possibility is that they are effectively precursors of cemetery C1, located c. 35m to the northwest, and that the shift in burial focus relates to the expansion of activity in Phase RB.2D.2. These two burials are different from those of C1, and most of the other Romano-British period cremations at the site, in that they occur in readily identifiable relatively deep pits, which are substantially larger than required for the cremations themselves. There is no indication that these pits were dug for any purpose other than the cremations themselves. The cremations are located in relatively close proximity to one of the small square enclosures (F.683) and the possibility that they are in some way related cannot be discounted. There are a number of other pits in the area that could be related to this phase of activity (F.616, 632, 641); all contained no dateable material, had been dug with a noticeably greater degree of care than many of the Romano-British pits at the site and are morphologically relatively similar to the cremations. These pits may equally well be later in date, although with the benefit of hindsight it is unfortunate that they were not 100% excavated. It is possible that some other undated or poorly dated isolated features may belong to Phase RB.2D.1, but this cannot be convincingly demonstrated in any individual instance. The duration of Phase RB.2D.1 is difficult to assess, it could span anything from a few years to several decades falling in the period c AD. It is also impossible to determine any particular associations between this activity and contemporary settlements. It could conceivably be linked to RB.2B.2 c. 200m to the south-southeast, but there is no particular evidence to support this and the features might equally well relate to an unidentified settlement located to the east or north. Overall, the evidence suggests that Phase RB.2D.1 represents a direct precursor to Phase RB.2D.2, this is not to suggest any inevitably to the development merely that there was no discernible hiatus between the two phases and alignments etc. indicate that they both existed within the same overall conceptual landscape. The small square enclosure F.683 was 5.0m by 5.0m in extent, with ditches up to 0.58m wide with a maximum surviving depth of 0.18m. Stratigraphically this was the earliest feature in this area of the site and the fills were noticeably paler and contained less material than later Romano-British features in the area. The fills of the ditch of this enclosure contained very little material culture, but there was a small quantity of pottery dated to 50BC 50AD. No contemporary features were present within the interior of the enclosure; however, a significant proportion of this area had been removed by later ditches. The small square enclosure F.699, located 31m to the south-southeast of enclosure F.683, was 5.2m by 5.1m in extent, with ditches up to 0.72m wide with a maximum surviving depth of 0.32m. Stratigraphically this was the earliest feature in this area of the site and the fills were noticeably paler and contained less material than later Romano-British features in the area. The fills of the ditch of this enclosure contained very little material culture and produced no dateable material. No contemporary features were present within the interior of the enclosure. 171

69 F.850 C metres F.858 Archaeological feature Excavated slot F.505 F.504 F.869 F.503 F.867 (2304) F.823 (2306) F.507 F.502 F.535 F.542 F.506 F.508 F metres F.508 F.507 Figure Site RB.2D: Cemetery 1

70 F.722 F.715 F.713 F.712 F metres 100 F.1127 F metres Archaeological feature Excavated slot 10 Figure Site RB.2D: Double-ditched entranceway

71 Phase RB.2D.2 - Initial Ditched Enclosure System (late-1 st to mid-2nd century) Phase RB.2D.2 began around c AD, the first act was the creation of two adjacent rectangular ditched enclosures of broadly similar size with a shared boundary (Enc.2D.1 2). There is little evidence for activity in the southern of these two enclosures (Enc.2D.2), with few features present. The northern enclosure (Enc.2D.1) witnessed more activity; it appears to have rapidly been subdivided. The precise nature of these subdivisions is unclear as there were clearly several different iterations, but they cannot be disentangled. The nature of the enclosure suggests that buildings may well have been present, but none could be convincingly recognised. Located towards the northern edge of Enc.2D.1 was a probably wattle-lined well (RB Well 05). Due to localised groundwater issues the northwestern corner of Enc.2D.1 was badly preserved; however, it appear that rather than being fully enclosed there was a 7.6m wide ditched routeway with ditches that were markedly less substantial than the actual enclosures running to the northwest. Located at the junction of Enc.2D.1 and the routeway was a small cremation cemetery (C1), with nine burials associated with Phase RB.2D.2 (Fig. 3.21). Lying c. 30m south of Enc.2D.2 was a relatively elaborate double-ditched entranceway (Fig. 3.22). Although this lay outside Enc.2D.1 2, and effectively straddles the spatial boundaries between settlements RB.2C and RB.2D, it appears most likely to be associated with the enclosures of RB.2D, particularly given the lack of evidence for contemporary activity in RB.2C. No trace of any structure or other evidence for activity was associated with this doubleditched entranceway. Presumably whatever was present was sufficiently shallow that it left no archaeological traces. The ending of Phase RB.2D.2 and transition to Phase RB.2D.3 is marked by a number of deposits in Enc RB.2D.2. The most notable is a significant quantity of material associated with the creation of RB Well 04 (Phase RB.2D.3), there was also a deposit containing a significant quantity of pottery in the eastern terminal of a ditch (F.634). The end of Phase RB.2D.2 probably occurred c AD, if the creation of RB Well 03 (Phase RB.2D.3) is also linked to this event then the dating can be narrowed to c AD. This gives a duration for Phase RB.2D.2 of c years. Enc.2D.1 - Enc.2D.1 was a relatively regular rectangle 49m+ (west-east) by 45m in extent, covering an area of 2350 square metres. The eastern boundary of the enclosure is unclear, no ditches on this side were identified and the relatively shallow ditches leading to it simply petered out. On balance it is likely that this side was enclosed by a shallow ditch that has not survived, in terms of the broad layout and orientation of activity along the gravel ridge it is likely that the principal entrance to the enclosure was on this eastern side. The southern boundary which was shared with Enc.2D.2 (F.637, 653, 673) displays no evidence for any re-cuts and was up to 1.1m wide with a maximum surviving depth of 0.6m. The western boundary (F.659, 695, 786) had no evidence for any re-cuts and was up to 1.7m wide with a maximum surviving depth of 0.4m. The northern boundary (F.832, 897) had no evidence for any re-cuts and was up to 1.25m wide with a maximum depth of 0.18m. The northwestern corner of the enclosure was poorly preserved, however it appears that there was a southeast to northwest aligned 7.6m wide ditched trackway located in the corner. The northeastern ditch (F.565) was up to 0.82m wide with a maximum surviving depth of 0.08m and the southwestern ditch (F , 868) shows signs of being re-cut once and is up to 1.25m wide with a maximum surviving depth of 0.18m. Enc.2D.1 was subdivided by a series of ditches (F.589, 603, 611, 613, 620, 634, 640, 667, 769, 891). These effectively created six rectangular and square sub-enclosures of varying extent, 174

72 although it is unclear if they are all contemporary and some of the narrower rectangular enclosures are unlikely to have been functional. In some instances particularly the southwestern corner of the enclosure it seems likely that some of the additional ditches that created small enclosures simply represent minor re-alignments due to ground conditions becoming too wet. RB Well 05 (F.548) is roughly oval 7.6m by 4.1m in extent with a maximum surviving depth of 1.22m and was investigated during the evaluation phase (F.247). The profile of the well suggests that it was wattle-lined, although no trace of this survived. Given the survival of other organic material in the well this suggests that the lining was made from a tree species that does not survive well in waterlogged conditions. Material recovered during the evaluation phase included a considerable quantity of 2nd 3rd century pottery, fragments of a glass cup, a piece of a wooden lathe-turned furniture finial and waterlogged and charred plant remains. Plant remains indicate that the environs were covered in rich, open vegetation with species indicative of a nutrient-rich, disturbed landscape such as a settlement. The evaluation had entirely removed the base of the well, but further excavation produced more pottery of similar date plus vessel glass, shale and a 1st 2nd century coin. There is no material that can be definitely associated with the construction of the well; it appears to have been backfilled at some point in the 3rd century. Cemetery C1 consisted primarily of cremations (12 identified), plus a few inhumations (2 definite and one possible; Fig. 3.21). Heavy disturbance by later ploughing, which had almost obliterated several cremations, makes it entirely feasible that some burials have been removed in their entirety. The core of the cemetery consisted of a tight group of nine cremations, covering an area of 5.0m by 4.5m. There were also three outlying cremations, these appear to be somewhat later and probably relate to Phase RB.2D.3 as do the inhumations. There is no evidence for strong layout or organisation of the cremation cemetery, although there is no inter-cutting of burials. All of the cremations were placed within substantial coarseware vessels and at least seven had inverted Samian dishes used as lids to seal the cremated bone. At least nine of the cremations possessed secondary ceramic vessels, these come in a range of forms and fabrics and were typically placed beside the primary container, although in one case it was placed within the cremation container (F.506). Additional items are relatively rare and it is unclear if nails in some of the cremations (F.507 and F.508) are deliberate addition or not. More plausible as deliberate additions are some hobnails and stones (F.506). The dating evidence suggests that C1 spans the mid/late-1st to 2nd centuries AD, with no evidence for continuity into the 3rd century. A substantial pit (F.738) just outside the southwestern corner of the enclosure and a connecting ditch (F.741) presumably relate to either water management of the enclosure ditch system or represent some form of light industrial activity that utilised the water supply of the ditches. Enc.2D.2 - Enc.2D.2 was a relatively regular rectangle 58m+ (west-east) by 38m in extent, covering an area of 2000 square metres. The eastern boundary of the enclosure is unclear, no ditches on this side were identified and the relatively shallow ditches leading to it either simply petered out or were truncated by recent quarrying. On balance, it is likely that this side was enclosed by a shallow ditch that has not survived, in terms of the broad layout and orientation of activity along the gravel ridge it is likely that the principal entrance to the enclosure was on this side. The southern boundary (F.657, 662, 688) displays no evidence of being re-cut and is up to 1.5m wide with a maximum surviving depth of 0.7m. The western boundary (F.669, ) appears to have been re-cut at least once and was up to 1.45m wide with a maximum surviving depth of 0.7m. The northern boundary which was shared with Enc.2D.1 (F.637, 653, 673) has already been described (see above). There is minimal evidence for activity within Enc.2D.2; the only features that can conclusively be associated with it are three pits (F.685, 691 2) and a posthole (F.705). Double-ditched entranceway: The double-ditched entranceway consisted of an outer ditch (F and 722), an inner ditch (F ) and a beam-slot to the west that is probably associated (Fig. 3.22). The outer ditch was 19.7m north-south with a 1.6m wide gap in its centre and its longest east-west arm 8.4m. The inner ditch was m long. The ditches were m wide, with a maximum surviving depth of 1.28m. 175

73 F.831 F.841 F.589 F.842 F.621 F.590 F.853 F.854 F.855 F.840 F.838 F.839 F.734 Str 2 F.890 F.870 F.815 F.871 F.758 F.759 F.633 F.733 F.750 F.889 F.891 F.873 F.872 F.760 F.753 F.752 F.814 F.754 Str 5 F.790 F Str F.893 F.816 F.892 F.817 F.628 F.626 F.627 F.625 F.624 F.623 F.635 F.788 F.755 F.756 F.778 F.779 F.780 F.781 F.782 metres Archaeological feature Excavated slot F.894 Str 2 Str 2 F.663 F.664 F.795 F.796 F.797 Str 5 F.798 F.702 F.671 F.799 F.719 F.716 F.717 F.726 F.732 F Str 6 metres 5 Figure Site RB.2D: Romano-British structures 2, 5 and 6

74 Figure Site RB.2D Romano-British well 4

75 Figure Site RB.2D Romano-British well 3

76 Figure Site RB.2D pit F.770

77 Phase RB.2D.3 - Modified and Expanded Ditched Enclosure System (mid-2nd to late-3rd century) In the mid-2nd century the area went through a major re-organisation, various strands of evidence suggest that this occurred c AD. Although the boundaries of the two earlier enclosures (Enc.2D.1 2) appear to have still exerted an influence probably as relict banks/hollows rather than as ditches proper a new much larger sub-rectangular enclosure was established (Enc 2D.3). This enclosure represented an expansion of the area of the two previous enclosures to the south (c. 18m), west (c. 19m) and north (c. 19m) and a contraction of unknown extent on the eastern side. Enc.2D.3 was subdivided, principally along its west-southwest to east-northeast axis into three or four areas, although these distinctions appear to be too minor to warrant designating the areas as separate enclosures. Whilst most of the boundary ditches of Enc.2D.3 appear to be relatively short-lived although they presumably continued to exist as banks perhaps with associated shallow hollows the eastern boundary showed evidence for numerous re-cuts and additionally shifted markedly to the east, noticeably changing its alignment at some point. There is good evidence for at least one posthole and beam-slot building (RB Str. 02; Fig. 3.23) located on the eastern side of the enclosure and it is likely that at least two more also existed (RB Str. s 05 06). The pre-existing well (RB Well 05) continued in use, although rather than being located near the edge of the enclosure as previously it now lay c. 30m from it, with an irregular set of shallow channels presumably linked to water management linking the two (F ). Another, probably wattlelined, well (RB Well 04; Fig. 3.24) was established c AD within a small sub-enclosure on the northern edge of Enc.2D.3. The construction deposits associated with this well contained a considerable quantity of pottery and quern, this material may well relate to the group who utilised this area during Phase RB.2D.2 and could indicate some form of clearance occasioned by the transition. A third box-lined well (RB Well 03) was established in the mid-1st century dated by a coin issued in AD and lay c. 27m north of Enc.2D.3 (Fig. 3.25). Whilst its association with Enc.2D.3 cannot be categorically demonstrated it appears likely as a group of shallow channels (F ) similar to those associated with RB Well 05 ran from the well to the northern ditch of Enc.2D.3. There were also some more regular shallow ditches to the east of these (F ), which probably represent a form of boundary. A further well (RB Well 06) was established on the eastern edge of Enc.2D.3, but this did not take place until the early-3rd century. The earlier cemetery (C1) continued in use, at least for a time, although the dating evidence is slightly ambiguous in some cases as to whether particular burials relate to Phases RB.2D.2 or RB.2D.3. It appears to have changed in character, as the cremations (F.535, 542, 549) are no longer located within the earlier core of the cemetery but some c. 6 9m away. Instead in the area that had formerly been the core there were some inhumations (F.850, 858 and possibly F.867; Table 3.24). Evidence for other activity within the enclosure is limited; however, along its western side there are a number of pits (F.739, 740, 180

78 744, 770, 776; Fig. 3.26) that suggests some form of specialised activity was taking place. In general there is relatively little pottery dated to after 250AD from Enc.2D.3 and a total absence of the late-3rd mid-4th century coins that are present at settlements RB.2A and RB.2C. Overall, this indicates an end of activity by c AD. Therefore Phase RB.2D.3 lasted for c years. There are indications that the backfilling of one well (RB Well 03) was used as an opportunity to dispose of material, perhaps indicating a form of clearance of the enclosure. There was also a relatively substantial deposit of pottery, including material dated to after 250AD, in the eastern terminal of a ditch (F.693). Enc.2D.3 - Enc.2D.3 was an irregular rectangle that is markedly wider at its northern end, it was 120m long north-south, 46m wide at its southern end and 26m+ but probably c. 83m at its northern end. It covered an area of c square metres, representing a 95% increase of the combined area of Enclosures 2D.1 2. The southern boundary (F.711, 736, 765, 861, 863) appears to have been re-cut at least once and is up to 2.9m wide with a maximum surviving depth of 0.95m. The western boundary (F.737, 747, 749, 783, 789, 806, , 862) appears to have been re-cut at least once and is up to 3.1m wide with a maximum surviving depth of 0.95m. In addition, there was a shallower outer ditch c. 3.5m west of the main ditch (F.805), which was 0.8m wide with a maximum depth of 0.2m. The western boundary (F.561) shows no sign of being re-cut and is up to 2.8m wide with a maximum surviving depth of 0.85m. The eastern boundary is the most complex, with evidence of multiple re-cutting up to eight or nine times (F.559, 582, 595, 601, 612, , 636, 652, 660, , , 694, , 706, , 766, 771, , 830, 932). Individual re-cuts of the ditch were up to 1.8m wide with a maximum surviving depth of 1.1m). The central portion of this boundary shifted noticeably eastwards at some point, although the northern and southern corners of the enclosure remained stable. There is also evidence for some shallower ditches which connected some of the wells (RB Wells 3, 5) to the northeastern corner of the enclosure, these presumably relate to water management/drainage in some manner (F , ). Additionally, some short lengths of ditch leading into this enclosure ditch probably represent ad hoc drainage channels dug to deal with local water issues (F.600, 830, 1091). The enclosure was divided into at least four main sub-enclosures, which were predominantly aligned west-east. From south to north the southernmost sub-enclosure (northern boundary F.661, 693, 793) was 30m wide and it seems likely it appears that a possible relict internal mound of one of the Phase RB.2D.1 small square enclosures (F.699) may have influences its location. The next sub-enclosure (northern boundary F.608) was 40m wide. The third sub-enclosure was rather irregular as it re-used the line of an earlier ditch creating a distinct dog-leg (northern boundary reused ditch F.634, plus new ditches F.670, 676). As a result it was between 24m and 31m wide. The final subenclosure was 34m to c. 50m wide, additionally there were one or possibly two small subenclosure on its northern side (F.567, 569, 587), one of which contained RB Well 4. One well-preserved and two much more fragmentary Romano-British structures were identified. RB Str. 02 was the most substantial Romano-British structure discovered at Sites RB.2A D. It is a west-east aligned rectangular posthole and beam-slot structure 9.6m long by 5.2m wide. Although the basic plan of the structure remained stable there is clear evidence for several phases of modification as posts and other structural elements were replaced. 54 features can be associated with RB Str. 2; 49 postholes and five beam-slots (F.590, 621, , 633, 635, , , , 788, 790, , 831, , , , , 1062). The postholes are m in diameter with surviving depths of m. The beamslots are all highly truncated; they are up to 0.26m wide with surviving depths of m. The evidence suggests that the eastern end of RB Str. 02 was the most complex and was probably the front of the structure. It seems probable that the structure relates to the phase after the enclosure ditch immediately to the east had silted up and the new ditch several metres to the east was in use. Stratigraphically RB Str. 2 is relatively late in the sequence, cutting through several earlier ditches. The dateable material is very mixed; it all appears to relate to have been deposited at the abandonment of the structure. Some of the pottery dates to after 200AD but none is definitely later than 250AD, suggesting that the building went out of use in the 3rd century. 181

79 Based upon parallels with RB Str. 2, RB Str. 05 probably represents the rather more substantial eastern end of a poorly preserved rectangular timber structure. It consists of five surviving postholes (F , ) m in diameter with a maximum surviving depth of m, forming a line 5.8m long. Although two of the postholes contained pottery none of this was closely dateable. Similarly RB Str. 06 probably represents the rather more substantial eastern end of a poorly preserved rectangular timber structure. It consists of ten surviving postholes (F.671, 702, , 726, 732, ) m in diameter with a maximum surviving depth of m, forming a line 7.6m long. Although two of the postholes contained pottery none of this was closely dateable, although some is later than 150AD. RB Well 05 continued from Phase RB.2D.2 and three new wells were constructed (RB Wells 03 04, 06). RB Well 03 (F.564) is roughly circular and c m in diameter with a maximum surviving depth of 2.9m. In the centre of the well there was a square shaft c m by 0.85m in extent with a box lining c. 0.9m by 0.9m in extent constructed from thin tangentially sawn oak boards supported by upright in the corners. The pottery associated with the construction with the well is not closely dateable; two coins that probably relate to the construction of the well were issued in 81 96AD and AD respectively, whilst a third coin also probably related to the well construction was issued in AD. The pottery associated with the backfilling of the well is not closely dateable, but there were some trimmed discs and quern concentrated in initial backfilling. RB Well 04 (F.586) was roughly circular c m in diameter with a maximum surviving depth of 1.8m. It has a distinct funnel-shaped profile with the lower 0.5m consisting of a vertically-sided circular shaft 0.6m in diameter. This clearly indicated that this was a wattlelined well, although no trace of the lining survived. After the wattle-lining had been placed in position but before the bulk of the construction cut had been backfilled a considerable quantity of refuse was deposited in a band across the northern portion of the construction cut of the well. This consisted primarily of pottery, including several semi-complete or complete vessels that had broken in situ, and substantial fragments from at least two Old Red Sandstone querns. Animal bone was conspicuously absent, although bone survival was relatively poor traces were usually visible and it is likely that if any larger or more robust bones had been present they would have survived. The dumping of refuse into the construction fills of a well appears counter-intuitive, as the water accumulating in the wattlelined shaft would have percolated through this material. Either those building the well did not care which appears highly unlikely or the material was deliberately added perhaps in an attempt to improve the rate of water percolation. If this was the case then animal bone and other material deemed inappropriate may have been deliberately excluded. The pottery indicates a date c AD for the construction of the well and of after 300AD for the backfilling. The construction cut for the well is separated from the enclosure ditch (F.610) to the north, which appears to be contemporary, by only c. 0.55m. The implication of this is that there was no internal bank. It should be noted that the actual shaft of the well was a further c. 0.9m from the ditch line, creating a less perilous space but still probably implying the existence of a fence or hedge of some kind. The well appears to have been located in a small sub-enclosure >16m long and 11 9m wide defined by some small ditches (F.569, 587), with a 4.7m wide entrance at the eastern end. The western end of the sub-enclosure was not revealed, but there are hints of further sub-division (F.567). The presence of a probably Post- Medieval dice towards the top of the backfilled shaft suggests that this remained a shallow hollow or soft-spot for a considerable period. RB Well 06 (F ) was roughly circular and c. 4.1m in diameter with a maximum surviving depth of >1.1m. Although initially interpreted as two separate features this is likely to represent a single well whose form suggests that it was wattle-lined and probably c. 1.3m deep. Some material that can definitely be associated with the construction of the well dates to after 200AD. This well is relatively late in terms of stratigraphic sequence, as it is cut through several phases of ditch associated with Enc.2D.3. Some burials from cemetery C1 appear to relate to Phase RB.2D.3 (Table 3.23), including cremations F.535 (outlier to southwest with some pottery of 150AD+), F.542 (outlier to southwest with some pottery of AD) and F.549 (outlier to north with some pottery of AD+). Also belonging to Phase RB.2D.3 are the inhumations F.850, 858 (which contains some pottery of 150AD+) and the possible example F.867 (Table 3.24). Whilst some 182

80 of these burials contain pottery of AD150+ the argument is also based upon the fact that inhumations are unlikely to predate the mid/late-2nd century and also upon spatial logic. Neither of the inhumations had associated grave-goods or hobnails, but one appeared to have been interred in a coffin (F.858). Other relatively nondescript features of this phase include pits (F , 725, 731, 746), postholes F.787, , ) and a hollow (F.745). A group of five pits located close to the western edge are quite carefully shaped and probably represent some form of light industrial activity (F , 744, 770, 776). As these pits were cut into the natural clay it is likely that they held water and several associated gullies probably relate to water management for these features. A number of other pits in the vicinity may also be related, although these generally show less care in their digging. There are also some postholes (F ) that hint at some from of flimsy structure or fence line. Phase RB.2D.4 - Relict Well Continues in Use (late-3rd early/mid-4th century) The evidence suggests that general activity in Enc.2D.3 ceased by c AD. The sole exception to this is that RB Well 04 appears to have continued in use until the early/mid-4th century, as pottery dated to after 300AD was found in the backfilling of its shaft. One other find that may relate to this period but requires further analysis is a large Late Romano-British or Anglo-Saxon leaf-shaped spearhead (ditch F.711). It is plausible that the banks associated with Enc.2D.3 still existed, this suggests that the area of the enclosure and well were utilised for agricultural purposes conceivably by the groups associated with the continuing activity in RB.2A and RB.2C. This is supported by the care apparently taken to backfill the other wells and some other features plus the deliberate disposal of material in RB Well 03. All of this points to an interest in maintaining the area as a useful space, as opposed to outright abandonment. Unphased A range of features cannot be assigned to a particular phase of this sequence. The majority are isolated relatively nondescript pits (17) or postholes (7), to which no particularly meaningful interpretation can be attached (pits F , 571, 583, 588, 597, 616, 632, 641, 649, 655, 672, 698, 703, 708, 1021, 1061; postholes 580, 599, 650, , 938) There was also a tree throw (F.684) and a pair of linear features c. 1.5m apart (F.568, 570) are best interpreted as cart/wagon ruts, which match the standard width of c. five Roman feet. Material Culture and Environmental Evidence In general only low densities of material culture were recovered from individual features at RB.2D, there were however several significant assemblages of pottery and other material, the most important being associated with the construction of RB Well 05. The pottery associated with cremations is also significant. Other intrinsically important material includes a possible pilum head (ditch F.674) and a Late Romano-British or Anglo-Saxon spearhead (ditch F.711). The overall assemblages from RB.2D are of low to moderate size. The analysed environmental samples from RB.2D contained only low densities of charred material; there was good waterlogged preservation from at least one of the wells. 183

81 Summary RB.2D consists of a short-lived mid-1st century AD phase of small square enclosures and cremations (Phase RB.2D.1), which was succeeded by a ditched enclosure system (Phase RB.2D.2) with an associated well and cremation cemetery. This enclosure system was modified and expanded in the mid-2nd century AD (Phase RB.2D.3), when there is evidence for four wells and at least one moderately sized timber building. Phase s RB.2D.2 3 both appear to relate to a small occupied rural settlement of some type. In the late- 3rd century AD occupation proper ended, although the continued use of a well until the mid-4th century AD (Phase RB.2D.4) indicates that the area continued to be utilised as part of a wider agricultural landscape. Isolated Romano-British Features Only a single isolated Romano-British feature located away from the settlements and routeways was identified (posthole F.2102). In part this is a reflection of the fact that features located at a distance from the settlements are inherently less likely to contain dateable material, and it is therefore plausible that a number of the isolated undated features are in fact Romano- British. It is also possible that some of the features assigned to Prehistoric Settlements 1 2 that contained little dateable material may in reality be isolated Romano-British features. Even with these caveats, isolated Romano- British features must have been relatively uncommon, as there are fewer than 30 undated features and the bulk of the features from Prehistoric Settlements 1 2 are extremely unlikely to be Roman. It therefore appears clear that within the Romano-British landscape the digging of features was restricted to the areas of settlements and routeways. Material Culture and Environmental Assessments Although densities were low, the investigations scale means that substantial assemblages of many material types were recovered. The material has been broadly divided into artefacts, building materials and environmental remains. As the Romano-British material represents the most substantial assemblages from the site, where material is discussed under more than one period the methodology section is included here. With most specialist analyses undertaken prior to the full elucidation of the component areas of RB.2, this is not reflected in the work and this will be incorporated at a later stage. In such instances the designation RB.2A relates to RB.2A, whilst the designation RB.2B relates to RB.2B D. The main exception to this is the Romano-British pottery that takes into account the final division into Sites RB.2A D. It should be noted that no eggshell was recovered even from the environmental samples. Additionally, molluscs were very sparse in the environmental samples and these do not warrant further analysis. These are part of a more general paucity of what might be termed micro level material, as indicated by the extremely low quantities of fish, small vertebrates and non-domesticated bird bones recovered. This absence is true even of the mid- 3rd century and later darker quasi-midden deposits identified at RB.2C. 184

82 19th Century Quarries Archaeological features Pottery by count: Pottery by weight (g): metres Figure Distribution plot of all Romano-British pottery by count and weight

83 Romano-British Pottery Katie Anderson A substantial quantity of Romano-British pottery totalling 21,972 sherds, weighing 372kg and representing EVEs (estimated vessel equivalent) was recovered from North West Cambridge (Fig. 3.27). Of this, 19,510 sherds weighing kg ( EVEs) were recovered from the excavations, with a further 2462 sherds, weighing 38kg totalling EVEs, collected from the evaluation phase. The size of the Romano-British assemblage presented some issues in terms of the methodologies used for analysis and interpretation at this stage of the reporting process. All of the pottery was examined and recorded in accordance with the guidelines set out by the Study Group for Roman Pottery (Darling 1994). This involved the sorting of each context assemblage primarily by fabric. As expected from such a large assemblage, a wide variety of vessel fabrics were identified, including many from known sources, which were recorded by acronyms adapted from the Museum of London Roman Pottery Fabric Reference Collection (Tomber & Dore 1998). There were however a significant number of sherds deriving from unknown sources and kilns, which were given their own codes and once analysis was completed, these were often amalgamated into larger fabric groups as often there was little to distinguish between these fabrics, and without implementing this, the number of fabric types would have been unusable in any meaningful way. Fabric groups such as sandy greywares (one of the most ubiquitous) did not require a detailed fabric description; instead, they were grouped by coarseness. Other fabric types, particularly early Romano-British groups did require more detailed fabric descriptions in order to distinguish between different types. A full list of all fabrics recorded at North West Cambridge, along with a fabric description (where it is felt necessary) can be found in Table For the purposes of this report, all fabric tables utilise the fabric codes rather than the full name. Once fabrics had been established, the sherds were counted, weighed and sorted by form type; the latter conducted at two levels. Firstly, a basic form type, e.g. jar, bowl etc, and secondly a more detailed description of the vessel form (where available), such as indented for a beaker. When a sherd comprised of solely a base or rim, a description of this element was made, even if nothing else could be established, such as beaded rim and pedestal base. The diameter of all rim and base sherds was recorded along with the percentage of the vessel that remained. This enabled not only an indication of vessel size but also allowed for the calculation of Estimated Vessel Equivalents (EVEs). Other elements recorded included any decoration and/or use-wear with a note made of where on a vessel this occurred. A note was also made of any stamps, as well as whether a vessel was complete. A note was made if there were any refitting sherds between different contexts, though given the often generic nature of Romano-British pottery; a specific program of cross-context refitting was not attempted. Instead, sherds with distinctive vessel forms or decoration were the only ones likely to be recognized as coming from a single vessel. 186

84 Fabric code Full fabric name/description AHFA Alice Holt/Farnham ware AMPH Amphora - unsourced BAET Baetican amphora BAETL Baetican amphora (late) BB1 Black-burnished ware (Dorset) BB2 Black-burnished ware 2 BLKSL Black-slipped - unsourced BUFF Buff sandy ware - unsourced Buff soft Soft buff sandy ware - unsourced CALC Calcareous tempered ware CC Colour-coated ware - unsourced CC import Imported Colour-coated ware - unsourced CGBL Central Gaulish black-slipped ware CGCC Central Gaulish colour-coated ware CHFOX Cherry Hinton/Foxton fineware COLCC Colchester colour-coated ware CSGW Coarse sandy greyware - unsourced CSMGW Coarse sandy micaceous greyware - unsourced CSMRDU Coarse sandy micaceous reduced ware - unsourced CSOX Coarse sandy oxidised ware Flint CSGW Coarse sandy greyware with flint - unsourced FSBLK Fine sandy black-slipped ware FSBUFF Fine sandy buff sandy ware - unsourced FSGW Fine sandy greyware - unsourced FSMBLK Fine sandy micaceous black-slipped ware - unsourced FSMGW Fine sandy micaceous greyware - unsourced FSMOX Fine sandy micaceous oxidised ware - unsourced FSMRDU Fine sandy micaceous reduced ware FSOX Fine sandy oxidised ware - unsourced Grey slipped Grey-slipped ware - unsourced GROG Grog-tempered ware - unsourced HADBB Hadham black-burnished ware HADOX Hadham oxidised ware HADRDU Hadham reduced ware HADRS Hadham red-slipped ware HORNBB Horningsea black-burnished ware HORNGW Horningsea greyware HORNOX Horningsea oxidised ware HORNRDU Horningsea reduced ware HWC Highgate Wood C ware Imit BB Imitation black-burnished ware - unsourced 187

85 KOLN Cologne colour-coated ware LONDONW London fine reduced ware Metallic GW Metallic finished greyware - unsourced MOSBLK Moselkeramik black-slipped ware NVGW Nene Valley greyware NVCC Nene Valley colour-coated ware NVSC Nene Valley self-coloured ware NVWW Nene Valley whiteware OXFRS Oxfordshire red-slipped ware OXFWS Oxfordshire white-slipped ware OXFWW Oxfordshire whiteware OXIS Oxidised sandy ware - unsourced PAKCC Pakenham colour-coated ware PORD Portchester D ware Q1 Medium sandy fabric, bit abrasive to touch. Common to frequent silver mica. Frequent small quartz under 1mm. Sandwich fired grey core, oxidised edges black ext Q2 Coarse sandy-no other inclusions visible Q3 Q1 but with fewer and small flint inc and occasional small red inclusions Q4 Medium fine sandy, frequent silver mica Q5 coarse sandy fabric with rare large flint inclusions up to 4mm, poorly sorted Q6 Medium sandy fabric, rare to occasional flint up to 1mm Q7 Fine sandy, moderate silver mica, occasional to common buff inclusions QG1 Medium fine sandy fabric with common very small to small grog-inclusions-well sorted QG2 Medium fine sandy fabric with common small sub-rounded grog inclusions and rare to occasional flint inclusions, up to 1mm, poorly sorted RDUS Reduced sandy ware - unsourced RHOD? Rhodian amphora RS Red-slipped ware - unsourced SAM Samian - unsourced SAMCG Central Gaulish Samian SAMEG East Gaulish Samian SAMSG South Gaulish Samian SHELL Shell-tempered ware SWNWS Swanspool white-slipped ware TR? Terra rubra VEG Vegetable-tempered ware VRW Verulamium whiteware WATT Wattisfield greyware West Stow West Stow fine reduced ware WS White-slipped ware - unsourced WW Whiteware - unsourced Table 3.26: Roman pottery fabric codes. 188

86 The final element of the pottery recording relates to dating. For the purposes of this assemblage, each sherd/group of sherds was given an earliest date and a latest date ; the division of this information into two columns allowing for more in-depth analysis. Throughout this report there are figures for each of the main Romano-British sites, which illustrate the breakdown of number of sherds by earliest date. There are obviously issues associated with calculating the date of an assemblage using this method, most significantly the fact that certain dates are commonly used for sherds which can only be broadly dated. For example AD50 is generally used as an earliest date for sherds which can only be dated as Romano-British; therefore there is a greater chance of a bias at this time. On the same note, there are many dates which are unlikely to produce an earliest date, unless a specific vessel type with a tight chronology was identified, e.g. Samian. Therefore some of the peaks and dips noted in these dating figures are perhaps more dramatic than in reality. In order to try and address some of these issues, it was decided that all sherds/group of sherds which could only be very broadly dated (earliest and latest dates being more than 200 years apart) were excluded from the figures. At a future stage of analysis, it might be appropriate to compare data based on earliest date to those if latest date was used instead. Finally each context was allocated a context date, which was based on either the latest dating sherds or else those which formed the bulk of the group. The distribution of Romano-British pottery that dates to pre-150ad, AD and post-250ad has been plotted regardless of the date of the features that it derives from (Figs ). Material from the evaluation has been included and, where feasible, were included with the designated settlement-portion analyses. There were also a number of isolated features, as well as post-roman features, which produced pottery. For the purposes of this report, the material is initially considered as an entire assemblage, followed by individual site, followed by a discussion on each individual sub-site. Due to the sheer number of contexts and features that contained Romano- British pottery, it would be impractical to attempt to discuss them all. Therefore particular features have been selected as case studies for more indepth analysis, for each of the settlements. These features were selected based on the composition of the ceramic assemblages, which stood out in some way as being interesting; sometimes due to the pottery itself, but also because of what the pottery could tell us about these features as a whole. However, for each individual settlement the features have been broadly grouped by assemblage size, to give a better impression of the composition from each site. Small assemblages were those features containing between 1 30 sherds. Medium assemblages comprised those features containing between sherds and large assemblages were groups in excess of 100 sherds. Within North West Cambridge, there were also a small number of features which contained considerable quantities of pottery (1000+ sherds), which are defined as extra-large assemblages. 189

87 19th Century Quarries Archaeological features Pre-150AD Pottery by weight (g) metres Figure Distribution plot of all Romano-British pottery dated to pre-150ad by weight

88 19th Century Quarries Archaeological features AD Pottery by weight (g) metres Figure Distribution plot of all Romano-British pottery dated to AD by weight

89 19th Century Quarries Archaeological features Post 250AD pottery by weight (g) metres Figure Distribution plot of all Romano-British pottery to post 250AD by weight

90 Assemblage Composition As discussed above, pottery was recovered from five main settlements, four of which also had associated cemeteries. The quantity, composition and date of each of these assemblages was varied, and each site is discussed individually below. Table 3.27 shows the quantities of pottery and date range for each of these sites, based on the earliest dates of the sherds. This does mean that the end date of the sites may be somewhat underestimated, and/or in some cases the use of earliest date may give a misleading impression of when a settlement had its origins, as in some cases, the earliest dates may reflect only a small number of sherds. However, no single technique for dating of settlement assemblages is without issue, for example, using context spot-dates would fail to deal with any residual/intrusive sherds. Therefore, earliest date is utilised; however, other datable finds from the site should be used in conjunction with the pottery to provide the most detailed and thus accurate dating. Site No. of sherds Wt (g) EVE (rim and base) Date Range (based on earliest date of sherds) Associated Cemetery Cemetery Date RB AD C5 AD RB.2A LIA AD300+ C3 AD RB.2B LIA AD150 N/A N/A RB.2C LIA AD300+ C2 AD C1 & isolated AD50 RB.2D LIA AD340 cremations 300/AD40 70 Total Table 3.27: Pottery from five main North West Cambridge sites and their associated cemeteries. The remainder of the pottery was derived from isolated features and pre- and post-roman features which did not belong to any of the defined sites, as well as material from the 2009 evaluation which did not correspond to any sites identified in the excavation. Before considering the individual sub-site assemables, it is essential to consider the North West Cambridge assemblage as a whole. Table 3.28 shows the entire assemblage by fabric group and highlights three industries which feature prominently. The largest is Horningsea (2152 sherds, 80.9kg), which is perhaps not surprising given the sites relative proximity to the production centre. A range of fabrics/finishes were associated with this material, comprising the greywares, reduced wares, oxidised wares and the imitation black-burnished wares. Nene Valley wares were also prevalent, totalling 759 sherds weighing 17.9kg. This included colour-coated wares, greywares, whitewares and self-coloured wares. Finally, Hadham (Herts) products were well represented (307 sherds, 4066g), comprising imitation black-burnished wares, the greywares and the red-slipped variety. 193

91 Fabric No. Wt (g) EVE AHFA AMPH BAET BAETL BB1 1 4 BB BLKSL BUFF BUFF SOFT CALC CC CC IMP 3 11 CGBL CGCC CHFOX COLCC CSGW CSMGW CSMRDU CSOX Flint CSGW FSBLK FSBUFF FSGW FSMBLK FSMGW FSMOX FSMRDU FSOX Grey slipped GROG HADBB HADOX HADRDU HADRS HORNBB HORNGW HORNOX HORNRDU HWC Imit BB KOLN LONDONW Metallic GW MOSBLK NVGW 1 5 NVCC NVGW NVSC NVWW OXFRS OXFWS OXFWW OXIS PAKCC PORD Q Q

92 Q Q Q Q Q QG QG RDUS RHOD? RS SAM SAMCG SAMEG SAMSG SHELL SWNWS TR VEG VRW WATT West stow WS WW Total Table 3.28: All Romano-British pottery from North West Cambridge by fabric. Romano-British coarseware fabrics dominate the North West Cambridge assemblage representing 92.3%, which is not surprising for a rural, domestic site. Coarse, sandy fabrics were the most commonly occurring, with sandy greywares in particular featuring highly. Within this category several unsourced coarseware fabrics were identified, many of which were early Romano-British in date. While it is not suggested that sherds from these fabrics were produced at the same source, in the most part they are likely to have been made relatively locally to the site as was the trend during the Romano-British period. A range of sourced coarsewares were identified, with Horningsea products being the most frequent, totalling 2152 sherds (80.2kg), thus representing 9.8% of the entire assemblage. Jars and bowls were the most frequently occurring vessel forms, with large/very large storage jars representing a minimum of 90 different vessels. Coarsewares from further afield included 42 sherds from Wattisfield (Suffolk), 33 Verulamium sherds, twelve Portchester D wares, and single examples from Alice Holt and Highgate Wood. Romano-British finewares accounted for a further 5% of the assemblage, with Nene Valley colour-coated vessels representing the largest single group, with 584 sherds (11.11kg). Pakenham colour-coated wares were also moderately well represented, as were Oxfordshire red-slipped wares, Colchester colourcoated wares and Hadham red-slipped wares. Of interest were 21 early Romano-British wares sherds with red-painted decoration, which are likely to have come from either Cherry Hinton or Foxton (Cambs.). Imported vessels made up the remaining 2.7% of the assemblage, of which Samian was the most commonly occurring, totalling 501 sherds weighing 9059g and representing 2.3% of the overall assemblage (Fig. 3.31). All three production centres were present in varying frequencies. Central Gaulish Samian was represented the largest group, totalling 46% of all Samian, 195

93 followed by East Gaul (36%) and South Gaul (8%). The remaining 10% of the Samian assemblage comprised sherds that could not be sourced. Central Gaulish colour-coated wares and black-slipped wares accounted for a small but significant quantity of all imported wares, totalling 6.3% and 2.9% respectively. Twenty-four amphora sherds were recorded, with Baetican vessels comprising the bulk of the assemblage (Fig. 3.32). Finally four Colonge colour-coated sherds were identified, along with two Moselkeramik blackslipped sherds and one Terra Rubra sherd. A minimum of 2676 different vessels were recovered (based on the number of individual rims and bases). Within this a wide variety of vessel forms were identified within this assemblage, albeit in varying quantities. Figure 3.33 shows all diagnostic sherds by main vessel group and highlights that jars dominated the assemblage, followed some way behind by beakers, flagons, bowls and dishes. It is of note that neither mortaria or amphora sherds featured highly within this assemblage (Figs. 3.32, Overall, with the exception of the cemetery material, the North West Cambridge repertoire is indicative of a rural, domestic site, with wares used for the storage, preparation and serving of foodstuffs and drinks. A minimum of 441 vessels were noted as being decorated (2227 sherds 47.6kg), of which burnishing, combing/rilling, rouletting and cordons were the most commonly occurring decorational techniques. A further 356 vessels were noted as having use-wear evidence, of which exterior sooting was the most prevalent, followed by interior limescale and burnt residue. 13 different sherds were noted as being trimmed bases and were distributed across the five main Romano-British settlements, which reflect secondary use/modification of vessels from their original uses/intended uses. Five were recovered from RB.2D, including two from a single Feature, well F.564. Four more trimmed bases were collected from RB.2C, with two more from RB.2A and single examples from both RB.2B and RB.1. These bases occurred in a range of fabric types including imitation black-burnished wares, greywares and oxidized wares, as well as three Samian examples. A variety of vessel forms were utilized including bowls, dishes, jars, cups and a beaker, with the trimmed bases ranging in diameter from 2cm at the smallest to the largest at 14cm. The range of sizes selected for modification suggests that these sherds had a variety of uses/intended uses once they had been trimmed. Commonly, smaller trimmed bases were thought to have been used as gaming pieces. The function(s) of the larger trimmed bases is somewhat uncertain. Given the size of the overall assemblage, the number of trimmed bases identified is not necessarily significant. Sixteen different vessels had post-firing holes in the bases, suggesting that they had been modified for use as sieves/strainers. Settlement Summaries Settlement RB.1 - The assemblage from RB.1 totalled 2008 sherds weighing 35.5kg and representing EVEs, thus representing 9% of the total North West Cambridge Romano-British assemblage. Pottery was recovered from 44 different features, the majority of which (29) contained small assemblages of between 1 30 sherds, with eleven features containing medium-sized 196

94 assemblages (31 99 sherds) and four which contained large pottery assemblages, in excess of 100 sherds This includes a quantity of material from the associated cemetery (C5) which represented 13% of the total RB.1 assemblage. Table 3.29 shows the quantity of pottery by feature type and highlights that 69% of the assemblage derived from ditches. Feature Type No. Wt (g) EVE Burial Ditch Pit (quarry) Posthole 1 6 Well Other Total Table 3.29: RB.1 Romano-British pottery quantities by feature type. The settlement can be broadly divided into three separate areas: RB.1 East, RB.1 West and C5. The pottery assemblage from RB.1 East totalled 70 sherds weighing 1075g, most of which was dated as early Romano-British (AD43 120). The majority of the pottery however, was derived from RB.1 West, accounting for 1543 sherds weighing 29.3kg. Although there were eight earlier Romano-British contexts identified, the majority of contexts dated to the mid/later Romano-British period. The difference between the two sites is of interest, with the pottery demonstrating that RB.1 East was earlier in date and shorter-lived than RB.1 West, although there does appear to have been a period when the two areas of the settlement were occupied at the same time. It is of interest that C5 was earlier Romano-British in date and therefore was contemporary with activity on RB.1 East and the earliest phases of RB.1 West, spatially C5 is more likely to have been associated with RB.1 West. Plotting the pottery from RB.1 by its earliest date (Fig. 3.36) shows a peak in activity between AD , with a smaller peak between AD Evidence of activity post AD200 was very limited and suggests a definite decline at this time. A range of fabric types were identified (Table 3.30), with coarse sandy greywares accounting for the largest fabric group, representing 47% of the total assemblage. Overall, Romano-British coarseware fabrics dominated the assemblage, representing 93.2%, with Roman-British finewares totalling 4.5% and imported wares accounting for the final 2.3%. Although most of the coarseware vessels are unsourced, it is likely that the majority would have been produced within the local area. Samian wares were the most commonly occurring imported ware, with all three regions represented. Overall however, Samian represented just 1.4% of the total RB.1 assemblage. Other imported wares included nine amphora sherds (six Rhodian in all likelihood representing a single amphorae given their spatial proximity, two Baetican, one unsourced), and one Cologne colour-coated ware. 197

95 19th Century Quarries Archaeological features Samian Ware by weight (g) metres Figure Distribution plot of all Romano-British Samian ware by weight

96 19th Century Quarries Archaeological features Amphorae by weight (g) metres Figure Distribution plot of all Romano-British amphorae by weight

97 19th Century Quarries Archaeological features Roman pottery 0 metres 100 Figure Romano-British pottery from all sites and cemeteries by main vessel forms

98 19th Century Quarries Archaeological features Mortaria by weight (g) metres Figure Distribution plot of all Romano-British mortaria by weight

99 Figure 3.35a Romano-British pottery (all sites and cemeteries) by main vessel forms Figure 3.35b All Romano-British pottery by earliest date from RB1

100 Figure 3.36 All settlement RB2 Romano-British pottery by earliest date (sites A D) Figure 3.37 All Late Iron Age and Romano-British pottery by earliest date from RB2A

101 Fabric No. Wt (g) AMPH 1 37 BAET 1 81 BAETL BLKSL CC CSGW FOX/CH 2 6 FSGW FSMOX 4 34 FSMRDU 1 7 HORNBB HORNGW HORNOX IMITBB 4 79 KOLN 1 83 NVCC NVGW 1 10 NVWW OXIS PAKCC Q Q Q Q4 1 7 QG QG RDUS RHOD? RS 4 63 SAM 3 7 SAMCG SAMEG SAMSG 5 24 SHELL WW Table 3.30: All Roman pottery from settlement RB.1 by fabric. The assemblage was domestic in nature, with a range of vessels used for the storage, preparation and serving of foodstuffs (Table 3.31). Jars were the most commonly occurring vessel form, as is expected from a Romano-British rural domestic settlement, occurring in a range of sizes and forms. Dishes, beakers and bowls were also well represented. A possible crucible was recovered from ditch F.2514, fill [9195]. 203

102 Form No. Wt (g) EVE Amphora Beaker Bowl Closed Crucible? Cup Dish Flagon Jar Lid Mortaria Open 7 85 Unknown Table 3.31: All Romano-British pottery from RB.1 by vessel form. Settlement RB.1 Feature Case Studies - Two ditches in particular stand out as containing large assemblage of material. F.2513 contained the largest assemblage of pottery from a single feature, totalling 700 sherds weighing 12.7kg and representing EVEs, thus representing 35% of all RB.1 pottery. The pottery dated AD and was derived from two contexts; [8044] which contained fourteen sherds (186g) and [9174] which contained 686 sherds of pottery weighing 12.5kg. The pottery from Ditch F.2513 was indicative of a domestic rubbish dump, with a large number of coarseware vessels, with sandy greywares and Horningsea greywares well represented. The mean weight of pottery from this feature was moderately high at 18.1g, though this was influenced by a relatively large quantity of Horningsea greyware storage jar sherds. That said, the assemblage was not noted as being particularly badly abraded, suggesting material had not moved far from where it had been broken. Overall, the quantity of pottery recovered from this feature suggests that it became one of the primary places for rubbish disposal within RB.1. That the majority of the sherds were from a single context indicates that the backfilling took place within a relatively short period of time. A second ditch, F.2514 contained 461 sherds in total, weighing 8380g and representing 7.78 EVEs, deriving from four different contexts. Context [9195] contained the bulk of the pottery, representing 82% of all pottery from this feature, totalling 379 sherds weighing 6306g. Context [8046] contained 60 sherds (1568g), with context [8047] totalling twelve sherds (347g) and [9138] contained ten sherds in total (159g). The material from all of the contexts was broadly the same date, with a date range of AD for the entire feature, thus making it contemporary with that from F The mean weight of the pottery from this feature was 18.2g and comprised a range of jars, bowls and dishes indicative of a domestic repertoire. This feature was not only contemporary with F.2513, but was also very similar in character, thus suggesting very similar depositional histories. There are also some particular similarities, notably the presence of sherds of Rhodian amphorae in both ditches, which may well derive from a single vessel. F.2768, a well, contained a total of 102 sherds of pottery, weighing 2865g and representing 8.90 EVEs. At this point it is worth making several general observations concerning the pottery from wells, which applies to this well and those that will be discussed subsequently. The first is that the pottery relates to the construction, use and backfilling of a well. These three stages often represent a considerable period and in many instances it was impossible to completely differentiate the pottery from the three stages. In most instances at North West Cambridge the bulk of the pottery relates to the construction of the well, with little material related to use and some material in the backfill of the well shaft. Another factor is that because of their relative size and depth plus the presence of significant material in their bases wells were frequently more intensively investigated than other features. This means that they will tend to produce large assemblages of pottery. Whilst many of the large pottery assemblages from other features consisted of concentrations of material that were identified 204

103 during excavation in most cases the material associated with wells was dispersed throughout the fills of the feature. The material was recovered from ten contexts and had a MNV (minimum number of vessels) of 33, based on just the rim and base sherds. The contexts contained varying quantities of material, though there was no primary dump of material, with all contexts containing between one and 23 sherds. There was also little obvious difference in terms of date between the different contexts, with a range of AD for nearly all of the fills. Context [9190] contained thirteen sherds weighing 778g which comprised large freshly broken sherds, including a semi-complete indented beaker, sherds from a Central Gaulish Samian Dr27 cup and an East Gaulish Dr31 dish. The pottery from Well F.2768 was domestic in nature, though included sixteen sherds of Samian (348g), which represented a maximum of fourteen different vessels and represented 16% of the entire assemblage from this feature, which is somewhat higher than the site average of 2.3%. The similarity in date between the different fills is indicative of backfilling of the feature taking place within a relatively short period of time, or else may represented material which had been re-deposited from elsewhere. That said this seems unlikely in the case of context [9190], this material relates to the backfill of the well shaft and parts of a colour-coated beaker appear to be deliberately deposited. It is intriguing to note that all the pottery is earlier than the backfilling, as a leather shoe dates to c This phenomenon of pottery being earlier than other dating evidence including coins occurs on a number of occasions. The quantity of pottery recovered implies that this feature was not a primary rubbish disposal area. Settlement RB.2 - This represented the largest assemblage from North West Cambridge, totalling sherds weighing 318.6kg and representing EVEs, thus representing 85% of all Romano-British pottery recovered. This site has been subdivided into four different sites (A D), which are considered here as single assemblages, although a discussion on RB.2 as a whole is included below. A breakdown of each site by quantity of pottery is provided in Table 3.32, and shows all four sites produced sizable assemblages, with RB.2D producing the largest assemblage overall. Occupation at this settlement appears to have been continuous from the Late Iron Age to the later Romano- British period. The pottery evidence implies that there were shifts in the focus of this settlement; with different sub-sites seeing peaks in activity at different times. That said, this area of the site as a whole clearly saw activity throughout the duration of the Romano-British period, as demonstrated in Figure Site No. of sherds Wt (g) EVE (rim and base) Date Range (based on earliest date of sherds) RB.2A LIA AD250 RB.2B LIA AD150 RB.2C LIA AD300 RB.2D LIA AD340 Table 3.32: All RB.2 sites by quantity and date. A relatively large number of sherds were recovered from later infilling deposits. In total sherds weighing 24.6kg were collected, representing EVEs, from 49 different features. This accounted for 5.5% of the complete North West Cambridge assemblage. The material had a mean weight of 20.3g. These deposits were all identified at RB.2, occurring on all four sub-sites in varying quantities. RB.2A totalled 94 sherds (2403g), 116 sherds weighing 2279g were recovered from RB.2B. The majority of sherds from later infilling deposits occurred on RB.2C, totalling 808 sherds weighing 16040g. Finally 196 sherds (3874g) were recovered from RB.2D. The two largest later infilling assemblages were both from RB.2C, comprising

104 sherds (7626g) from later infilling F.1167 and a further 165 sherds (12541g) from Well F The former included some 4th 5th century contexts and included a Horningsea greyware jar with an X Scratched into the side, as well as a black-slipped jar with two post firing holes in the side of the vessel. The later infilling deposits contained primarily later Romano-British material (3rd 5th century AD), with the exception of six contexts dated as early Roman. However, these contexts accounted for only fifteen sherds weighing 346g. Settlement RB.2A - A total of 4559 sherds of pottery weighing 78.8kg and representing EVEs were recovered from RB.2A. This therefore represented 24.4% of all pottery from RB.2 and 20% of the entire North West Cambridge assemblage. Within this group 17% of the pottery was derived from the associated cemetery C3 (see below), with the remaining 83% coming from settlement features. This totalled 3780 sherds weighing 73.4kg (98.51 EVEs), recovered from 127 features, the majority of which (106) contained small assemblages of pottery (0 30 sherds), with thirteen containing mediumsized groups (31 99) and eight containing large assemblages (100+). 39% of the pottery was recovered from ditches, 15.8% from quarry pits and 11% from wells, with smaller quantities coming from pits, postholes and tree throws (Table 3.33). Feature Type No. Wt (g) EVE Burial Ditch Pit 4 58 Pit (Quarry) Posthole Tree throw Well Other Total Table 3.33: All Romano-British pottery from RB.2A by feature type. The pottery indicates that this site had its origins in the Late Iron Age. Figure 3.37 which shows sherds by earliest date is of interest, demonstrating that the largest peak was in the early Romano-British period around AD43. However, these early Romano-British sherds frequently occurred alongside later dating sherds, being residual, and a likely result of the re-cutting and consequent truncation of earlier features. Therefore, the early component of this assemblage may be overlooked if just considering overall context dates and would not necessarily show the true extent of early Romano-British activity at RB.2A. There was a second peak visible at AD200, with a third smaller peak at AD100. The figure also shows occupation continuing into the 3rd 4th century AD. It appears that activity was likely to have been continuous, albeit fluctuating in intensity. The latest dating features comprised ditches and wells, with nine ditches dating to AD

105 Fabric No. Wt (g) AHFA 1 48 BLKSL BUFF CALC 1 6 CC CGBL 5 35 CGCC COLCC 4 67 CSGW CSOX FSBLK 9 99 FSGW FSMGW FSMOX 2 19 FSMRDU 1 27 FSOX 1 31 GREYSLIP 3 62 GROG HADBB 2 40 HADOX HADRDU HADRS HORNBB HORNGW HORNOX 3 87 HWC 1 2 IMITBB KOLN 2 9 NVCC NVGW 3 42 NVSC NVWW OXFRS OXFWS 1 12 OXFWW OXIS PAKCC PORD 2 33 Q Q Q Q Q

106 Q QG RDUS RS 3 86 SAM 3 29 SAMCG SAMEG SAMSG 2 32 SHELL SWNWS 1 13 VRW WATT 6 98 WS WW Table 3.34: All Romano-British pottery by fabric from RB.2A. Fifty-seven different fabric types were identified from this assemblage (Table 3.34), with Romano-British coarsewares being the most commonly occurring, representing 96% of the pottery. Romano-British finewares accounted for 3%, with imported wares making up the remaining 1%. The largest individual fabric group were the coarse sandy greywares, totalling 1141 sherds weighing 16919g. Horningsea products and shell-tempered wares were also well represented along with unsourced black-slipped wares. The earlier Romano- British sandy fabrics Q1, Q3 and Q4 combined, accounted for 11% of the total pottery. A variety of vessel forms were identified from this site (Table 3.35). Unsurprisingly, jars were the most commonly occurring vessel form, representing 20% of all sherds and 44% of all diagnostic sherds. Beakers also featured highly at this site, accounting for 40% of all diagnostic sherds (18% of all RB.2A pottery). Bowls and dishes were less common, though still accounted for significant quantities of the diagnostic sherds. Form No. Wt (g) EVE Beaker Bowl Closed Cup 1 12 Dish Flagon Flask 1 84 Jar Lid Mortaria Open Sieve 1 15 Unknown Table 3.35: All sherds by vessel form from RB.2A (settlement and C3). 208

107 Figure 3.38 Main vessel forms for settlement versus cemetery assemblage from RB2A Figure 3.39 All Late Iron Age and Romano-British pottery by earliest date from RB2B

108 This data includes the pottery recovered from associated cemetery C3; therefore, the frequency of certain vessel types is somewhat misleading. Figure 3.38 shows a comparison between the main vessel form types for settlement versus cemetery. The most striking element is the very high frequency of beakers occurring within the cemetery assemblage, representing 80% of all cemetery pottery, compared to just 12% of settlement material (a more detailed discussion of the pottery from C3 is below, in the cemeteries section). It is also noteworthy, though not unexpected, that bowls, dishes, lids and mortaria occur exclusively within the settlement assemblage. Settlement RB.2A Feature Case Studies: Ditch F.1901 comprised the largest feature assemblage from RB.2A, totalling 492 sherds, weighing 5956g and totalling 6.30 EVEs. The pottery had a relatively low mean weight of 12.1g and was recovered from three contexts, two of which contained only small quantities of material; [5748] totalled eight sherds (207g), while [7483] contained three sherds weighing 46g. The majority of the pottery was collected from context [5948], totalling 481 sherds weighing 5703g. A minimum of 42 different vessels were identified, including twelve beakers, sixteen jars and thirteen closed vessels. This difference is partly due to the fact that the larger group comes from a corner of a ditch where a wider slot was excavated, nonetheless there was a greater density of material as was noted during excavation. There was no clear difference in date between the different contexts, with all being mid/late-1st century AD in date, thus this ditch represents one of the earliest features on the site. Three wells within RB.2A contained significant quantities of pottery. F.2023 contained 118 sherds, weighing 4411g (10.81 EVEs, with a high mean weight of 37.4g, recovered from six different contexts. Fill [6083] contained 24 sherds (638g) which included an unusual Hadham reduced ware bowl with embossed dot decoration, as well as four straight-sided dishes. Context [6084] comprised 56 sherds weighing 2055g (47% of F.2023), which comprised lots of fresh sherds, resulting in the high mean weight of 36.7g. This included two beakers, five bowls, eleven dishes and five jars with Nene Valley colour-coated sherds, Hadham wares and shell-tempered wares commonly occurring. Context [7462] contained nineteen sherds (1005g) with a high mean weight of 52.9g, including many fresh sherds, although no complete or semi-complete vessels were identified. There was little to distinguish the six contexts in terms of date from one another, with all dating to the later Romano-British period, c. AD , although [6084] which is stratigraphically late in the fill sequence could be more specifically dated AD Well F.2044 contained 173 sherds weighing 6359g and totalling 5.67 EVEs, recovered from six contexts with a mean weight of 36.8g. Context [7111] contained the largest quantity of material totalling 75 sherds (2228g) which represented a minimum of eighteen different vessels. This included nine jars, two bowls and four dishes, with shell-tempered wares, Horningsea greywares and Hadham wares all well represented. A further 35 sherds (1627g) were collected from [7113], with similar fabrics and forms represented. The material from all contexts was later Romano-British in date (AD ), with [7112] which is stratigraphically late in the fill sequence more closely dateable to AD Therefore as was the case with Well F.2023, the pottery from this feature suggests a fairly rapid backfilling of the well once it had gone out of use. It is of note that all of the contexts had a high mean weight and comprised fresh unabraded sherds. That said there was no evidence of any specially placed deposits within these features. Box-lined well F.2056 contained a fairly large assemblage totalling 212 sherds weighing 7471g and representing 8.03 EVEs, with a mean weight of 34.2g. The pottery was recovered from eleven different contexts. Fill [7163] which is stratigraphically late in the fill sequence contained 58 sherds (1882g) dating AD Context [7446] which relates to the initial backfilling of the well shaft contained sixteen sherds weighing 1253g, eleven of which (989g) were from a single shell-tempered jar which was semi-complete when refitted. Interestingly, all of the sherds from this context were heavily sooted, suggesting that they had been used over a fire. Other sherds of note include a Hadham oxidised bowl base from [7443] which had been trimmed and burnt, and an unusual Hadham reduced ware flagon with a rippled neck from [7280] As with the two previous wells, the contexts from this feature were 210

109 later Romano-British in date, dating 3rd 4th century AD, with context [7295] which is stratigraphically late in the fill sequence dating AD The three wells discussed above have very comparable assemblages, not least because they all contained fairly large assemblages of pottery. The pottery evidence indicates that by the mid- 3rd century AD all three features had stopped functioning as wells and instead were utilised for the disposal of domestic waste. The individual deposits suggesting that once they started being used for rubbish disposal, they were backfilled within a relatively short period. That significant quantities of pottery from all three features comprised large fresh sherds is of note and suggests material had not travelled far post-breakage. Quarry pit F.1936 contained 140 sherds of pottery weighing 4500g and totalling 5.18 EVEs, all of which came from a single context, [5826]. The assemblage recovered had a high mean weight of 32.1g and was dated AD , although there were some earlier dating sherds suggesting re-deposition or truncation of earlier features. The sherds comprised primarily non-diagnostic coarsewares, although there were a small number of finewares including an East Gaulish vessel with a clipped base. A minimum of nineteen different vessels were represented within this feature. The supply, consumption and deposition of pottery from RB.2A highlight the changing function/status of this particular settlement. For example, it is of note that the associated cemetery was only utilised in the early Romano-British period, and not used during the subsequent periods. Also the in-filling of the wells discussed above is of interest, with these three features having very similar histories in terms of them going out of use at around the same time, which in turn has interesting implications for understanding how this particular site changed and evolved overtime. Local wares dominated, with Horningsea products particularly well represented. That said the presence of non-local wares in the later component of the assemblage (post-ad200), including Oxfordshire wares and an example of an Alice Holt vessel is of interest and demonstrates that the site had access to goods from outside of the immediate local area. These elements of the assemblage are not however considered as representing high status activity, with the pottery indicative of a jar dominated, domestic assemblage. Settlement RB.2B - The pottery assemblage from RB.2B totalled 3048 sherds, weighing 44kg (mean weight 14.5g) and represented EVEs, thus equating to 16.3% of all RB.2 pottery and 13.9% of the entire North West Cambridge assemblage. The material was collected from 80 different features, the majority of which contained small assemblages (61 features) with twelve features containing medium-sized assemblages and seven which contained large groups totalling more than 100 sherds. 75% of the pottery derived from ditches (Table 3.36), which also yielded all of the large pottery assemblages. 18% of the pottery was collected from quarry pits, with the remaining 7% coming from pits, postholes and soil. Feature Type No. Wt (g) EVE Ditch Pit (non-quarry) Pit (quarry) Posthole 3 13 Soil Evaluation TOTAL Table 3.36: All Romano-British pottery from RB.2B by feature type. 211

110 Settlement RB.2B comprised one of the earlier sites at North West Cambridge, with the pottery indicating that the site had its origins in the Late Iron Age (Fig. 3.39). Four features were dated Late Iron Age/early Roman, with a suggested date range of 50BC AD50. The majority of features were, however, early Romano-British in date (mid-1st late-1st/early-2nd century AD). A further seven contexts were dated mid-2nd century AD+, comprising four quarry pits and two ditches. There was very little evidence of activity after the mid/later 2nd century AD. There was no associated cemetery with RB.2B, although occupation at this site was contemporary with the earliest three cemeteries in North West Cambridge (Cemeteries 3, 5 and 7). The majority of the assemblage comprised coarseware vessels (98%), most of which are likely to have been locally made. Romano-British finewares accounted for 1%; vessels of note include ten sherds (53g) from a fineware painted beaker dating to the mid-1st century AD, with affinities with the Cherry Hinton (Evans 1990) and Foxton kilns in South Cambridgeshire. A total of 31 imported sherds were identified, which represented just 1% of the RB.2B assemblage. However, this low figure is not unexpected given the early date at which this site peaked, and its rural location. Imported wares identified comprised two unsourced amphora sherds and 29 Samian sherds, with the South and Central Gaul production centres well represented. Seven East Gaulish sherds were also identified (Table 3.37). Fabric No. Wt (g) AMPH 2 46 BLKSL BUFF CALC CC 1 2 CH/FOX CSGW CSOX 1 61 CSRDRU FOX OX FSBLK FSGW FSMBLK FSMGW FSMOX 2 12 FSMRDU 3 10 FSOX 8 24 FSRDU 1 11 GROG GSLIP 3 14 HORNBB

111 HORNGW HORNOX 1 31 IMITBB 3 90 NVCC 1 4 OX SHELL 1 10 OXIS Q Q Q Q Q Q Q QG QS RDUS SAMCG SAMEG 7 74 SAMSG SHELL VEG 1 40 VRW WATT 5 42 WSOX 2 52 WW Table 3.37: All pottery by fabric from RB.2B. The range of vessel forms identified is detailed below in Table As with all of the North West Cambridge sites, jars dominated the assemblage, with a minimum of 230 different vessels identified. These occurred in a wide range of sizes and therefore functions. The smallest vessels had rim diameters of 8cm, with the largest jars comprising rims measuring 40cm. The latter comprised storage jars, most of which were from the Horningsea kilns. Other vessel forms were less well represented, with beakers occurring in what initially appears to be a sizable number (78 sherds, 343g). However, this figure equated to a minimum of just six vessels. Bowls and dishes were very poorly represented at this site, which is probably due to the relatively early date of the site, before these particular vessel forms had become commonplace. 213

112 Form No. Wt (g) EVE Amphora 2 46 Beaker Bowl Closed Cup Dish Flagon Jar Lid Mortaria Open Platter? 1 57 Unknown Table 3.38: All sherds by vessel form from RB.2B. The condition of the pottery from RB.2B was somewhat mixed, with a relatively low mean weight 14.5g. It is therefore perhaps not surprising that such a high proportion of sherds could not be accurately categorised by form. This is highlighted by the presence of a very high number of generic closed vessels, which accounted for 33% of the entire assemblage, totalling 1006 sherds. This factor may therefore be masking the true character of the pottery in terms of vessel forms. Settlement RB.2B Feature Case Studies: The largest single assemblage derived from Ditch F.1810, totalling 576 sherds weighing 10583g and representing EVEs, thus representing 20% of all pottery from RB.2B features. The material was recovered from just two contexts. Fill [5408] contained 81 sherds (1050g), which included a semi-complete South Gaulish Samian Dr33 cup, dating AD50 100, which had a complete stamp reading II.TAIUSSAD. The bulk of the pottery came from fill [5410], totalling 495 sherds (9533g), with a minimum of 45 different vessels recovered. The pottery dated AD and included a butt-beaker and 41 different jars, four of which were partially complete when refitted. Ditch F.1697 contained 346 sherds weighing 4491g and representing 5.38 EVEs, with a mean weight of 13g. The pottery was collected from eleven different contexts, of these fill [5573] was the earliest, with a date of AD This comprised a single vessel; a partially complete Q3 small jar with a short neck and rounded rim, with rilled decoration. Fill [5109] contained 60 sherds weighing 730g, which dated AD50 100, and included two handmade sherds in the Late Iron Age tradition. With the exception of a small sherd from a South Gaulish Samian Dr33 cup, all of the other fabrics identified within this context comprised coarsewares, which are likely to have been made in the local area. A group of material from ditch F.1688 [5115] comprised largely of fine sandy micaceous black-slipped jar which appeared to represent a single vessel (166 sherds, 531g). This therefore accounted for a significant proportion of the overall assemblage from the site (186 sherds, 745g) and demonstrates how a single well-preserved specific depositional event can potentially have a significant influence on the overall representation of a ware. 214

113 Settlement RB.2C - In total, 5136 sherds of pottery weighing 99.6kg and representing EVEs were recovered from RB.2C. Of this 4476 sherds (90.7kg, EVEs) were collected from settlement associated features, with the remaining material deriving from the associated cemetery (C2). The pottery from the settlement therefore represented 27.5% of all RB.2 pottery and 23.4% of the entire North West Cambridge assemblage, thus making it very similar in size to RB.2A. Material was recovered from a total of 184 different features; 154 of which contained small assemblages of pottery, 22 contained medium-sized assemblages and eight contained large groups of pottery. The pottery came from a variety of different feature types (Table 3.39), with ditch accounting for the largest quantities, although a significant quantity of material was also derived from wells. Feature Type No. Wt (g) EVE Beam slot Burial Ditch Later infilling Pit (non quarry) Pit (quarry) Posthole Soil (occupation) Soil (subsoil) Tree throw Well Evaluation Total Table 3.39: All Romano-British pottery from RB.2C by feature type. RB.2C appears to have had its origins in the Late Iron Age/early Romano- British period, with subsequent occupation spanning the entire Romano- British period, until the 4th century AD at least, thus making this one of the longest lived settlements at North West Cambridge. Plotting of the sherds by earliest date (Fig. 3.40) shows several peaks visible, the highest being at AD200, closely followed by a second peak at AD150. Two smaller peaks are also recorded at AD100 and AD50, with a final small, yet significant peak at AD250. Although a small number of 1st century AD contexts were identified, the quantity of pottery recovered from these was negligible, with much of the material of this date occurring residually in later features, which is a likely result of truncation due to the cutting and re-cutting of features. The thus suggesting that activity in the early Romano-British period was limited; RB.2C being somewhat peripheral during this time. RB.2B was associated with C2, one of the latest dating within North West Cambridge. A total of six graves contained grave good vessels, all of which comprised a single colour-coated vessel, with a date range of AD

114 Figure 3.40 All Late Iron Age and Romano-British pottery by earliest date from RB2C Figure 3.41 All Late Iron Age and Romano-British pottery by earliest date from RB2D

115 Fabric No. Wt (g) AMPH BAET BAETL BB BLKSL BUFF CC CGBLK CGCC? COLCC CSGW CSMGW CSOX CSRDU CSRS 1 20 FSBB 1 29 FSBUFF FSGW FSMBLK FSMGW FSMOX FSMRDU 1 30 FSOX FSWW 1 12 Grey slipped 8 56 GROG 4 72 GS1 1 8 HADBB HADOX HADRDU HADRS 6 51 HORNBB HORNGW HORNOX IMITBB KOLN 1 11 Metallic GW MOSBLK 2 5 NCGW 1 5 NVCC NVGW NVSC

116 NVW 1 61 NVWW OXFRS OXFWW 2 31 OXIS PAKCC PORD/ 1 7 Q Q2 1 8 Q Q RS SAM SAMCG SAMEG SAMSG 3 23 SHELL TR? 1 14 VFSGW VRW 4 60 WATT WS WW Table 3.40: All pottery by fabric from RB.2C (including C2). As expected from an assemblage of this size, a wide variety of vessel fabrics were identified (Table 3.40) Romano-British coarsewares represented 87% of all sherds, which although still considered to dominate the assemblage, also represents the lowest percentage of coarsewares overall, when compared to the other RB.2 sites as well as RB.1. Romano-British finewares totalled 10% of the pottery (excluding material from the cemetery), while imported wares accounted for 3% of the assemblage, thus representing the largest percentages at any site as well as compared to the North West Cambridge material as a whole. The division of pottery into the three groups has interesting implications for the interpretation of RB.2C, in terms of function and status, compared to the other settlements. The longevity of this particular site is one explanation; however, this is perhaps too simplistic. The higher frequencies of finewares and imported wares instead are evidence that this particular settlement had more wealth, access and/or desire for fineware vessels. While these percentages are not suggestive of a high status settlement or villa etc. per se, they do indicate a somewhat elevated status compared to the other settlements. In terms of individual fabric types, coarse sandy greywares were the most commonly occurring fabric. Other moderately well represented fabrics included shell-tempered wares, black-slipped wares, Horningsea wares and Nene Valley colour-coated wares. Jars dominated the assemblage in terms of vessel form (Table 3.41), although beakers, bowls and dishes were all well 218

117 represented. The high number of beaker sherds is affected by the material derived from associated C2, which included 105 sherds (693g) from beakers. Overall, beakers dominated the cemetery assemblage, representing a minimum of nine vessels. RB.2C represented the largest number of mortaria sherds from a single site, totalling 48 sherds, weighing 2356g and representing a minimum of 20 different vessels. Form No. Wt (g) EVE Amphora Beaker Bowl Closed Cup Dish Flagon Jar Lid Mortaria Open Unknown Table 3.41: All sherds by form from RB.2C (including C2). Settlement RB.2C Feature Case Studies sherds weighing 6052g and representing 7.12 EVEs were recovered from Well F.1236, from four contexts. Context [3599] which is stratigraphically late in the fill sequence contained the largest number of sherds totalling 243 (2904g) equating to a minimum of 61 different vessels, including Nene Valley colourcoated sherds, Hadham oxidised wares, Horningsea greywares and shell-tempered wares, giving a context date of AD , although there were also some earlier elements dating AD sherds (1319g) were collected from fill [3712], dating AD A minimum of thirteen different vessels were identified, including two which had evidence of exterior sooting indicative of being used over a fire. The remaining two fills [3721] and [3848] were also dated AD A second well F.1020 contained 183 sherds of pottery weighing 9249g (8.29 EVEs) and representing a minimum of 42 different vessels. The material was recovered from seven contexts in varying quantities and had a high mean weight of 50.54g, although this was in part, due to the presence of a sherd from a shell-tempered jar with a very large beaded rim, with a 36cm diameter, which weighed 937g. Fill [2922] contained 63 sherds (3499g) dating AD , which included an interesting Nene Valley colour-coated sherd with embossed X decoration and a metallic slip. Sherds from the same vessel, or at least a very similar vessel, were recovered from fill [3099] which is stratigraphically and physically adjacent to [2922], as well as a Nene Valley self-colour imitation Dr37 bowl with red painted decoration, dating this context to AD Features F.1020 and F.1236 therefore follow a pattern identified within several other well assemblages, (including the three from RB.2A, see above) in terms of the composition of material and also in the fill sequences. All of the fills containing pottery within these wells were seemingly deposited within a relatively short period of time, after the wells had gone out of use. Interestingly at both Sites RB.2A and RB.2C this occurred sometime after AD250. The quantity of the pottery demonstrates that these two features and the other wells discussed had become key features for the disposal of domestic waste in the later Romano- British period, while the condition of many of the sherds, suggests that the material is unlikely to have travelled far from its place of breakage. 219

118 A group from ditch F.1076 [3086] (223 sherds weighing 3181g) was strongly dominated by Horningsea ware (196 sherds, 2903g); where identifiable this material was all from jars. This was identified as a significant deposit during excavation; however, given the size of many Horningsea ware jars it may well represent only a few vessels. Settlement RB.2D - A large assemblage of pottery was recovered from RB.2D, totalling 5949 sherds weighing 96.1kg and representing EVEs. This equated to 27.1% of the entire North West Cambridge assemblage and 31.8% of all RB.2 pottery, thus representing the largest single site, in terms of pottery, identified. The majority of the assemblage (71.4%) derived from settlement features (4249 sherds, 77.9kg, EVEs), with the remaining 28.6% (1700 sherds, 18.3kg, EVEs) coming from associated Cemeteries 1 & 7 (see below). Pottery was recovered from 121 different features, with one (F.586) containing an extra-large assemblage (+100 sherds), sixteen containing large assemblages, fourteen containing medium-sized assemblages and 90 features containing small assemblages. The pottery was derived from a variety of different features (Table 3.42). Feature Type No. Wt (g) EVE Beam slot Burial Ditch Pit (non quarry) Pit (quarry) Posthole Tree throw 1 14 Well Evaluation Total Table 3.42: All Romano-British pottery from RB.2D by feature type. The pottery suggests that the earliest phases of activity occurred in the Late Iron Age/early Romano-British period, although this appears to have been somewhat limited before AD50 (Fig. 3.41). However, there was a small peak at AD50, before the two largest peaks visible at AD100 and AD150, demonstrating the site peaked in the 2nd century AD; there was a final peak visible at AD200. After the beginning of the 3rd century AD, activity at this site appears to have gone into decline, with only a small quantity of sherds which post-dated the mid-3rd century AD. This therefore suggests that the by the mid-3rd century AD the focus of the settlement had shifted, perhaps relocating elsewhere. This view is supported by the ceramic evidence recovered from two of the wells on the site (F.548 and F.586, discussed below), which had begun to be backfilled by the later 2nd century AD. A total of nineteen contexts were dated to the earlier Romano-British period (AD43 150), comprising primarily ditches. A further nineteen contexts were dated to the later Romano-British period (3rd 4th century AD), which again comprised primarily ditch fills. 220

119 RB.2D had one associated cemetery, plus some isolated cremations; C1, which appears to have been in use throughout the occupation of the settlement, with the earliest graves dating AD and the latest dating AD (although there are grounds to believe that the cemetery went out of use considerably earlier than AD300). Two isolated cremations within settlement RB.2D were early Romano-British in date, with the pottery suggesting a date range of AD However, this comprised material from just two graves. A total of 63 different fabric groups were identified (Table 3.43). As expected, the coarse sandy greyware group dominated the assemblage. Horningsea products were also well represented, as well as early Romano-British fabric Q1. Overall Romano-British coarsewares accounted for 93% of the pottery, with finewares representing 4% and imported wares totalling the remaining 3%. Fabric No. Wt (g) AMPH BB1 1 4 BB BLKSL BUFF CC CGBLK 1 2 COLCC CSBLK 1 85 CSGW CSMGW CSOX CSRDU FMBLK 1 7 FMSGW 1 7 FS Buff FSBLK 6 22 FSBUFF 8 41 FSGW FSMBLK FSMGW FSMOX FSMRDU FSOX 5 44 FSRDU FWW 1 8 GS GW HADBB HADOX

120 HADRDU 3 37 HADRS 1 16 HORNBB HORNGW HORNOX IMIT BB LONFRW 2 3 MOXI 2 11 NVCC NVCC NVGW NVSC 2 42 NVWW OXFWW OXIS PAKCC 1 18 PORD Q Q Q Q RDUS SAM SAMCG SAMEG SAMSG SHELL VCSGW VRW 3 30 WATT WS WSFRDU WW Table 3.43: All pottery by fabric from RB.2D (including Cemeteries 1 & 7). RB.2D comprised the most diverse range of vessel forms from any of the individual settlements (Table 3.44). That said, the ratio of different vessel forms was consistent with the site wide pattern, with jars being the most commonly occurring vessel form, with beakers, bowls and dishes also featuring highly. The number of flagon sheds stands out within this assemblage; however, this is largely influenced by the pottery recovered from the two associated cemeteries. Figure 3.42 shows the comparison of key vessel forms for RB.2D settlement versus Cemeteries 1 & 7 pottery and demonstrates the high ratio of flagons recovered from the cemeteries, especially compared to those from the settlements, with a few exceptions (see F.586 below). However, it should be noted that this graph displays the number of sherds, and in terms of minimum number of vessels, this equates to eleven different 222

121 vessels from the cemetery. That said, the settlement produced a minimum of just three flagons (based on the number of individual rims/bases), therefore this is one of the few examples of cemetery assemblages utilising wares which may be considered somewhat periphery within the contemporary associated settlement, thus suggesting that in this instance particular vessels/vessels forms were purposefully chosen for inclusion within graves. Form No. Wt (g) EVE Amphora Beaker Bowl Closed Cup Dish Flagon Jar Lid Mortaria Open Platter Sieve Spindle whorl Unknown Table 3.44: All sherds by vessel form from RB.2D (including Cemetery 1 & isolated cremations). RB.2D Feature Case Studies - The largest assemblage from a single feature derived from well F.586, totalling 1379 sherds weighing 27.7kg and representing EVEs, with a mean weight of 20g. This feature therefore produced 32.5% of the entire assemblage from RB.2D. The bulk of this pottery derived from a dense spread of material deposited during the construction of the well. Material was recovered from twelve contexts a represented a minimum of 247 different vessels. In addition some material was recovered from a later ditch that truncated this deposit (F.610 see below). A small but significant number of sherds were noted as being abraded, with 36 sherds in total having use-wear evidence. Exterior sooting and interior limescale were the most commonly occurring, although one beaker base had been trimmed, suggesting a secondary function. 551 sherds of pottery weighing 11.11kg (mean weight 20g), equating to a minimum of 102 different vessels, were recovered from context [1574]. The pottery was somewhat mixed in date, with some early Romano-British material, however, the bulk of the pottery dated AD This included a Nene Valley colour-coated beaker with phallic barbotine decoration, as well as an unusual fine sandy micaceous greyware bowl, which appears in form at least, to be an imitation of a Samian Dr37, with cordon decoration. Interestingly, sherds from this vessel refitted with sherds from [1573] the fill immediately overlying it. Context [1573] contained 347 sherds, weighing 5039g, with a minimum of 66 different vessels identified, based on the number of individual rim and bases recovered. The majority of sherds dated AD ; however, there were a small number of later dating sherds including a Nene Valley colour-coated imitation Dr38 bowl (AD ) noted as being abraded. These later sherds probably relate to the backfilling of the well shaft rather than the construction related deposit. There was also a Nene Valley greyware castor box, which is unusual as these forms are generally restricted to colour-coated vessels. Some individual vessels that had fragmented in situ could be identified; these included a WW flagon [1615], a Q4 jar [1616], a possible second Q4 jar although it may be the same vessel [1617], a colour-coated beaker [1618] and a sandy ware jar [1619]. 223

122 Figure 3.42 Comparison of key vessel forms for Settlement RB2D versus associated Cemeteries 1 & 7 Figure 3.44 Percentage frequency of pottery by earliest date for the five main sites

123 Overall, flagons appear to be well represented in this group, with a minimum of five different vessels identified based on the number of rims and bases, which represented half of all flagons (based on minimum number of vessels), recovered from RB.2D settlement features. Interestingly, there is a similar story with beakers, with a minimum of fifteen vessels recovered from F.586, from a settlement total MNV of 38 vessels In comparison to the overall site assemblage for RB.2D, the pottery recovered from F.586 is much more comparable to C1, in terms of vessel forms. This has interesting implications for how we interpret pottery assemblages, in particular the extent to which general settlement assemblages can be taken as a normative baseline. It is arguable that the assemblage recovered from well F.586 could be typical of the settlement as a whole, perhaps only visible as such due to the large quantity of material recovered. It is also possible that the relatively high number of vessels related to drinking in particular, may indicate this particular feature assemblage as reflecting something beyond the everyday activities of the site and instead may be indicative of more specialised, less frequent event(s). A second well, F.548, produced a comparable assemblage, although much smaller in size, totalling 256 sherds weighing 2731g and representing 7.49 EVEs. The pottery had a mean weight of just 10.7g, which was recovered from four contexts. The primary dump was context [1381], which comprised 221 sherds weighing 2158g and representing a minimum of 32 vessels. The pottery from this fill dated AD , although there were also some earlier Roman, residual sherds identified. Several sherds were noted as being abraded/heavily abraded including five different Samian sherds. The remaining three contexts [1551], [1552] and [1553] contained much smaller quantities of pottery, with all three dating AD The two wells discussed are somewhat different in character to those already discussed from settlements RB.2A and RB.2C. Firstly, the RB.2D wells appear to have been slightly earlier in date than the others, or at least went out of use as wells at an earlier stage, with the primary backfilling episodes occurring between AD , the other wells at North West Cambridge showing evidence of going out of use sometime after AD250. Secondly, the pottery itself was more fragmentary, with a lower mean weight of 20g for F.586 and just 10.7g for F.548. Finally, a greater number of sherds were noted as being abraded/heavily abraded, with a higher number of sherds noted as having use-wear evidence, primarily consisting of exterior sooting or interior limescale. A group from ditch F.634 [1659] is largely comprised jars in Q1 and Q3 fabrics and although identified as significant in the excavation records is probably similar to F.1076 (see above). Other Sites/Isolated Features - The remainder of the Romano-British pottery assemblage derived from a number of features not associated with any specific settlement. Some of the material was residual, occurring in later dating features, while further sherds were intrusive in earlier features. Overall, the remainder of the assemblage accounted for 5.9% of the assemblage by count and 4.9% by weight. 296 sherds were recovered from three routeways, weighing 3178g and representing 3.16 EVEs. 30 sherds weighing 351g were collected from RB Way 1, dating to the 1st century AD. The material was generally small and several sherds were noted as being abraded. A further 34 sherds (216g) were recovered from F.2584 RB Way 1 & 2. The bulk of the pottery derived from RB Way 3, totalling 232 sherds, from sixteen different features. The material was quite mixed in date with some early Romano-British (mid/late-1st century AD) occurring alongside some mid/later Romano-British pottery. 310 sherds weighing 5508g and representing 4.77 EVEs were collected from Post-Roman features. This included ditch F.772 which contained 116 sherds of Romano-British pottery weighing 2845g, dating to the mid/later Romano-British period. This material presumably derives from the underlying Romano-British ditch (F.932). This included several large Horningsea greyware sherds and a large sherd from a Horningsea black-burnished ware beaded, flanged bowl. A further 94 sherds (1226g, 1.72 EVEs), dating to the 2nd 3rd century AD, were recovered from a Second World War ditch (F.610). This pottery all derives from a single slot, located directly over a Romano-British well (F.586, see above) and the ditch 225

124 truncated the dense pottery concentration. Finally, the Double-Ditch system contained ten sherds weighing 71g and representing 0.10 EVEs. The pottery was recovered from three features, and dated AD Cemeteries Pottery was recovered from five cemeteries across the site. This totalled 2901 sherds weighing 28.4kg and representing EVEs. This figure includes all sherds recovered from cemetery features. However, there is a distinction between grave goods and sherds which had been caught up in the backfill of the graves. A minimum of 64 grave good vessels were identified, totalling 2669 sherds and weighing 26.75kg, thus representing 92% of all the cemetery pottery. As with the settlements, the cemeteries spanned the Romano-British period. That said there does appear to have been some overlap with some of the cemeteries, rather than each reflecting a different phase of occupation. The earliest dating cemeteries were the cremation and mixed cremation/inhumation cemeteries plus isolated cremations: C3, C5 and isolated cremations in Settlement RB.2D, which appear to have been, at times, contemporary with one another, having their origins in the mid/later 1st century AD. Unsurprisingly, the inhumation cemetery C2 appears to have been the latest dating cemetery, spanning c. AD and possibly beyond this date. The earliest individual graves were F.1958 and F.2109 (C3) and F.936 (isolated cremation), while the latest individual graves were F.629 and F.849 dating AD (C2). With a few exceptions the cremations were relatively shallow, they have therefore suffered post-depositional plough damage and few of them are complete. Additionally, many vessels have been crushed, presumably by heavy farm vehicles repeatedly passing over them. The North West Cambridge cemeteries data provides a unique opportunity to compare and contrast the burial practices over the Romano-British period, in terms of the ceramics, occurring within a single community. Patterns observed from the analysis of the ceramic grave goods demonstrates that earlier graves tend to have a greater number of ceramic vessels associated with them, with the mid-1st 2nd century AD examples having between one and five vessels per grave. The mid/later Romano-British graves contained between one and four pots, with half containing just a single vessel (although this excludes the inhumations which contained no pottery grave goods). That said, there does appear to have been longevity over the forms of vessels selected for burial, with the most commonly occurring multi-pot group comprising a coarseware jar, a dish (often Samian) and a flagon and/or beaker. Thus representing a complete set of vessels used for the serving of food and drink, and in the case of the graves, for the offering of foods stuffs as part of the burial practice, a custom commonly identified within Romano- British cemeteries in Britain. These multi-pot groups effectively constitute the container for the cremation (the coarseware jar), a lid placed over the cremated bone within the main container (an inverted dish) and secondary/ancillary vessels (flagon and/or beaker) located either within the main container beside/above the lid or next to the main container. In some cases the vessel containing the cremation was inverted and the lid was located underneath it. In some instances where lids and secondary/ancillary vessels are absent this reflects a genuine pattern, in other cases when burials 226

125 have been more heavily disturbed it is possible that these vessels have been entirely removed. Another aspect of the North West Cambridge cemeteries that forms part of a much wider tradition was the ritual killing of a small number of the grave goods. This process refers to the deliberated breaking of a vessel before deposition in the grave, and commonly includes the clipping of a piece of rim, or the stabbing of a vessel. A total of six vessels from three of the cemeteries were noted as having been ritually killed, comprising three Samian dishes with parts of the rim clipped-off (F.638, isolated cremation; F.507 C1; F.502 C1). F.535 from C1 contained two vessels with post-firing holes [1334] and [1337], likely to have been done as part of the burial process. Finally, a Nene Valley colour-coated beaker from F.629 (C2), had a missing rim, which appeared to have been deliberate rather than accidental, although this is hard to prove conclusively (based upon its location etc. the missing rim must at least pre-date deposition in grave). Although the ritual killing of vessels could not be considered as common-place with the North West Cambridge cemeteries, it is of note that the ritual killing of vessels occurred with both earlier and later dating graves, and shows that certain practices were relatively long-lived. The final element to explore is the possible curation of certain grave goods. This practice is often most visible amongst graves containing Samian, as the forms, fabrics and stamps often give a much closed dating bracket that other vessel types. F.502 (C1) for example, comprised small closed vessel dating AD50 100, a large jar dating AD and a Central Gaulish Samian Dr31 dish dating AD , thus suggesting the two locally made vessels had been curated for a period of time before being deposited in this grave. Again, this practice is not uncommon; however, it is debatable as to whether these curated vessels were kept especially for these occasions or whether there is a more utilitarian explanation. C1 (Settlement RB.2D) - A mixed cremation and inhumation cemetery, with evidence that its use spanned two phases of activity at the site. The assemblage from C1 was the largest of all the cemeteries, totalling 1024 sherds of pottery weighing 15.75kg and representing EVEs. Grave goods accounted for 94% of the total assemblage, the remainder comprising sherds caught up in the backfill of graves. Eleven graves contained pottery grave goods, with a further two containing non-grave good pottery. The graves appear to span the early/mid Romano-British period, with the earliest dating AD50 150, and the latest dating AD This therefore implies that the cemetery was in use for a relatively long period of time. F Contained three pottery grave goods, comprising one jar, a Samian Dr31 dish and a closed buff sandy ware vessel, with a date of AD The Samian vessel is the latest dating of this group, with a range of AD Interestingly, the other two vessels are slightly earlier in date (AD50 150) which implies that they may have been curated. (1211) cremation vessel Q3 large jar with everted, square rim, AD sherds weighing 1746g. (1212) lid Central Gaulish Samian Dr31 dish with stamp MA..., AD Six sherds weighing 282g. (1210) secondary vessel external Fine sandy buff ware closed vessel, AD sherds weighing 170g. 227

126 F A single vessel was recovered from F.503 comprising a coarseware jar dating AD , although a further nine sherds (5g) were collected from the backfill of the burial. (1220) cremation vessel Q3 medium-sized jar, AD sherds weighing 374g. F A medium-sized greyware jar was recovered from this feature, dating AD A single sherd (22g) from an oxidised sandy ware jar was found in the backfill. (1227) cremation container Coarse, sandy greyware jar with everted, slightly beaded rim. 23 sherds, 242g. F Contained two ceramic grave goods, comprising a sandy ware jar and a buff sandy ware flagon, dating AD (1224) cremation vessel Q2 medium-sized jar with limescale on the interior, dating AD sherds weighing 552g. (1225) secondary vessel external Buff sandy flagon with limescale on the interior, dating AD sherds weighing 189g. F Three complete/near complete grave goods were collected from F.506, comprising a complete South Gaulish Dr18 dish, a buff sandy flagon and a large sandy jar. The vessels had a date range of AD , although the Samian vessel is slightly earlier than this, suggesting possible curation. The flagon had been clipped at the rim. (1341) cremation vessel Q1 large necked jar with a beaded rim, almost complete. AD sherds, 2467g. (1320) lid A South Gaulish Dr18 dish, complete, though with a very worn interior, suggesting it was used prior to being a grave good. One sherd 367g. (1321) secondary vessel internal Buff sandy small globular flagon, with one handle. Part of the rim had been clipped prior to deposition. AD One sherd, 384g. F Four semi-complete vessels were recovered from F.507, with a feature date of AD This included a large coarseware jar, a flagon, an indented beaker and a Samian Dr18/31 dish, which had a possible clipped rim. This feature has a date of AD (1270) cremation vessel Q1 large necked jar with beaded rim, combed. Semi-complete, AD sherds, 1516g. (1271) lid South Gaulish Samian Dr18/31 dish, with a possible clipped rim dating AD Two sherds, 321g. (1272) secondary vessel internal Whiteware flagon, AD sherds, 131g. (1273) secondary vessel internal Q1 medium-sized jar, eleven sherds, 106g. F Four grave good vessels were recovered from F.508, giving a date of AD This comprised a Central Gaulish Samian Dr31 dish, an oxidised flagon, an oxidised indented beaker and a large coarseware jar. (1323) cremation vessel Q3 large necked jar with beaded rim, semi-complete. AD (1276) lid Central Gaulish Dr31 dish, almost complete. Stamp VA...RV. AD , eleven sherds, 397g. (1277) secondary vessel external Oxidised sandy flagon, rim missing. AD sherds, 188g. (1278) secondary vessel external Oxidised sandy indented beaker, half complete. AD , eleven sherds 242g F Contained three grave good vessels, comprising a coarseware jar, a colour-coated flagon and a black-slipped beaker. The jar had a post-firing hole in the side, which can be considered as evidence of ritual killing. Four sherds (25g) were also recovered from the backfill of this feature, including one sherd from a Nene Valley colour-coated flagon and three oxidised sandy ware body sherds. This feature has been dated AD (1334) cremation vessel Coarse sandy oxidised medium jar. Semi-complete with a ritually killed hole in the side. AD sherds, 1091g. (1336) secondary vessel external Colour-coated globular flagon, with roulette decoration. AD sherds, 413g. (1337) secondary vessel external Black-slipped plain rim beaker, with large hole in the base. AD One sherd, 110g. F This cremation had been extremely heavily damaged; it effectively consisted of scattered material in a plough scar. Two vessels were recovered from F.542, consisting of a 228

127 greyware closed vessel and a colour-coated cornice rim beaker. Sherds from a further five vessels were also recorded, although these appear to have been caught up in the backfill rather than reflecting additional grave goods. The pottery from this feature suggests a date of AD (1361) cremation vessel Coarse sandy greyware closed vessel. AD sherds, 171g. (1361) secondary vessel Colour-coated cornice rim beaker with barbotine scroll decoration. AD , eleven sherds, 17g. F This grave comprised two near-complete vessels; a fine sandy oxidised beaker and a greyware closed vessel with a pedestal base. Ten sherds (7g) from a fine sandy greyware beaker were also identified. (1408) cremation vessel Fine sandy micaceous oxidised ware beaker, with a long neck and small beaded rim. Rouletted decoration and a cordon. AD sherds, 706g. (1409) secondary vessel external Fine sandy micaceous greyware closed vessel with pedestal base. AD sherds, 181g. F Grave 823 consisted of three ceramic vessels; an East Gaulish Curle 15 dish, a Horningsea greyware jar and an oxidised flagon. The pottery suggests a date of AD (2304) cremation vessel Horningsea greyware necked jar with beaded rim, semi-complete. AD sherds, 843g. (2302) lid East Gaulish Samian Curle 15 dish, complete. AD Four sherds, 279g (2305) secondary vessel external Oxidised sandy ware small flagon. AD sherds, 123g. F This was an inhumation with no recognised pottery grave goods. The pottery from this feature comprised a small number of sherds, with no definite grave goods identified. Three sherds from a Q3 closed vessel (51g) were recorded, along with one sherd (31g) from a Q1 lid and a further sherd (15g) from a Q1 closed vessel. The pottery dates AD which supports the view that this material had been caught up during the backfilling of this feature, rather than reflecting grave goods. F An inhumation with no recognised pottery grave goods. As with F.850, there were no definite grave good vessels identified within F.858. Eleven sherds of pottery were collected, comprising two Horningsea greyware sherds from a jar, three sherds (33g) from a buff sandy closed vessel and six sherds (33g) from a coarse sandy greyware closed vessel. C2 (Settlement RB.2C) - This cemetery consists solely of inhumations. Pottery was recovered from ten of the twelve or thirteen graves within C2, totalling 161 sherds weighing 1171g and representing 8.60 EVEs. However, not all of this material can be considered as grave goods, with four features containing no obvious grave good pots. The remaining six features contained a single ceramic grave good suggesting that this cemetery was mid/later Romano- British in date, with a range of AD for most of the graves, thus making this one of the latest dating cemeteries on the site. F One grave good vessel was identified from this feature, comprising a Nene Valley colour-coated beaker, dating AD Sherds representing a maximum of six other vessels (11 sherds, 64g) were also recovered, although these are unlikely to have represented grave goods, instead having been caught up in the grave backfill. This included sherds from a Q1 closed vessel, shell-tempered sherds and coarse sandy greyware sherds. (1642) secondary vessel Nene Valley colour-coated bag-shaped beaker. AD One sherd, 179g. F This grave contained one semi-complete vessel, as well as sherds from several other residual vessels, totalling fifteen sherds weighing 80g. The grave goods comprised a small colour-coated beaker dating AD The remaining sherds included a shell-tempered sherd, a Verulamium whiteware sherd and seven coarse sandy greyware sherds. (1690) secondary vessel Colour-coated beaker with a small beaded rim, semi-complete, dating AD sherds, 80g. 229

128 F This feature contained no human bone and appears to represent a child or infant burial. A single grave good comprising a Colchester colour-coated beaker, dating AD was recovered from this feature. Two further sherds (8g) were collected from the backfill, comprising one Q2 sherd and one coarse sandy oxidised body sherd. (1748) secondary vessel Colchester colour-coated funnel-necked beaker, with white barbotine decoration. AD Six sherds 62g. F The grave good from this feature comprised an almost complete Colchester colourcoated mini bag-shaped beaker with white painted decoration, dating AD (2494) secondary vessel Colchester colour-coated mini bag-shaped beaker with white painted dot and scroll decoration. Almost complete, dating AD Three sherds, 100g. F A colour-coated beaker dating AD was recovered from this feature, along with seven residual body sherds (27g). (2423) secondary vessel Colour-coated beaker with a long neck, and decorated with barbotine swirls and rouletting. AD sherds, 130g. F This grave contained a Nene Valley colour-coated beaker, dating AD Three further sherds (135g) were collected from the backfill of this feature, comprising a Horningsea greyware bifid rim (106g) a greyware jar sherd (25g) and a further Nene Valley colour-coated sherd (4g). (2500) secondary vessel Nene Valley colour-coated beaker with a long neck and small beaded rim, dating AD sherds, 126g. F Twelve sherds weighing 84g were collected from the backfill of this feature, none of which could be classed as grave goods. This included a Horningsea greyware jar sherds, a body sherd from a Nene Valley colour-coated vessel and several coarseware body sherds. The pottery from this feature dates AD F No ceramic grave goods were present within the feature. However, three sherds weighing 18g were recovered, comprising one Pakenham colour-coated body sherd, one coarse sandy greyware jar sherd and one Q3 body sherd, all dating AD F.630: Five sherds (45g) from two vessels were identified, neither of which constitute grave goods. These comprised four coarse sandy greyware body sherds (11g) and one Q1 jar sherds (34g), all of which could only be broadly dated as Romano-British. F.848: Three sherds weighing 28g, dated as Romano-British. This consisted of one oxidised sandy sherds (2g) and two Q1 jar sherds (26g). C3 (Settlement RB.2A) - A mixed cremation/inhumation cemetery, which spanned several distinct phases of activity at the site. A total of eight graves within C3 contained Romano-British pottery, although in three cases, this did not include any definite grave goods; the pottery recovered coming from the backfilling of the features. The pottery collected from C3 totalled 779 sherds, weighing 5370g, and are predominantly earlier Romano-British in date. F A single vessel was recovered from this feature, comprising a white-slipped closed vessel, dating AD (2295) cremation container White-slipped closed vessel, AD Fifteen sherds, 1435g. F This feature contained a black-slipped jar, dating AD No other pottery was recovered from this grave. (2609) cremation container Black-slipped medium-sized jar, AD sherds, 409g. F A large assemblage of pottery was recovered from F.1958, totalling 298 sherds weighing 2021g. This comprised four semi-complete grave goods were collected from this feature, comprising four beakers, dating AD There were also 20 further sherds (77g) recovered from the backfill, including fourteen coarse sandy greyware body sherds. 230

129 The additional pottery from (5870) a CSGW beaker must have been inside the Grogtempered cornice rim beaker. This suggests that it was a further grave good. Additionally the CSGW in the backfill might also come form this, this idea is supported by the fact that this cremation is probably amongst the earliest events in the area and there shouldn t be much domestic debris kicking around yet. (5870) cremation vessel Grog-tempered cornice rim beaker, semi-complete, AD sherds, 489g. (5870) secondary vessel internal CSGW beaker. (5872) cremation vessel QG1 beaker with large pedestal base.ad sherds, 405g. (5874) cremation vessel QG1 beaker, with an everted rim and large pedestal base, AD sherds, 496g. (5876) cremation vessel QG1 beaker with everted, rounded rim. AD sherds, 554g. F A single butt-beaker was deposited within this grave, dating AD (5988) cremation vessel Q4 butt-beaker, with rouletted decoration. AD sherds, 633g. F F.2109 contained five different grave good vessels, as well as six coarse sandy greyware body sherds (62g). The semi-complete vessels comprised two butt beakers, a jar, a flagon and a further closed form. The pottery recovered suggests a grave date of AD Although not recognised in the field it appears that flagon (7357) contained jar (7357). (7356) cremation vessel Buff sandy butt beaker, semi-complete. AD sherds, 602g. (7357) secondary vessel external Q1 small jar with beaded rim, AD sherds, 187g. (7357) secondary vessel external Coarse sandy greyware flagon, AD sherds, 253g. (7358) secondary vessel external Coarse sandy greyware closed vessel, AD sherds, 441g. (7359) secondary vessel external Fine sandy micaceous greyware butt beaker, semicomplete. AD sherds, 237g. F a heavily disturbed feature that contained some human bone and probably represents the remnants of an inhumation. Although eight sherds of pottery were recovered from F.1931, there were no obvious grave goods. The pottery comprised four sherds (12g) from a colour-coated beaker, with barbotine and rouletting decoration, and one sherd (91g) from a Horningsea greyware jar. F an inhumation with no recognised pottery grave goods.11 sherds of pottery weighing 121g were recovered from this feature, although none of these comprised grave goods. The sherds included three Verulamium whiteware sherds and seven greyware body sherds. F the ring gully of part of C3. Nine sherds of pottery, weighing 75g were recovered from this grave; however, there were no clear grave goods. The sherds present included two sherds from a Horningsea greyware jar, a coarse sandy greyware jar and a black-slipped vessel. The pottery could only be broadly dated as Romano-British. C5 (Settlement RB.1) - a purely cremation cemetery. The pottery assemblage recovered from C5 totalled 261 sherds weighing 3576g and representing 8.56 EVEs. Material was recovered from five different features, and all dated to the early Romano-British period (AD50 100), thus making this one of the earlier cemeteries identified. F A single sandy ware jar was recovered from this grave, dating AD (8743) cremation vessel Q1 medium-sized jar, AD50 100, 27 sherds, 514g. F This feature contained two ceramic grave goods dating AD50 100, comprising an imitation CAM12 platter and a greyware jar. A further ten sherds (19g) from a buff sandy ware vessel and a single Q4 body sherd (27g) were also recovered. The ten sherds from the buff sandy ware vessel appear to be a heavily truncated grave good. (8705) cremation vessel Coarse sandy greyware jar, AD50 100, 36 sherds, 603g. 231

130 (8704)/(8705) lid Q4 imitation terra nigra CAM12 platter, AD40 100, five sherds, 194g. (8704) secondary vessel external buff sandy ware vessel. F A single greyware beaker was placed within this grave, dating AD A further two sherds (99g) were recovered from the backfill. (8728) was the principal container, with bone present. It was just heavily truncated. (8728) cremation vessel Q1 jar. (8729) secondary vessel external Coarse sandy greyware beaker with a pedestal base, dating AD sherds, 30g. F A total of three grave goods were identified from F.2666, comprising a mediumsized jar, a platter and a closed vessel. The pottery collected suggests a date of AD Nine sherds (49g) from a sandy greyware jar were recovered from the backfill. (8713) cremation vessel - Coarse sandy greyware medium-sized jar, AD sherds, 49g. (8711) lid Coarse sandy greyware imitation terra nigra CAM12 platter, AD Fourteen sherds, 208g. (8710) secondary vessel internal Buff sandy closed vessel, AD sherds, 169g. F A single medium-sized jar was recovered from this grave, dating AD (8742) cremation vessel Q1 medium-sized necked, everted rim jar, dating AD sherds, 761g. Isolated Cremations (Settlement RB.2D) - Two isolated graves within settlement RB.2D contained ceramic grave goods, totalling 675g weighing 2527g and representing 0.77 EVEs. The pottery recovered from this cemetery suggests that the cemetery was in use in the early Romano-British period, between AD40 70, thus potentially the earliest of the site cemeteries. F Three semi-complete ceramic grave goods were recovered from this feature, comprising two flagons and a South Gaulish Samian Dr18 dish, which had part of the rimmed clipped, prior to deposition, a process known as ritual killing. The coarse sandy oxidised flagon from (1667) had been completely dispersed within the general fill and was not recognised as a grave good, suggesting it was originally placed higher than the other vessels. (1669) cremation vessel Buff sandy flagon, AD sherds, 775g. (1671) lid South Gaulish Samian Dr18 dish, with part of rim deliberately clipped. AD sherd, 224g. (1667) secondary vessel, probably internal Coarse sandy oxidised flagon, AD sherds, 79g. F The pottery from F.936 contained 401 sherds weighing 1448g, from three grave goods. These comprised a flagon, an imitation Terra Nigra platter and a closed vessel. 25 sherds (57g) were recovered from the backfill. The buff sandy closed vessel was not individuated during excavation and must have been placed within the buff sandy flagon (2670), (2671), (2706) cremation vessel Buff sandy flagon, AD sherds, 1039g. (2706) lid Fine sandy micaceous greywares platter, imitation CAM12 form, AD sherds, 245g. (2670) secondary vessel internal Buff sandy closed vessel, AD sherds, 107g. Discussion The North West Cambridge assemblage represents one of the largest groups of Romano-British pottery recovered from Cambridge in recent years. The quantity of material recovered, along with the sites longevity make it a very important site within the Cambridge environs. The presence of five cemeteries amongst the settlements offers further opportunity to understand not only about everyday life at the site, but also customs and rituals 232

131 associated with death. In terms of the ceramics, the two groups of pottery can be used to try and gain a deeper understanding of the fundamental differences between settlement and cemetery assemblages. The ceramics offer some of the best opportunity for dating the sites, especially when used in conjunction with other well dated artefacts, such as coinage and leather. The pottery allows for a detailed insight into when the individual settlements (and cemeteries) were peaking as well as when they were in decline. Figure 3.44 shows the percentage frequency of pottery for the five key sites. It is of interest that all five sites show different distributions of pottery overtime. That said, it is clear that many of the sites overlapped and therefore are likely to have been contemporary with one another during certain periods. RB.2B represents one of the earliest dating sites at North West Cambridge, with Late Iron Age origins. The pottery indicates that this settlement/part of settlement had ceased to be a focus of activity after the mid-2nd century AD. Although there were a few examples of pottery post-dating this time (e.g. Nene Valley colour-coated wares), these sherds make up only a very small proportion of the overall assemblage, and in several cases appear to have been intrusive. The relative lack of variety in the types of features that pottery was recovered, as well as the lack of associated cemetery has interesting implication of both the function and status of this settlement. The latest dating settlement appears to have been RB.2D, which had evidence of activity into the mid-4th century AD and possibly beyond. This is, however, slightly misleading as it appears that this relates principally to the use of a single well after the main settlement had gone out of use. This site also constituted the largest single settlement assemblage within North West Cambridge. Site Coarsewares % Finewares % Imported wares % RB RB.2A RB.2B RB.2C RB.2D C C C C Isolated cremations Table 3.45: All Romano-British pottery from main sites and cemeteries by percentage of coarsewares, finewares and imported wares. Table 3.45 shows the breakdown of coarsewares, finewares and imported wares for each of the five main sites as well as the cemeteries. Overall, there is very little variation in the composition of the assemblages, with a few notable exceptions. Firstly, C2 is the only site where finewares, not coarseware vessels dominate, although it should be noted that the pottery from this cemetery 233

132 represented a minimum of just fifteen vessels. The other exception to the overall pattern identified at North West Cambridge was RB.2C, which had the highest frequency of Romano-British finewares and consequently the lowest percentage of Romano-British coarsewares. As discussed above, this has interesting implications for this specific site, suggesting that the sites relative status and/or function may account for this difference. Imported wares never accounted for any meaningful percentage of any assemblage, with settlements RB.2C and RB.2D comprising the highest proportion of wares, totalling just 3%. This is however, not unexpected from a rural site in Cambridgeshire, with these wares generally representing less than 5% of a total assemblage. Since the material from North West Cambridge as a whole spans the period from the Late Iron Age to the later Romano- British period, it is unlikely that the date of occupation is the prime factor in this low figure. Instead it seems that the infrequency of imported wares was either due to supply, cost and/or desire. The overall low frequency of imported wares within the North West Cambridge assemblage is not unexpected, given the sites location and function/status and it is comparable to other rural, domestic sites in the region in terms of assemblage composition and character. That said, although imported wares never feature highly, they are informative about trade networks operating around this site and this area of Cambridge as a whole. Comparisons of settlement versus cemetery assemblages are of interest, and with the possible exception of the high frequency of flagon sheds from settlement C1 (RB.2D), there is little evidence to suggest that the pottery utilised within graves as grave goods was coming from a separate source to the settlement material. Although it might be assumed that the cemeteries would have produced higher numbers of finewares and even imported ware vessels, this was simply not the case. This is not to say that the selection of ceramic goods that accompanied the dead was random or unconsidered, rather that these vessels were not noticeably different from the pottery being utilised on the contemporary associated settlements. In fact both the repertoire of grave goods selected, as well as the evidence for ritual killing apparent within some of the cemeteries demonstrate that the residents of North West Cambridge were following wider regional traditions when it came to using pottery within burial contexts. There was very little evidence for stratified ceramic sequences within individual features across the site, in that very few features comprised fills with pottery that was clearly of different dates. While on the one hand, this may be due to the nature of the pottery e.g. coarseware sherds with longevity, it may also suggest that many, if not most of the features had been filled within a relatively short period of time, rather than the gradually and slowly backfilling over a long period of time. The lack of stratigraphy in terms of date between features with multi-contexts may also have occurred as a result of re-deposition of material, when new features were dug and/or the clearing out/maintaining of others. In conclusion, the North West Cambridge pottery assemblage offers a unique insight into both the everyday workings and site dynamics taking place throughout the Romano-British period. The quantity of material recovered 234

133 and the scale of the excavations allowing for in-depth inter and intra0site comparisons. There are several sites, both published and unpublished in the immediate vicinity of the site that provide interesting comparisons and will allow for the assemblage to be understood in its wider landscape context, including Castle Hill, (where sherds from recent excavations have been recorded to a comparable standard), Langdale Hale (14,000+ sherds), Camp Ground (73,000+ sherds, 169kg), Hutchison Site Addenbrooke s (20,000+ sherds, 273kg) New Hall (168kg) and Vicar s Farm (12,000+ sherds). This aspect is recommended for the publication, once all the phases of North West Cambridge have been completed. Glass Vicki Herring The material from the Romano-British contexts at North West Cambridge has been quantified by minimum number of vessels which is calculated using colour, form and the presence of diagnostic elements such as bases, rims and decoration. A total of 26 small fragments of Romano-British glass vessels were collected from four sites of excavation (RB.2A D), representing a minimum of thirteen vessels (Table 3.46). A further five fragments representing a minimum of two vessels were also retrieved during the evaluation carried out in The assemblage as a whole is extremely fragmentary and with most pieces being smaller than 20mm, form, function and date are very difficult, and in many cases impossible, to distinguish. The material from Site RB.2A was collected from three Romano-British features, with a minimum of one vessel from each. A further vessel fragment was collected during evaluation making a minimum vessel count of four for the site. However, most of the fragments are so small as to provide little if any diagnostic information. Only two of the fragments from this site have any diagnostic elements remaining. A small colourless shard from well F.2023 has a tiny amount of what appears to be dark black/brown paint suggesting that this was possibly a tableware vessel, most likely a 2nd 3rd century cylindrical cup, with enamelled decoration. Vessels with this form of decoration are rare throughout the Romano-British period in general and have only been found in small fragments in Britain, the greater concentration of these being around Hadrian s Wall. Complete vessels, almost always cylindrical cups, have been found in Germany and Denmark and it is likely that examples found in Britain were imported from that region (Price & Cottam 1998). The second diagnostic fragment is around half of the rim and top of the neck from a cylinder or prismatic bottle which was collected from Trench 51 of the 2009 evaluation. Bottles were common utility vessels of varying sizes used for transport and storage and those with this rim type were very common in the last quarter of the 1st century, with similar examples found at Stansted, Essex (Price & Cottam 1998), though the prismatic forms were longer lived and are most common at Colchester from the late-1st to the late-2nd century (Cool & Price 1995). A single fragment of colourless green glass representing one vessel of unknown form was collected from sl.481 of Romano-British ditch F.1076 of Site RB.2C. A minimum of six vessels are represented from Site RB.2D, one from each of six Romano-British features in RB.2D. A further vessel fragment was collected from this area during the 2009 evaluation. Three rim fragments, two of which refit, and a small body shard from F.247 of the 2009 evaluation, are probably derived from the same cylindrical cup, 12cm in diameter. With a vertical fire rounded rim. Cylindrical cups such as these were the most common form of drinking vessel in Britain prior to the 4th century, although they appear most commonly in the 2nd 3rd centuries. Close parallels can be found at Verulamium, Hertfordshire, in a late-2nd century well deposit (Price & Cottam, 1998, 100), and from Periods 4 6 ( AD) at Colchester (Cool & Price 1995). A very tiny fragment (4x3mm) of glass from cremation F.504 retains no diagnostic features but is very thin walled (less than 0.5mm thick) and pale yellow green in colour which are characteristics of 4th century vessels (Price & Cottam, 1998). 235

134 Feature Site Type Object Colour Date Romano- F.1933 pit RB.2A Vessel? Unk. Colourless British F.2023 Well RB.2A Vessel Cylindrical cup with painted decoration? Colourless with dark (black/brown) enamelled paint F.2023 Well RB.2A Vessel? Unk. Colourless F.2023 Well RB.2A Vessel Unk. Colourless F.2044 Well RB.2A Vessel Unk. Colourless F.2044 Well RB.2A Vessel Unk. Colourless NWC09 TR158 RB.2A Vessel Cylinder/prismatic bottle F.1076 ditch RB.2C Vessel Unk. Blue green Colourless green NWC09 F.247 RB.2D Vessel Cylindrical cup Colourless F.504 Cremation RB.2D Vessel Unk. Pale yellow green F.548 Well RB.2D Vessel Cylindrical bottle Blue green F.586 Well RB.2D Vessel Unk. Colourless F.617 Ditch RB.2D Vessel Unk. F.673 Ditch RB.2D Vessel Unk. Colourless green Colourless green F.681 Ditch RB.2D Vessel Unk. Colourless Table 3.46: Romano-British vessel glass. 2nd 3rd Century Romano- British Romano- British Romano- British Romano- British 1st 2nd Century Romano- British 2nd 3rd Century Romano- British Romano- British Romano- British Romano- British Romano- British Romano- British Leather Quita Mould The following assessment is based on examination of the leather at the CAU premises. A basic record of the assemblage was made at that time and is appended to this document. The basic record notes all the diagnostic features present, measurements where necessary and the contextual information as known to date. Nailing patterns are classified according to Carol van Driel- Murray s nailing pattern classification (2001) and constructional thonging as that commonly used for Romano-British nailed footwear (e.g. Mould 2009, 834 fig. 507). The material is summarised below, any further work necessary is outlined and recommendations for conservation given. The leather was wet and washed when examined. The wet material is currently packed wet in double, self-sealing polythene bags or within air-tight storage boxes. Some of 236

135 the leather was delicate, liable to tear easily and to fragment further. In view of this it was considered advisable to limit handling until those items have been conserved. All the leather found was Romano-British footwear. No waste leather from the manufacture or repair of leather goods was recovered. The leather was recovered from wells (F.1020, 1168, 1402, 2056, 2756, 2768) and two ditches (F1173, 1402; Fig. 3.45). No large groups were present. The shoes were of two principal constructions: nailed shoes and one-piece shoes (Fig. 3.46). Seven or possibly eight shoes of nailed construction were present. Highly fragmentary remains of shoes of nailed construction were recovered from F.1173, 1214, 2756 and Three nailed shoes were better preserved and are detailed below: Well F A shoe of nailed and stitched construction <1578> came from well F.1020 associated with later third and fourth century pottery, along with fragments probably from a second shoe of nailed construction. The shoe of nailed and stitched construction <1578> had an upper with a single side seam stitched with leather thong, possibly with a tall leg of soft leather, and may potentially be a boot of Balland s Tall-J style (Volken 2012, 122 fig. 105) a 3rd century style dated c AD. Study of the conserved leather may reveal features that allow identification. Well F A complete nailed shoe bottom with remains of the shoe upper surviving on the left side was found in well F Well F A well-preserved shoe of adult-size 3(35) was recovered from a well F Much of the shoe upper survives at the toe and back part (heel) but unfortunately the central section (around the instep area of the foot) is broken. The shoe has a decorative feature known as a toe stripe running from the broken throat area vertically down toward the toe created using a decorative stitch of running thonging. This distinctive feature identifies the shoe as of Portchester-J style (Volken 2012, 126, fig. 105) dated to the later third and fourth centuries. A small but growing number of examples of this style have been identified in Britain in recent years occurring notably in fourth century well fills in rural communities in the southeastern counties (Cambridgeshire, Bedfordshire, Buckinghamshire, Lincolnshire, and Northamptonshire); only three have been recognised from Europe so far. To date no direct evidence of how this shoe style was fastened has survived. Though broken in the central section the NWC12 example does have further pieces broken from the upper present that may provide valuable information regarding its fastening. The frequency of damage to this central section may suggest that a fastening in this area created a point of weakness. In order to prevent any deterioration of the upper, examination of the sole and broken upper fragments will be undertaken once the shoe has been conserved and is more robust. Four shoes of one-piece construction were present. A one-piece shoe was found in the box-lined well F.2056 and another in well F.2768 with two examples in well F.1168 associated with pottery dated to AD. In addition a large repair piece from a one-piece shoe was found in well F.1402 associated with the nailed shoe bottom from [4218]. While two of these shoes (<1580.1> F.1168 and <000> F.2768) were of child size, another (<2989> F.2056) was of a large size to fit an adult man possibly adult size 7(41). The two child-sized shoes are likely to be of Amcotts-U style dating to AD. Features on <1580.2> F.1168) suggest it is of Zugmantel-U style (Volken 2012, 121, fig. 103) dating earlier to AD. The large example (<2989> F.2056) is certainly of a different style to these having a decorative cut out and lobed fastening loop close to the back seam and long plain fastening loops also associated with it. Further examination of the fragments may result in identification. Some of the material is highly fragmentary and it is unlikely that further examination would add significantly to the information already gathered so 237

136 that no further work is considered necessary. A proportion of the material survives in a better condition and comprises one-piece shoes and nailed shoes with part of the shoe upper preserved (seven examples in all). Details of these items will be of interest to other researchers. It is suggested that they be conserved by freeze-drying to allow them to be re-examined, photographed or illustrated if necessary and safely stored allowing access to other researchers. These seven items for freeze-drying and re-examination are detailed below: F.1020 Romano-British Well ([3102] sl.482 <1578>) Leather shoe of nailed and stitched construction, left foot, adult size adult 5+ (38+) Bag 1: Nailed shoe bottom with oval toe (toe broken), wide natural tread tapering to a wide seat with no distinct waist present. The bottom comprises insole, lamina, midsole and lasting margin of upper on the left side, and broken heel stiffener. The insole has type 2 constructional thonging with thong 7mm wide. The lasting margin of the upper survives on the left side of the tread and has the lower part of a side seam sewn with thong preserved. Running thonging, 4mm wide, is present along the left edge joining the upper lasting margin to the midsole; the thong is 4mm wide. Insole surviving length 250+mm, Width tread 101mm, seat 67mm. Bag 2: Numerous fragments of very fragile, thin leather including several pieces with thonged stitching from a shoe upper; may represent the leg of soft leather from a boot. Bag 3: two pieces broken from a nailed shoe bottom: Fragment 164+x67+mm with three layers present apparently from another shoe bottom. Fragment 67+x65+mm from oval toe area of shoe bottom of nailed construction comprising a sole with two hobnails and insole. Smaller fragments broken from the material above. F.1168 Romano-British Well ([3544] sl.536) <1580.1> Leather one-piece shoe, right foot, child size 6 7 (23 24) One-piece shoe with holes worn through the tread area of the toe and seat. Fastening with a series of long plain loops, appears asymmetrical. The back seam extends high up the back of the leg and survives to a height of 65+mm; a second fragment of back seam may join on to it. The back seam and seat seam are butted edge/flesh stitched. Leather cattle hide 2.30mm thick. Estimated length c. 155mm. <1580.2> Leather one-piece shoe, adult size. Bag 1: three pieces torn from a one-piece shoe: 63+x42+x2.47mm, 93+x36+x1.89mm, curved piece from side of shoe 106+x35+x1.82mm. Bag 2: small with a tooled (rounded) edge from the top edge of a shoe 31+x20+x3.24mm. Other small fragments. Bag 3: three pieces broken from a one-piece shoe: the largest piece is broken from the right side of the heel area with a butted edge/flesh back seam, stitch length 7mm, surviving to a height of 47+mm and seat seam. It has two long, plain fastening loops with a broken connecting bridge surviving; the uppermost loop is c. 80mm long. Cattle hide 3.72mm thick. F.1173 Romano-British Ditch ([3539] sl.536) <1579> Leather fragmentary shoe bottom unit of nailed construction Eight fragments and 3+ very small fragments broken from a bottom unit of nailed construction, including: Piece with central constructional thonging likely to come from an insole. Cattle hide 2.05mm thick. 67+x34+mm tapering fragment of forepart lamina with constructional thonging. Cattle hide 2.50mm. 81+x43mm lamina from seat area with awl made tunnel stitching holes to attach the lasting margin of the shoe upper and a nail hole. Cattle hide 1.72mm thick 67+x34mm. F.1214 Romano-British Ditch ([3115] sl.482) <1360> Leather fragmentary shoe bottom unit of nailed construction, adult size Eight principal and other smaller fragments of broken shoe bottom unit, no indication of the shape of the bottom survives. The fragments include: Heel stiffener c 40+mm high of worn cattle hide 2.78mm thick Insole and lamina with constructional thonging, thong 6mm wide, and a single row of nailing along each side (van Driel-Murray type 1). Insole cattle hide, insole width 61mm. 238

137 F.1402 Romano-British Well ([4218] sl.536) <000> Leather shoe bottom of nailed construction, left foot, child size 11(29) Bag 1: complete heel stiffener 35mm high at centre back of worn cattle hide 3mm thick belong tto the bottom in Bag 2. Bag 2: bottom of nailed shoe with a short, pointed toe, petal-shaped tread tapering to a medium seat with no distinct waist area. The insole and sole are adhering with remains of a closed upper surviving on the left side of the forepart. The upper is of worn bovine leather 2.57mm thick. The insole has constructional thonging of type 1, thong 6mm wide. The sole has widely spaced nailing of van Driel Murrays type 1D. Insole length 195mm, width tread 70mm, seat 37mm. F1402 Romano-British Well ([4218] sl.536) <000> Leather sole repair piece for one-piece shoe, adult size Bag 3: Large clump sole for a one-piece shoe broken along two sides. Cattle hide 2.36mm thick. 97x96+mm. F.2056 Romano-British Box-lined Well ([7417] sl.1380) <2988> Leather shoe of nailed construction with decorated vamp, right foot, adult size 3(35) Nailed shoe with much of the upper present but fragile so details of construction including nailing pattern to be added following conservation Bottom unit with oval toe, natural tread tapering to a wide seat with no distinct waist. The insole has constructional thonging visible in the waist area and lower tread and a single row of nailing is present down each side but other details are uncertain at present. Estimated length of insole 235mm, width tread 85mm, seat 52mm. The upper is well-preserved with the toe area surviving to a height of 75+mm but the throat area is broken, and the back surviving to a height of c. 70+mm at centre back, both sides are broken away. The centre front has a decorative vertical line of thonging, thong 2mm wide, passing through small thong slots spaced 2mm apart, running from the throat down to the toe. Upper leather worn bovine 1.94mm thick. Separate small pieces of upper also present coming from the sides of the upper. F.2056 Romano-British Well ([7445] sl.1380) <2989> Leather fragmentary one-piece shoe, adult male size Fragmentary remains of a large, heavily worn, one-piece shoe, now in more than 20 pieces. The following features were present: Part of the right side of the shoe has remains of the back seam surviving to a height of 39+mm; it appears now as a series of large grain/flesh stitches, stitch length c. 5mm but this is the result of delamination. Fastening loops present at the heel have decorative lobes and a triangular, lobed, cut-out beneath. Separate, torn fastening loops are long and plain. Two potentially joining pieces suggest it may have been c. 270mm long originally. Much of the leather is delaminated one piece is 2.63mm thick. A seat seam is 50mm long and sewn with a grain/edge seam. F.2756 Romano-British Well [(9081] sl.1824) <000> Leather fragments of nailed shoe bottom Four fragments broken from two layers of a nailed shoe bottom unit, apparently broken from the lower tread and waist area of an insole and midsole, one original edge only remaining. Insole cattle hide 2.38mm thick. Largest fragment length 41+mm, width 60+mm. F.2768 Romano-British Well ([9190] sl.1861) <000> Leather one-piece shoe, left foot, child-size 9(27) Near complete one-piece shoe with holes worn through at the toe and seat of the tread area but was not repaired. The left side is better preserved that the right with the fastening loops at the toe end of the shoe torn. Broken or possibly cut at the toe possibly with the suggestion of a toe seam here (check following conservation). The back seam, edge/flesh seam stitch length 8 9mm, is now broken but extended high up the back of the leg surviving to a height of 80+mm. A narrow, plain border separates the long fastening loops from the back seam. A broken fastening loop has a single decoratively scalloped edge with five smaller and one larger lobe. Leather cattle hide. 239

138 19th Century Quarries Archaeological features metres Figure Distribution of features that produced Romano- British leather (with MNI shoe count)

139 centimetres Figure Romano-British shoes Figure 3.47 Relative Romano-British nail lengths

140 F.2768 Romano-British Well ( [9317]sl.1896) <000> Leather fragmentary shoe bottom of nailed construction Eleven fragments from the waist area of a nailed bottom unit including an insole and matching lamina with impression of thong running down each side of the lower face (flesh side) of the insole. A larger piece of the insole 70+x49mm has a single line of nailing running down each side and infilling at the lower tread area (van Driel Murray type 1). A smaller piece of insole has a construction thong slot. Insole cattle hide 2.70mm thick. Metalwork Grahame Appleby Some 2586 pieces of metalwork (weight c. 24.7kg) were recovered during metal-detecting and excavation of archaeological features across the site and reflects the strategy of systematic metal-detecting and excavation. Of these items, one piece was manufactured from aluminium (2g; discarded) and the vast majority were composed of iron objects mainly nails, nuts and miscellaneous un-diagnostic pieces recovered from the top- and sub-soils during metal-detecting (estimated 2000). In addition, 391 copper alloy, lead and composite items were recovered items were recovered from 182 features, 1406 consisting of iron items, 27 of copper alloy, and the remainder lead. Although the assemblage is composed largely of unstratified iron nails and unidentifiable and Post-Medieval objects found during metal-detecting, that exercise and hand-excavation of archaeological features resulted in the recovery of several important items. These include a Romano-British copper alloy steelyard arm from well F.1236 (also found in this feature was a copper alloy brooch, a Romano-British coin and 45 iron objects), two disc brooches (one gilded: F.1236 sf.212), a trumpet brooch fragment (F.910), Romano-British bracelets (F.1020); hair pins (F.2023) and toilet items (F.1088, 2092) and needles (F.695). Other items include unidentified items such as a rhomboidal-shaped copper alloy, tapering pin/stud from ditch F An unusual find was the recovery of a Romano-British pilum head (c. 20cm long) found in waterlogged well F.1148, a possible socketed pilum (F.677) and ballista bolt-head (F.1278), two spearheads (F.711, 1031) and a potential arrowhead (Burial F.856). A possible scalpel was recovered from pit F.906, and several knife or blade fragments and potential tools (gouges and chisels from spread F.1167, wells F.1020, F.1236, and ditches F.1169, F.2040 and F.2066) have also been recovered. Romano-British coins are dealt with separately (see Wells below), a further Iron Age and a Trinovantes/Catuvellani coin dating from the 1st Century BC/AD transition also found during metal-detecting. In addition to this material there was also some metalwork from the evaluation phase that can be directly related to the excavation phase. This includes six fragments of copper alloy sheet, possibly from a thin walled vessel or decorative object. An inhumation (F.76) produced four coffin nails plus three stains of additional nails plus 23 hobnails recovered from the area of the feet and a bar. In addition 40 other iron artefacts, the majority undiagnostic nail fragments, were recovered. Amongst the remainder, however, significant Romano-British finds include a probable candlestick fitting and a possible arrowhead. 242

141 Copper Alloy Excluding coins and a lead-copper alloy object described below, a total of 157 copper alloy items (weight 1074g) were recovered, 24 pieces from 17 features. The overall total includes 29 buttons, all recovered during metal-detecting and Post-Medieval in date and, unless of particular note or earlier in date, not described further. Brooches <10121> (sf.212) - Well preserved cast copper alloy umbonate disc brooch with scalloped edge and two concentric rings, with a concave moulding with pale grey concretion (this may be corroded elements of a white metal or insert as part of the brooch s decoration). The pin survives, but the latch has broken. Diameter 35mm. A smaller, but similar flat example is known from Colchester, attributed to the 1st century AD (Crummy 1983, cat. no. 85, fig. 14). <10196> (sf.299) - Well preserved cast copper alloy disc brooch with central umbo with (fragmentary) blue glass inset. The circumferential edge is flat on the outer side, with scalloping on its inner edge. The umbo has radiating spiral sun rays with separate raised circles at the end of each ray. The surface is enamelled (blue and black), with traces of glass inlay present. The pin is missing, although the catch plate is intact. Diameter 32mm. Recovered from the upper fills of ditch F.2043, broadly dated to the 2nd 4th century AD, this type of brooch dates to the 1st 2nd century AD. A very similar example was recovered during excavations at Vicar s Farm, c. 2km to the southeast of the site. Jewellery <11164> Pits F.1999/F.2000 ([5970], sl.1279) - Fragment of a of copper alloy cable bracelet manufactured, unusually, from five twisted wires or cables; internal diameter c. 55mm, weight 7g. 3rd 4th century AD. Toilet Implements <11156> Ditch F.1932 ([5813], sl.1228) - Well preserved example of a toilet or cosmetic spoon, the top section missing, but the flat scoop is intact. Probably manufactured by beating flat and finishing the end of a copper alloy rod, this example measures c. 80mm in length and the slightly unequal bowl 5mm across; weight 2g. Roman. Associated with pottery spot-dated to AD. <11200> Ditch F.2092 ([7292], sl.1383) - Set of well-preserved tweezers with decorative medial ridge and flared blades; length c. 48mm, weight 3g. Forming part of a chatelaine, tweezers were used for the removal of unwanted hair (Crummy 1983, 58). Associated with pottery spot-dated to AD. <11260> (sf.90) - Small, relatively well-preserved copper alloy nail cleaner measuring 41mm long. The suspension loop has broken, but is at right-angles to the blade, the points are also missing. Similar to Crummy Type 2a (Crummy 1983, 58, cat. no. 1874); this example is undecorated apart from edge grooves. Textiles <10848> Ditch F.695 ([1861], sl.193) - Long, thin and heavily corroded needle. Although the eye is broken a groove is clearly visible below the eye, suggesting this would have extended above the eye. Length c. 90mm. Crummy describes these needles as more similar to modern needles, her Type 3 needles, and attributes them to the 3rd 4th century AD (Crummy 1983, 67). No associated dating evidence. 243

142 Miscellaneous <10843> Ditch F.680 ([1790], sl.173) - Irregular lump of melted and partially melted copper alloy sheet metal; weight 18g. Roman. Associated with pottery spot-dated to AD. <11067> Pit F.967 ([2778], sl.423) - Bent and folded thin copper alloy sheet fragments; max. Length 42mm, total weight 2g. Romano-British associated with pottery spot-dated to AD and a coin issued in AD. <11080> Pit F.1013 ([2901], sl.442) - Cast short, copper alloy pin with tapering end and rhomboidal head with six facets and fluted head. Length 51mm, shank diameter 5mm, head diameter 12mm, weight 12g. This is an unusual object, relatively heavy and short. Possibly part of a larger, more complicated object it may be a locking pin or similar. Roman. No associated dating evidence. <11093> Pit F.1088 ([3139], sl.489) - Flat copper alloy hook/strap with suspension eyelets/loops at either end (one opened up). The central bar measures 4mm in width and tapers to the flat eyelets. Part of a larger object, this is clearly a connector of some variety and possibly embellished with crenellations and transverse grooves. Length 50mm, weight 2g. Roman. Associated with pottery spot-dated to AD. <11175> Ring-ditch F.2025 ([7004], sl.1306) - Very small, spherical copper alloy object, weight less than 1g. Unidentified, probable droplet. Roman. <11202> Ditch F.2024 ([7314], sl.1389) - Irregular-shaped lump of copper alloy that has been melted; it is uncertain whether this is a casting spill, or a broken element from a larger object. Weight 10g. Roman. Iron A total of 2166 pieces of iron work, weighing 19.98kg, were recovered; 640 (9552g) pieces of metalwork were retrieved from archaeological and natural features, 590 (8921kg) attributed to prehistoric and Romano-British features, with some 710 pieces (1910g) retrieved from cremations and inhumations. The following text provides an assessment and catalogue of selected pieces attributed to or identified as of prehistoric or Romano-British origin and all of the cemetery material. Excluded from this assessment are items recovered during metal-detecting not directly associated with an archaeological feature or post-date the Romano-British period or are clearly modern. An assessment of all nails revealed these were generally broken, with few complete examples, and varied in length from a few millimetres to in excess of 100mm. These are retained in the archive for future research and distribution plotting. The iron work shows variable states of preservation, ranging from excellent (almost no corrosion) to very poor (friable and delaminating, disintegrating). The majority of the iron work falls between these two extremes, but evidence of deterioration post-excavation is evident on numerous pieces. Iron objects from wells and cemeteries are separately listed Ditches F.517 (<10013> sf.16) - Complete circular disc, concreted with rounded raised areas indicative of possible rivets/studs; diameter c. 40mm, weight 27g. Possibly intrusive. F.657 (<10832> [1747], sl.156) - Unequalled triangular-shaped blade fragment from the tip of the blade 60mm long and part of a large implement such as a knife; weight 14g. 244

143 F.662 (<10834> [1760], sl.161) - Fragmentary, but largely complete example of a socketed cleaver (Manning Type 2a); missing its tip, total length c. 195mm, weight 141g. F.674 (<10671> sf.777) - Narrow spiked socketed object, heavily concreted and corroded, length 110mm, weight 42g. Possibly a socketed pilum with broken shaft and missing head. F.711 (<10853> [1907], sl.205) - Large, leaf-shaped spearhead. The surface is heavily corroded and the socket has become detached. The blade is essentially flat with a more pronounced angle on one edge of the blade resulting in an asymmetric appearance. Length c. 210mm, weight 120g. Late Roman/Anglo-Saxon. F.866 (<10895> [2453], sl.349) - Seven items: a) fragment of knife blade (Manning Type 19), length 64mm, weight 14g; b) possible fragment of thin knife blade (Manning Types 4 6?), length 32mm, weight 3g; c) possible fragment of thin knife blade (Manning Types 4 6?), length 38mm, weight 5g; d) broken end of a piece of binding or strapping, or possibly a hinge, with a rounded end, length 35mm, width 22mm, weight 14g; e) fragment of nail, length 41mm, weight 3g; f) fragment of nail shank, length 48mm, weight 4g; g) fragment of clenched nail or possible hook, total length 65mm, weight 9g. F.888 (<11046> [2511], sl.366) - Conical headed suspension ring or loop; the loop has been formed by forging the shaft around and over itself. Length 60mm, loop internal diam. 10mm, loop external diam. 25mm, weight 27g. Possibly Post-Roman/intrusive. F.944 (<10129> sf.226) - Corroded and concreted rectangular cross-sectioned bar. The bar tapers at both end, with one end rounded (the other end is corroded and damaged). Length 65mm, width c. 17mm, thickness c. 10mm, weight 27g. F.952 (<11056> [2724], sl.413) - Four iron objects: a) heavily corroded and disintegrating iron ring (three refitting fragments), internal diam. c. 33mm, external diameter c. 45mm, weight 15mm; b) nail with broken head, heavily corroded, length 30mm, weight 5g; c) heavily corroded nail shank, 45mm long, weight 6g; d) heavily corroded and delaminating triangularshaped nail shank, length 36mm, max. width 11mm, weight 9g. <11057> [2725], sl.413. Five nail fragments and possible plate and blade all heavily corroded and concreted: a) concreted and relatively heavy curving fragment, possibly a broken knife blade, length 53mm, weight 43g; b) heavily concreted rectangular iron plate, with evidence of a possible central rivet/stud, c. 45mm x 55mm, weight 48g; c g) nail fragments ranging in length between 30 and 50mm and in weight between 6g to 9g. F.946 (<11055> [2698], sl.409) - Heavy iron bar, with rectangular cross-section measuring 10mm x 13mm x 40mm, weight 58g. F.963 (<11066> [2771], sl.421) - Two items: a) fragment of curved binding or decorative strap work with probable rivet on upper surface, length 72mm, width c. 30mm, weight 28g; b) delaminating lozenge-shaped lump measuring c. 25mm by 30mm, weight 12g F.972 (<11065> [3195], sl.421) Three objects: a) corroded rectangular plate (blade?) fragment with one rounded corner, length 40mm, width 30mm, weight 17g; b) concreted and corroded fragment of nail shank, length 45mm, weight 8g; c) corroded and delaminating nail, length 45mm, weight 6g. F.998 (<11073> [2865], sl.435) Two delaminating items: a) small nail, 22mm long, weight 6g; b) substantially complete (c. 75%) oval disc or plate measuring c. 40mm by 60mm, weight 39g no obvious evidence for a perforation. <11069> [2795], sl.425. Two heavily corroded and concreted items: a) probable blade fragment, length 50nn, weight 18g; b) nail or stud fragment, length 36mm, weight 10g. F.1003 (<11075> [2869], sl.438) - Large, tapering and narrowing rectangular cross-sectioned bar, rounded at one end with a spike at the other. The bar is S -shaped forged with rightangled bends. Length c. 145mm, weight 89g. Probable wall hook or similar. <11076> [2872], sl.438. Four objects: a) small knife fragment with upward slanting tang (Manning Type 23 or 24), length 49mm, weight 7g; b) large nail, shank broken, length 75mm, weight 23g; c) clenched nail, length 50mm, weight 5g; d) relatively thick, flat rectangular shape fragment 245

144 (lump) with a rounded corner measuring 25mm x 33mm x 8mm, weight 33g. <11077> [2871], sl.438. Four objects: a) possible blade fragment (Manning Type 4 6?), length 42mm, weight 6g; b d) three corroded and delaminating nail shanks, lengths 44 55mm, weights 4 8g. F.1010 (<10699> sf.814) - Concreted curved tapering bar, length 40mm, weight 14g. <10701> sf.818. Narrow, rectangular cross-sectioned and tapering hook, length 70mm, weight 5g. F.1018 (<11081> [2916], sl.447.) - Heavily concreted (numerous pieces of flint and gravel adheres to the surface) and corroded iron bar or clamp. The bar slightly tapers to where it has a slight twisted (broken by twisting/torque action?) short, right-angled bend. Length c. 160mm, width 40 45mm, weight 246g. Found with a nail fragment, length 40mm. F.1031 (<11084> [2954], sl.458) - Four objects: a) large, flat irregular-shaped lump measuring c. 35mm x 40mm x 15mm, weight 47g; b) corroded and concreted nail, length 55, weight 11g; c) corroded and concreted nail shank, length 29mm, weight 4g; d) complete T -shaped headed nail, length 35mm, weight 4g. <11084> [2954], sl.458. Poorly preserved and corroded small-bladed spearhead (Manning Group I). The socket is broken on the side where the two flanges are rolled together during the manufacturing process (Manning 1985, 163) and the head s edges are highly corroded or concreted. Length 74mm, weight 23g. F.1060 (<11089> [3048], sl.472) - Twelve objects: a) corroded and concreted tapering and slightly waisted bar with rounded terminal and transverse break, length 70mm, weight 46g; b) similar preservation condition to a), rectangular bar/plate measuring c. 25mm x 35mm, weight 24g; c) heavily concreted curving, thin fragment, length 40mm, weight 13g, possible blade; d f) three lumps, weight 3 5g; g) parallel side bar with rectangular cross-section tapering to an oblique terminal, length 58mm, weight 9g, possible small tool; h) heavily concreted and corroded small nail, length 38mm, weight 4g; i) large very corroded and concreted nail shank, length 72mm, weight 12g; j) fragment of tapering nail shank with square cross-section, length 32mm, weight 7g; k) very corroded nail fragment with concretions, length 31mm, weight 7g; l) very corroded nail fragment with concretions, length 37mm, weight 8g. F.1103 (<11096> [3174], sl.496.) - Three objects: a) reasonably well-preserved ferrule or similar, length 67mm, socket diam. 22mm, weight 55g; b) nail fragment, length 26mm, weight 4g; c) nail fragment, length 29mm, weight 6g. <11097> [3175], sl.496. Four objects: a) heavily corroded bar, delaminating, length 93mm, weight 27g; b) broken, irregular-shaped plate, most likely from a larger object, such as a knife, preserved curved edge on one side, width c. 45mm, weight 19g; c) tapering rectangular cross-sectioned rod or bar, length 45mm, width 7 9mm, thickness 2 5mm; d) rectangular cross-section bar with projecting lug with possible notch on internal corner, length 42mm, weight 9g. F.1160 (<11133> [4066], sl.748) - Eight heavily corroded and concreted iron objects: a) possible joiner s dog, length 65mm, width c. 20mm, weight 48g; b) possible plate fragment 28mm x 30mm, weight 9g; c) possible hinge fragment or similar fitting with possible bifurcation in the plate (requires x-ray to confirm partial perforation is not a product of corrosion), length 27mm, weight 11g; c) nail or stud head, diam. 20mm, weight 6g; d g) nail shanks, length 35 67mm, weight 5 18g. F.1169 (<11109> [3536], sl.536) - Corroded and concreted metal tube with crescent-shaped blade, similar to a gouge, weight 31mm. <11116> [3536], sl.536. Corroded and concreted metal bar with possible perforation and loop/ring at one end and large open ring/loop at the other with the terminal or remainder missing. Superficially similar to horse gear or potential one half of a shackle, length c. 130mm, weight 63g. F.1272 (<11129> [3707], sl.617) - Irregular-shaped plate fragment measuring c. 35mm x 45mm x 3mm, weight 18g. F.1278 (<1123> [3270], sl.582. Pyramidal-shaped object lacking shank or socket, measuring 40mm long, weight 7g possible end of a large stud or nail. Found with a possible nail shank measuring 82mm long; both heavily corroded. 246

145 F.1442 <11136> [4308], sl.824) - Corroded round bar with a possible pyramidal or faceted head. This is provisionally identified as a pin (Manning 1985, plate-33), length 75mm, weight 8g (requires x-raying to confirm identification). F.1599 (<11143> [4751], sl.961) - Rounded, rectangular object, relatively thin and possibly a fragment from a small bladed implement or tool; length 35mm, width c mm, weight 6g. F.1600 (<11169> [7050], sl.1299) - Two iron objects: a) relatively thick iron collar or spacer with rectangular perforation, diam. 25mm, height 15mm, weight 22g; b) complete nail with slightly domed head, possibly used in conjunction with the collar, length 40mm, weight 8g. <11170> [7022], sl Fragment of a flat oval or round object with a complete shaft with rounded end. The end is possibly perforated and this may be a pendant or decorative piece. Length c. 68mm, weight 6g. F.2040 (<11177> [7067], sl.1317) - Three corroded iron objects: a) triangular-shaped blade fragment with relatively narrow tapering cutting edge. The back-edge is relatively straight with a downward, concave slope to the tip, length 62mm, weight 13g; b c) two nail shanks, lengths 22 45mm, weights 2 3g. <11190> [7249], sl Six objects, all corroded and concreted: a) complete dome-headed hobnail, length 15mm, weight 2g; b) double-spiked loop formed by folding a bar with spiked ends, length c. 50mm, weight 25g. Double-spiked loops are common on Romano-British sites and used for a variety of purposes (Manning 1985, 130); c) tentatively identified as a small chisel square cross-sectioned bar with flanged and relatively sharp tapering blade, length 40mm, weight 12g; d) narrow rectangular cross-section bar with spilt end, length 78mm, weight 8g; e) nail shaft, length 38mm, 10g; f) small object, probably a nail shaft, length 37mm, weight 4g (requires x-ray for confirmation of identity). F.2062 (<11184> [7188], sl.1339) - Heavy circular to rectangular (?) cross-sectioned bar, slightly tapering to a rounded end. Length 12mm, weight 43g. F.2066 (<10674> sf. 780) - Small chisel or punch (Manning Type A23 or A24); length 58mm, weight 46g. Pits F.906 (<10696> sf.806) - Single nail, length 23mm, weight 2g. <11053> [2599], sl.386. Six objects, heavily corroded an delaminating: a) blade or strap fragment 55mm long, with a transverse break across the object one end appears to be rounded, suggesting this is a strap rather than a blade, weight 24g; b) irregular-shaped (tapering) fragment of a relatively thick (c. 3mm) plate or similar, weight 7g, measuring c. 20mm x 30mm; c) probable blade fragment, with a straight back, convex curve to the tip or tang and transverse break across the blade, length c. 40mm, width 35mm, weight 14g; d) nail, length 43mm, weight 5g; e) collection of five to six fused dome-headed hobnails, weight 12g; f) possible scalpel blade, with straight back and slightly concave, edge with slightly bulbous rounded corner on one side and rightangle edge on the opposite corner, length c. 60mm, weight 12g. F.928 (<11052> [2661], sl Fifteen iron objects, corroded and delaminating/crumbling: a) roughly square, pierced plate, the piercing is off-centre, measuring c. 30mm x 28mm, weight 9g, probable backing-plate; c) thin bar or rod, rectangular cross-section and wider rounded end, with broken narrower end, length 30mm, weight 4g; c n) nail fragments, lengths 18 48mm, weights 1 10g. F.967 (<11068> [2778], sl.423) - Three iron objects, heavily corroded and delaminating: a) possible spatulate-headed linch-pin missing its shaft, weight 21g; b) very concreted rectangular cross-sectioned bar, length 105mm, weight 52g; c) fragment of a relatively large socket or sleeve, length 65mm, weight 37g. F.1999 (<11165> [5970], sl.1279) - Four reasonably well-preserved iron objects: a) perforated plate fragment with slightly rounded lateral sides and two partial perforations, 35mm x 35mm, weight 22g; c d) one complete dome-headed nail and two nail shanks, length 40 60mm, weights 5 8g. The function of this piece is unclear it, although it is possibly either a furniture or door fitting or from a larger iron object. 247

146 F.1292 (<11130> [3756], sl.637) - Three heavily corroded objects: a) nail fragment, length 37mm, weight 4g; b c) two lumps, weight 7g and 9g respectively. F.1470 (<11138> [4388], sl.848) - Fourteen objects: a) perforated plate fragment, c. 25mm x 45mm, weight 17g; b) rectangular plate fragment, c. 28mm x 35mm, weight 24g; c l) nails and nail fragments, lengths 18 55mm, weights 2 21g; m) lump, weight 3g. F.1804 (<11153> [5383], sl.1107) - Tapering and rectangular cross-shaped bar or rod. Due to the concretion and corrosion it is not possible to identify this object, length 52mm, weight 20g. F.2539 (<11207> [8148], sl.1544.) - Four iron objects, corroded and concreted: a) fragment of a small- to medium-sized tanged blade with curving back, tang and cutting edge, tip and most of the tang are missing. Possibly a Manning Type 18 (Manning 1985, 117), length 78mm, weight 23g; b) large nail head with part of shaft surviving, length 28mm, weight 12g; c) large nail with flat head, length 115mm, weight 48g; d) large nail, length 48mm, weight 45g. The identification of the last two objects is provisional and subject to x-ray conformation. Postholes F.1187 <11112> [3401], sl.536. Total of twelve objects, including eight un-diagnostic lumps: a) short, triangular-shaped stud, length 40mm, weight 46g; b) small rod or bar measuring 55mm x 6mm, weight 6g; b) small rod or bar measuring 32mm x 7mm, weight 4g; c) small rod or bar measuring 38mm x 10mm, weight 9g. Spreads/Middens F.1167 (<11113> [3402], sl.536) - Two pieces of iron: a) rectangular cross-sectioned rod tapering and bent toward the point. Length c. 75mm, width 5mm, weight 5g; b) thin rectangular cross-section pin(?) fragment, length 57mm, width 3mm, weight 2g. <11114> [3403], sl.536. Seven iron objects: a) small, corroded, roughly square plate, 25mm x 25mm, weight 8g; b) curved, narrow strip, possibly a blade fragment (Manning Type 4 6), length 30mm, weight 4g; c) heavily corroded rectangular cross-sectioned tapering bar, length 40mm, weight 10g; d) irregular-shaped lump (possible corrosion-only product), weight 3g; e) complete nail, length 70mm, weight 14g; f) nail shank, length 35mm, weight 2g; g) probable nail fragment, length 15mm, weight 2g. <11115> [3404], sl.536. Ten pieces of iron: a) heavy and large, irregular-shaped lump weighing 166g; b) small bent narrow and thin bar with rounded ends, length c. 50mm, weight 2g; c) square cross-sectioned rod with one end tapering to a splayed, flat, screw-driver-like blade, length 56mm, weight 4g; d) potentially similar to c) measuring 115mm, with possible rounded terminal, weight 13g; e) as per c), but bent and both ends too corroded/broken to identify if one end is undamaged, weight 8g; f) square cross-sectioned bar tapering to a point, length 112mm, weight 9g; g) fragment of a relatively thick rectangular cross-sectioned bar or rod possibly tapering to a chisel-like terminal, length 70g, weight 16g; h) iron bar measuring 10mm wide by 3mm thick that possesses shoulders at one end tapering to a broken tang or loop, length 77mm, weight 12g; heavily corroded rectangular(?) cross-sectioned bar twisted to form a large ring with overlapping terminals, diameter c. 45mm, internal diameter c. 35mm, weight 20g; i) broken rounded piece of iron, part of a larger object with vertical side 12mm wide possible broken chain link, weight 88g. This is a curious sub-assemblage for which a post-roman attribution may be more likely. Post-Roman Features and Topsoil F.2138 (<10000>, <10001>) - Two hobnails: a) small dome-headed hobnail, clenched, wellpreserved, length 15mm, head diam. c. 7mm; 2) large, very corroded hobnail, length c. 15mm, head diam. 10mm. <10387> sf Small rounded spatulate/crescent-shaped object with tapering shaft, length 58mm, weight 5g. Possible small cosmetic spoon; found in upper fills of ditch F.1992; 2nd 3rd century AD. 248

147 <10406> sf Well preserved small dome-headed hobnail, clenched, length 10mm, head diam. 7mm. Found in the upper fill of ditch F Dated 2nd 3rd century, attributing a date is problematic due to the potential for this being either residual or intrusive. <10435> sf Well preserved dome-headed hobnail, length 9mm, head diam. 8mm. Found in the upper fill of ditch F Dated 2nd 3rd century, attributing a date is problematic due to the potential for this being either residual or intrusive. <10554> sf Well preserved dome-headed hobnail, clenched, length 14mm, head diam. 10mm. Hobnails are ubiquitous finds on Romano-British sites and used in footwear worn by many different members of a community. Consequently, finds of single hobnails, such as the two examples above, and this specimen, provide little further information as hobnails are/were easily shed during the use-life of a shoe or sandal; hobnails were frequently replaced as these were prone to loss and wearing out. Well Assemblages (all metalwork) F.1020 (<10725> sf.842.) - Two objects, both very concreted: a) large suspension loop with rectangular cross-section, length c. 95mm, width 12mm, thickness 6 8mm, weight 67g; b) nail, length c. 50mm, weight 9g. <10726> sf.843. Triangular gouge -shaped object with a possible flange, weight 36, length 40mm, possible tool similar to Manning s Type B48 gouge, but missing the shaft. <10734> sf Small, slightly tapering bar, heavily corroded, with wider round terminal, length 37mm, weight 8g. <11082> [2922], sl.448. Five objects were recovered from this slot: a) reasonably well-preserved S-shaped object. One end expands to a long, relatively thick blade-like end, with the end rounded. The bar is bent through a rounded right-angle and curves upwards to a tapering, squared-off end, possibly a tanged handle, length c. 115mm, weight 46g; b) two refitting elements of a broken bar, heavily corroded, length 73mm, weight 11g; c) small bar or rod, heavily corroded, length 32mm, weight 4g, probable nail fragment; d) similar to c) measuring 25mm, weight 2g; e) heavily concreted and corroded thin and narrow tapering bar with possible oblique edge at one end, length 95mm, width 4 6mm, weight 7g. <11090> [3099], sl.482. Fragment of copper alloy bracelet manufactured from two twisted wires or cables; internal diameter c. 45mm, weight 5g. These forms of bracelets are relatively common and often found as accompanying grave goods. Found in the same feature as copper alloy nails cat. no ; 3rd 4th century AD. <11091> [3100], sl.482. Concreted and irregular-shaped narrow, rectangular cross-sectioned bar or rod, length 115mm, weight 8g. Similar rods or bars have been recovered from other features and wells on the site and their narrowness hints that these are part of a large, composite object or objects. <11099> [3102], sl.482. Three small domed headed copper alloy nails, one nail head has separated from the shank. The largest nail measures 12mm in length and has been clenched. The two smaller nails measure c. 8mm in length with head diameters of c. 3 4mm. Copper alloy studs and nails were used to decorate fabric and wooden items and it is likely, due to their size, that these examples were used on leatherwork. Roman F.1148 (<11127> [3589], sl.521) - An iron object 248mm long with a pyramidal-shaped head and square cross-section, c. 60mm long, The shank, square in cross-section, slightly tapers s towards the both the end and the head. This is a very well-preserved Pilum head with part of the shank surviving (it is not possible to determine if this is a tanged or socketed pilum, although the former is more likely). The pilum is a throwing weapon used by the Roman Army, specifically by legionaries and underwent several design modifications and variations during its use by the army. The heavy bodkin-like head is designed to penetrate shields, armour and combatants, ideally at the same time. As observed by Bishop and Coulston, pila are frequently found on military sites and sometimes confused with drill bits and nails (2006, 75). They further describe pila with heads in excess of 70mm from Newstead dating to the 2nd century AD, with almost identically shaped and sized heads as this example recovered from the well in the headquarters building at Bar Hill, both sites being located on the Antonine Wall (ibid., 129). Although recovered from a feature attributed to the 4th century or later, the recovery of this item from the well should not necessarily be considered residual. F.1236 (<11119> [3599] SF 855, sl.582) - Gilded copper alloy disc brooch, consisting of a raised outer rim, inner concentric ring and raised umbo/boss with five petals. The pin is missing, but part of the latch-plate survives. Diameter, 34mm, weight 17g. Found in the same 249

148 features as the steelyard balance arm described below (cat. no ). Roman. <11121> [3712], sl.582. Eight pieces of ironwork: a) fragment of a rectangular iron plate or buckle measuring 43mm x 38mm. The inner edge, where present, has two notches, with notches also present in the internal angles of the corners, weight 20g; b) tapering rectangular crosssectioned iron bar with angled facet/face with possible ridge-stop or iron sleeve, length 47mm, weight 25g possible small chisel?; c) concreted flat, rectangular cross-sectioned tapering bar 75mm long, 4mm thick, weight 17g; d) complete nail c. 55mm long, weight 10g; e) rectangular cross-sectioned rod 55mm in length, weight 7g possible nail shank; f) curving bar 55mm long, weight 14g; g) thin rectangular cross-sectioned bar, length 52mm, width 4mm, weight 2g; h) relatively thing rectangular cross-sectioned, with irregular taper, with right-angled bend at 70mm, overall length c. 100mm, width 2 7mm, weight 2g. <11122> [3712], sl.582. Well preserved small copper alloy steelyard with arms of unequal length. The terminal is decorated with a lunate/crescent motif and two suspension hooks, one broken. Crummy reports this fulcrum arrangement is unusual, with respective scales for each fulcrum represented on the arm from the example found at Colchester (1988, 99, cat. no. 2508). Due to surface pitting the identification of scales on the example from North West Cambridge is problematic; however, regularly spaced grooves on one side, approximately 4mm apart, may represent the scale and further cleaning may reveal further detail. Steelyard balances are infrequent, but not uncommon finds from Romano-British sites (see discussion). <11124> [3721], sl.582. Five pieces of iron: a) fragment of twisted cable or spirally twisted rod, length 60mm, diam. c. 8mm, weight 10g; b) corroded nail shank with possible broken head/mineralised wood adhering to one end, length 42mm, weight 9g; c) nail fragment with corroded head and partial shank, length 27mm, weight 8g; d) heavily concreted and corroded?triangular cross-sectioned lump c. 45mm long, weight 15g; e) heavily corroded and concreted and partially folded relatively thick piece of iron, rectangular in shape and folded/distorted, thickness 4 5mm, length 40mm, weight 24g. <11126> [3848], sl.582. Two objects: a) complete rectangular cross-sectioned rod tapering to a point at one end and a oblique blade, working edge, at the other, length 100mm, weight 14g probable awl or similar type tool; b) thin, narrow rectangular cross-sectioned bar 4mm wide, 2mm and 70mm long. <11128> [3599], sl.582. Four objects, all heavily concreted: a) bifurcated item with thin blades/plates uniting to form a single collar (?) with a square cross-sectioned tang (broken). It is difficult to attribute a former function to this piece, although it may be part of a handle for a tub or bucket (see Manning 1985, cat. no. P38), length 70mm, weight 36g; b) hook with a wide shaft and irregular-shaped flattened head, possible perforated; length c. 60mm, weight 18g; c) small rectangular plate or blade fragment measuring 20mm x 25mm, weight 5g; d) squareheaded(?) nail fragment. F.2023 (<10667> sf.773) - Small, possible curved blade fragment, length 35mm, weight 3g. <10668> sf Small, broad oval-head tack or fitting; head 13mm x 17mm; weight 3g. <10675> sf.781. Relatively thick, lozenge-shaped lead object with possibly bevelled edges and melted on one side, length c. 64mm, weight 62g. Due to the degree of damage to this object it is unclear what its original shape may have been, although a lozenge shape is more likely. The slight bevelling may indicate that this was a decorative plaque, possibly a well-executed repair patch. <11171> [6084], sl Bent, cast copper alloy bead-headed pin with reel/bead motif with flattened sphere (Type 2; Crummy 1983, 28). Length c. 95mm, weight 6g. 2nd 4th century AD. Found with a probable, corroded coin. <11172> [6084], sl Four heavily corroded iron objects: a) probable loop-headed spike that could be attached to wall or suitable surface or timber (Manning 1985, 129), length c. 85mm, weight 47g; b e) one complete flatheaded nail and three nail shanks, lengths 30 60m, weights 2 13g. <11174> [6088], sl Three heavily corroded iron objects: a) bent nail shank with square cross-section, length c. 65mm, weight 12g; b) small knife or blade fragment, possibly complete with the exception of the missing tang, length c. 58mm, weight 16g; c) possible loop-headed spike, bent almost back on itself and most of the head/loop corroded or missing, straightened length estimated at c. 100mm, weight 23g. F.2044 (<11178> [7113], sl.1324) - Five corroded objects with some delamination taking place: a d) two complete nails, lengths 100mm and 63mm each, weights 31g and 15g respectively, and two nail shanks, lengths 26 75mm, weights 2 22g; e) rhomboid-shaped holdfast with in situ clenched nail or stud providing a material thickness of c. 25mm, weight 16g. (Manning 1985, 132, cat. no s R74 R81). <11179> [7112], sl Two heavily corroded and delaminating iron nails and a reasonably well-preserved cleat: a) complete possible cleat or similar with a broad flat end tapering to a point, length c. 60mm,weight 3g; b c) two nail fragments, lengths 35 45mm, weights 5 10g. The identification of cleats with boots or footwear is problematic 250

149 due to the length if the tangs and consequently a different function, such as fastening wood, for larger cleats has been suggested (Manning 1985, 131). F.2056 (<11192> [7446], sl.1380) - A very well-preserved but broken, square cross-sectioned bar tapering to a V-shaped groove with one break and a rectangular cross-section hooped extension, also broken. Towards the square broken end the bar has a lateral gouge on side and the other side is worn smooth and rounded; similar wear is seen further along the length of the bar on this surface and edges. Length c. 350mm, weight 177g. Possible tong handle (see Jackson 1996; Manning 1985), although another function cannot be ruled out due to the absence of a pivot hole. <11193> [7283], sl Scooped -shaped object with slightly curved and tapering tang, heavily concreted. Identification of this item is problematic and, although speculative, it may be part of a tanged and socketed reaping hook similar to Manning s Type 3 (Manning 1985, 55). Length 98mm, weight 68g <11195> [7279], sl Two corroded and concreted objects: a) triangular piece of iron with irregular transverse break and uneven tape to a rounded end. Possible large blade fragment, length 90mm, weight 35g; b) bar or plate with one straight and on slightly tapering edge. The piece is bent with one end obscured by heavily corrosion, possibly obscuring further detail or representing mineralised material, weight 71g, length c. 75mm; possible strapping. <11196> [7295], sl Eleven iron objects, all corroded an with varying degrees of concretion: a) probable blade fragment (with transverse break across the blade), with a straight back and convex curve to a very short vertical ricasso (it is unclear, however, if this edge is a deliberate design feature or a result of blade failure towards the tip or corrosion) length 40mm, weight 7g; b) two refitting fragments of a fitting with tapering sides and lozenge-shaped terminal, possibly indicating decorative piece no obvious riveting, length c. 60mm, weight 11g; c) rectangular-shaped plate measuring c. 35mm x 55mm x 3mm, weight 23g; d) possible single spiked-loop, length c. 60mm, weight 32g; e) rectangular cross-section bar with stepped end to narrow rounded projection, length c. 90mm, weight 14g; f) possible stylus, round shaft, missing its point with wider, flat smoothing end, length 72g. <11197> [7280], sl Two concreted and corroded objects: a) triangular-shaped plate fragment with possible right-angled lug at one end, length c. 50mm, weight 19g; b) small, possibly riveted, plate or strap fragment, heavily corroded, but in section 3mm thick and 30mm wide, weight 16g. <11198> [7163], sl Ten objects, all corroded and with varying degrees of concretion: a) common form of a small straight-backed tanged knife, the tang possibly rising slightly. The blade is relatively short for its width and has a strong convex cutting edge curving up to the tip, weight 23g, length 78mm Manning Type 12 (Manning 1985, 114); b) possible Manning Type 2a linch pin with crescentric head, length 88mm, weight 42g; c) fragment of a possible joiner s dog (staple used to join pieces of timber) with one arm broken, length 50mm, weight 28g; d) triangular-shaped plate/sheet fragment with possible in situ rivets, length 58g, weight 38g; e h) nails length 42 80mm, weight 9 12g; i) small rectangular cross-section and tapering bar, length 28mm, weight 4g; j) irregular-shaped lump, weight 5g. <11199> [ ], sl Well preserved large nail, 84mm long, with slightly domed head. At the mid-point of the shaft is a potential spacer, or corrosion products that have developed between the gap of two pieces of timber; weight 21g. <11261> [7163], sl Four pieces of copper alloy sheet measuring c. 80mm by 65mm riveted together. One of the sheets is a rectangle 50mm by 55mm in size to which the other two fragments have been riveted, and possesses at least one rivet that indicates the whole assembly was attached to another object. The other two upper sheet fragments have irregular transverse tears and a variable number of missing or in situ rivets, but have clearly defined and shaped edges where these are attached to the central piece. The fourth, and smallest sheet fragment, is riveted to the central rectangular sheet on the underside the side presumably attached to the object to which these sheets were fixed. It is possible that these riveted sheets are part of a decorative scheme, or binding on, for example, a bucket. Cemetery Assemblages (all metalwork) Some 786 pieces (2883g) of ironwork were recovered from cremations and inhumation burials. The vast majority of the overall assemblage consisted of hobnails (393) followed by nails (276). Of the 276 nails, 46 were identified during excavation as coffin nails, with a further 33 nails recovered from cremation burials (see below); Table 3.47 provides a summary of object type 251

150 by funerary rite. Of particular note is the recovery of a potential Hipposandel fragment from a burial and two discs from separate cremation burials. 252

151 Object Burial Cremation Total Bar 2 2 Binding? Blade Bracket? 1 1 Coffin nail Disc 2 2 Hipposandel 1 1 Hobnails Joiner's dog? 2 2 Lumps 9 9 Nail Riveted plate 1 1 Spike/Nail 1 1 Undiagnostic 5 5 Total Table 3.47: Cemetery assemblage by object type. Inhumation F.76 - Reported upon in the evaluation report. Inhumation F A total of 43 coffin nails were retrieved from this inhumation (Table 3.48). All of the nails were corroded with several retaining traces of mineralised wood; corrosion and mineralisation provided estimated plank thicknesses between 8mm and 18mm. No other iron objects were recovered from the burial. Cat No. Context Nail length (mm) Est. wood thickness (mm) Wt. (g) Head Head Comments Mineralised wood Mineralised wood Mineralised wood Mineralised wood Mineralised wood Mineralised wood Mineralised wood Mineralised wood Mineralised wood 253

152 Head Table 3.48: Metalwork from inhumation F.500. Inhumation F Associated with this inhumation were 45 hobnails and 28 nails all probably coffin nails; the single identified coffin nail provided an estimated plank thickness of 15mm (Table 3.49). A copper alloy ring was also found with this burial, although this was not directly associated with the skeleton and lay at the end of the grave a few centimetres from the right foot. <10769> [1206.1], sl.2. Well preserved, but corroded notched copper alloy ring; internal diameter 16mm, external diameter 18mm. Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Nail Nail Nail Nail Head Nail Head Nail Nail Nail Nail Hobnails Nail Hobnails Nail Nail Nail Nail Nail Hobnails Nail Nail

153 Nail Nail Nail Nail Nail Nail Nail Nail Nail Nail Hobnails Coffin nail Hobnails Table 3.49: Metalwork from inhumation F.501. Inhumation F A total of 27 dome-headed hobnails were recovered from this inhumation burial (Table 3.50). Cat No. Context Type Qty. Wt. (g) Hobnails Hobnails Table 3.50: Metalwork from inhumation F.510. Inhumation F Two hobnails, five un-diagnostic lumps and nineteen nails were retrieved from this inhumation (Table 3.51). The recovery of only two hobnails is intriguing as this may reflect local preservation conditions or truncation of the burial and subsequent loss of these items. Nonetheless, it is probable that the burial was accompanied by footwear, as the hobnails were found by the left foot within the probable coffin. The nineteen nails also most likely represent coffin nails with mineralisation providing estimated plank thicknesses of between 12mm and 15mm. Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Nail Nail Nail Nail Nail Nail Nail Nail Nail Nail Head Nail Nail Nail Nail Head Lumps Nail Nail Nail Nail Nail Head Hobnails 2 1 Table 3.51: Metalwork from inhumation F

154 Inhumation F A small assemblage of eight nails and 56 hobnails, the latter divided into two distinct groups indicative of a pair of shoes or boots (Table 3.52). Several of the hobnails were fused and the total number recorded is considered a minimum. Although the nail assemblage is small these are considered to be coffin nails with plank thickness estimated between 12mm and 17mm. Cat No Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Comments Hobnails Hobnails Hobnails Several fused Hobnails Several fused Nail Nail Nail Nail Nail Nail Nail Head Nail Table 3.52: Metalwork from inhumation F.630. Clenchedcomplete Inhumation F Including eight hobnails these assemblage included 29 nails all presumably coffin nails; two identified as such consisted of the nail head only (Table 3.53). Coffin plank thickness is estimated between 15mm and 18mm. Despite the small number of hobnails recovered, these represent the probable inclusion of footwear as grave-goods as part of the funerary rite. Cat No. Context SF no. Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Nail Coffin nail Nail Nail Nail Nail Nail Nail Nail Coffin nail Head Hobnails Hobnails Nail Nail Nail Head Nail Head Nail Nail Hobnails 6 4 Table 3.53: Metalwork from inhumation F.631. Inhumation: F.658 (<10833> [1749], sl.157) - Single corroded nail, length 20mm, weight 7g. The excavation records indicate that this may well be an incidental inclusion rather than a deliberate grave good. 256

155 Inhumation F Consisting of 30 items, 20 were hobnails, nine nails and one object that may have been a piece of binding or similar with a total length of 75mm (formed from two refitting pieces cat. no and 10840; Table 3.54). Unfortunately, the surviving elements of the object were insufficient to aid further identification or function. The recovery of a single group of 20 hobnails may represent the deposition of only one shoe or piece of footwear as part of the funerary rite. Cat No. Context Type Nail length (mm) Qty. Wt. (g) Nail Nail Nail Nail Nail Nail Binding? (inc ) 75 1 (2) Nail Hobnails Nail Nail Table 3.54: Metalwork from inhumation F.666. Inhumation F An assemblage of only eleven objects, this included hobnails, nails, lumps and a bar or rod (Table 3.55). The small number of items recovered from this inhumation possibly reflects local preservation conditions, later truncation or disturbance of the burial, or a combination of both. Nonetheless, the presence of hobnails further attests to the inclusion of footwear as grave good and the nails are indicative of a coffin burial. Cat No. Context: Type Nail length Qty. Wt. (g) (mm) Hobnails Nail Nail Hobnails Hobnails Lumps Bar Table 3.55: Metalwork from inhumation F.848. Inhumation F A relatively large assemblage of 97 items, objects included as grave goods include at least one knife (or bladed implement; cat. no ), 47 hobnails, a riveted plate measuring c. 20mm x 30mm and a large nail (cat. no ) with a large triangular-shaped head, the head measuring 18mm wide, 32mm long and the shaft 60mm (Table 3.56). This presence of this object and riveted plate (cat. no ) is of note as the former object may be a linch pin and the later part of a larger object. As such, although the majority of nails may be coffin related (two nails have traces of wood mineralisation, with estimated plank thicknesses between 12mm and 18mm), some nails may originally have been associated with other items included as grave goods, such as the footwear evidenced by the present of the hobnails or the knife, for example a wooden box (see below ref). The inclusion of a wooden box may account for the presence of the riveted binding or strapping and possible key fragment (cat. nos and 10995). The excavation records support the idea that there was possibly a wooden box originally placed to the right side of the head or shoulder. The distribution of the material does not indicate that any of the other metal items were contained within this box. The linch pin and blade were found together over the chest and are probably grave goods. There appear to have been two shoes associated with the left (<11016>) and right (<11017>) feet. 257

156 Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Nail Nail Nail Nail Nail Undiag Bar Lumps Undiag Binding 2 13 Comments 10mm x 25mm with shoulders and tapering point. Key or nail frag right-angled projection; 22mm long, 20mm rightangle projection, 13mm tine. c. 70mm long and 20mm wide, riveted. Illustrate Riveted plate mm x 22mm Nail Nail Nail Nail Nail mm x 20m rounded head Nail Nail Mineralised wood Nail Blade Nail Head 1 5 Small tanged blade frag 38mm long, 19mm wide, transverse break across blade Nail Large triangular shaped head; 18mm wide x 32m long. Linch pin? Nail Nail Binding? Nail Nail Nail Nail T-shaped head Nail Nail Nail Nail Nail Nail Nail

157 Nail Head Nail Nail Hobnails Hobnails 16 8 Table 3.56: Metalwork from inhumation F.849. Inhumation F A seemingly large assemblage of 119 items, 106 objects consisted of hobnails, with two clusters of 40 and 59 hobnails, highly indicative of a pair of shoes or boots, probably of adult size (Table 3.57). As well as twelve nails (one providing an estimated coffin plank thickness of 15mm) one un-diagnostic piece (cat. no ) is intriguing. This item, despite severe corrosion, has a smooth and rounded mid-rib on both sides. In profile, it has a distinct arrowhead shape although this identification is unlikely as it most likely relates to the coffin construction. Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Nail Nail Nail Qty. Wt. (g) Comments 20mm x 23mm triangular/rectangular head Nail Nail Nail Nail Clenched-complete Undiag Nail Nail Nail Nail Head Nail Hobnails Hobnails Hobnails 7 8 Table 3.57: Metalwork from inhumation F.856. Possible arrowhead frag 15mx22mm rectangular head Inhumation F The assemblage recovered from this inhumation consists exclusively of nails, one providing an estimated coffin plank thickness of 17mm (Table 3.58). Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Nail Nail Nail Nail Nail Nail Nail Table 3.58: Metalwork from inhumation F.858. Inhumation F Consisting of 116 items this assemblage includes 54 hobnails, 24 found beside the left foot and 17 beside the right foot (Table 3.59). Recovered from this burial is a broken but relatively heavy and looped object provisionally identified as part of a Hipposandel (cat. no ) and a bracket (cat. no, 11042) that is curved at one end and 259

158 folded back on itself with a central vertical bar. Broken just beyond this central bar, the bracket measured 58mm long, 20mm wide, with an overall bracket width of 23mm. The loop is much thicker (reinforcing?) and is 15mm thicker than rest of band/strip c. 4mm thick and heavier than the rest of the bracket, indicating this was a weight or load-bearing object (possibly refits with or joins with the binding cat. no ). In addition, two possible joiner s dogs, un-diagnostic lumps and 53 probable nails were found, these providing an estimated plank thickness of between 15mm and 20mm. Cat No. Context Type Nail length (mm) Est. wood thickness (mm) Qty. Wt. (g) Comments Nail Head mm round head Nail mm x 2mm square head Joiner's dog? Nail Nail Nail Nail Nail Head Nail Nail Nail Nail Nail Nail Nail Nail mm x 18mm rectangular head 25mm x 25mm rounded head 26mm x 30mm rectangular head 25mm x 23mm round head Nail Nail Nail Nail Clenched-complete Lumps Joiner's dog? Nail Clenched-complete Nail Nail Nail Nail Hobnails Hobnails Left foot Nail Clenched-complete Nail Head Nail Hobnails Right foot - two fused Bracket? 1 42 Illustrate: Part of 11044? Hipposandel 1 37 Illustrate: Frag with loop (diam. 22mm) length 30mm Binding? 3 33 Part of 11042? Nail Lumps Nail mm x 18mm 260

159 rectangular head Nail mm x 18mm rectangular head Nail Nail Nail Nail Nail Nail Clenched-complete Nail Nail Nail Head 1 6 Domed 27mm x Nail Head Nail Nail Nail mm rectangular head 27mm x 23mm rectangular head 25mm x 22mm rectangular head Nail Nail Mineralised wood Nail Nail Nail Large head-short shaft Nail 4 4 All fused Nail Hobnails 4 4 Three fused Hobnails 5 5 Three fused Hobnails 3 2 Table 3.59: Metalwork from inhumation F.865. Inhumation F.2666 (<11210> [8671], sl.1966) - Undiagnostic fragment weighing 10g. Cremation F Consisting primarily of 20 hobnails a fragment of a knife or tool blade was also recovered from this cremation, length 40mm, weight 6g (Table 3.60). The iron items were scattered around the grave without obvious patterning. Cat No. Context Type Nail length Qty. Wt. (g) (mm) Hobnails Hobnails Hobnails Blade Hobnails Hobnails Hobnails Hobnails Hobnails Hobnails 1 2 Table 3.60: Metalwork from cremation F.506. Cremation F.507 (<10783> [1343], sl.8) - Nail: length 55mm, weight 19g. <10783> [1343], sl.8. Nail: length: 50mm, weight 6g. Cremation F.508 (<10785> [1328], sl.9) - Corroded nail, 45mm long, weight 9g. <10785> [1328], sl.9. Small, thin corroded disc of unknown function, diameter c. 16mm, weight 2g. 261

160 Cremation F.549 (<10030> sf. 71) - Possible spike or nail, heavily corroded: length 58mm, weight 14g. Cremation F.910 (<11049> [2623], sl.388) - Fragment of a rosette brooch with decorated arched bow and circular plate with four concentric rings, also internally decorated with transverse grooves/ridges; weight 17g. A Similar example was recovered from Saham Toney (Brown 1986, cat. no. 30 & fig. 11) and dates to the mid-1st century AD; Late Iron Age/Romano- British. The pottery vessel containing the cremation and brooch is spot-dated to AD. Cremation F.1958 (<11157> [5868], sl.1246) - Fragmented and corroded iron disc with flat underside and convex surface. Diameter c mm, weight 73g. A similar item was recovered from the Butt Road cemetery, Colchester (Crummy 1983), although a function for these items is unclear. The pottery vessels from this cremation indicate a mid-1st century AD date and the location of the disc suggests that it probably represents a deliberate grave good. Cremation F.2109 (<11203> [7354], sl.1404) - A single, small hobnail, weight 1g. The excavation records indicate that this material may well represent incidental inclusions rather than deliberate grave goods. Cremation F Thirty nails and hobnails were recovered from this cremation burial. The presence of these nails suggests that the cremation or other accompanying grave goods were accompanied by a wooden box or casket, but which did not contain the actual cremated remains. Lead and White Metal Two steelyards weights (were recovered during metal-detecting. Steelyard weights are frequently found on rural Romano-British sites and conform to divisions or multiples of the uncia, one twelfth of the Libra roughly equivalent to 27 28g. Used in conjunction with a steelyard balance these weights were standardised within the system of weights and measures introduced during the Romano-British period and attest to commercial-like activity nearby. <10107> sf Bi-conical steelyard weight with remnant copper alloy suspension loops at either apex weighing slightly more than a Roman Sextans (2oz), one sixth of a Libra. Dimensions: diameter 25mm, height c. 25mm, weight 58g. Found in close association with pit F.1222, this feature has been attributed to the 3rd century AD. <10707> sf Bi-conical steelyard weight with remnant iron suspension loops at either apex weighing slightly more than a Roman Triens (3oz) and less than a Quadrans (4oz), or four unciae. Dimensions: diameter 30mm, height 40mm, weight 100g. Recovered from the upper fills of ditch F.824, this weight can be attributed to the 2nd 3rd century AD. Evaluation-phase Finds In addition to the metalwork catalogued and described above, a further 88 items were recovered during the evaluation phase of excavation; eight were of copper alloy, the remaining 80 of iron. During the evaluation two significant groups of artefacts were identified. The first of these was derived from a densely intercutting cluster of probable settlement-related features, whilst the second group was associated with a coffined inhumation (F.76 mentioned above). Both groups are Romano-British in date. Of the iron objects recovered during the evaluation, the majority of them some 23 items comprise undiagnostic nail fragments. Amongst the remainder, however, significant Romano-British finds include a probable candlestick fitting from pit F.163 and a possible arrowhead from ditch F.1003 (F.87). Eleven items were recovered via the metal-detecting of a 262

161 discrete cluster of intercutting features in Trench 158. Notably, the group included two nummi of Constantine II (born AD 316, Emperor ). Discussion The metalwork assemblage from this phase of investigation is somewhat challenging with respect to its importance and significance. Just under half (47%; 271) of the non-cemetery assemblage was recovered from ditches, with wells (163) and pits (96) forming the bulk of the remaining assemblage. The nineteen objects, including eight bars and a rod recovered from spread F.1167 may represent midden material. Interestingly, only three nails were retrieved from this spread. The number of nails from non-funerary contexts comprises the largest single artefact category recovered from features, amounting to some 25% of the overall ironwork total, with 287 nails (49.9% of the Romano- British assemblage) recovered from features attributed to the Romano-British period. As such, this value or quantity is not unreasonable, although 144 nails were recovered from boundary and enclosure ditches (to which should be added the various nail fragments found recovered during the evaluation), with 70 retrieved from wells. Regardless of their final deposition context, nails demonstrate the nearby presence of reasonably large structures and more finely constructed items. Of 287 nails recovered from the site, 109 exceed 50mm in length, with a maximum length of 115mm (Fig. 3.47; the vast majority of nails were fragmentary and lengths should be considered minima). In addition to the nails, several joiner s dogs, tacks and spikes, a clamp and a cleat were also found, reflecting the use of different fixtures for construction. Less clear in terms of function are objects such as the potential wall hook (ditch F.1003) and 21 fragments of iron plate, 45 bar fragments and three straps. Of these, the 17 bar and nine plate fragments recovered from the wells most likely relate to their construction, with one artefact including an in situ nail (F.2044; cat. no ). The recovery of a punch, two gouges and a chisel, although widely distributed across the site reflects woodworking and carpentry. Of note, is the apparent lack of hammers and saws from the assemblage; however, local preservation conditions may have resulted in the total loss of the latter of these items, and only one copper alloy saw blade has been recovered from a secure archaeological context from the Cambridge environs (Collins 2012). The recovery of the gouges is of particular interest as these may attest to skilled woodworking, including lathe turning or more elaborate wood carving (Ulrich 2007, 29). Similarly, some of the generically described plate fragments, such as the example recovered from well F.1020, may be plane blades, although this interpretation is currently speculative. Of the 20 blade fragments (and two examples found during the evaluation F.113) and large knife (and a further potential eighteen blade fragments recovered from archaeological features), all but two of these were found in Romano-British features, 26 from ditches and six from wells and the large knife from a pit (F.939). Consisting primarily of tanged forms (see Manning 1985), the examples from North West Cambridge are very similar to the numerous examples recorded from rural and urban sites and are ubiquitous tools used for everyday activities. The recovery of a cleaver (ditch F.662) further attests to the range of activities taking place and compliments the 263

162 faunal evidence for butchery and carcass dismemberment. A possible scalpel blade was also found in pit F.906, but this interpretation hinges on the shape and size of the artefact. F.548 F.564 F.586 F.614 F.996 F.1020 F.1148 F.1236 F.1364 F.1402 F.1821 F.2023 F.2044 F.2056 Bar Blade Cable 1 Cleat 1 Fitting 1 1 Gouge 1 Handle? 1 Hobnail 2 Hook 1 Joiner's dog? 1 Ladle 1 Lynch-pin 1 Lump Nail Nail? 2 Pilum 1 Plate Plate and nail 1 Scoop' 1 Spike 2 1 Strap 1 Stylus? 1 Suspension ring 1 Tack 1 Tong handle 1 Tool/handle 1 Tool? 2 Total Table 3.61: Ironwork from wells. In terms of ironwork assemblages from specific feature categories, the wide distribution of material across the site in pits and ditches provides evidence of probable rubbish disposal and incorporation of midden material as these were backfilled or levelled. It would, however, be beneficial to plot the material against evidence for metalworking debris so that due consideration can be given to arguments for smithing and recycling of ironwork in North West Cambridge. The well assemblages are informative, not least of all as secure deposits, as the quantity and variety of metalwork is limited. Only two well have more than ten pieces of ironwork, with nails the single largest 264

163 artefact category, with the aforementioned bars and plate fragments the next largest groups, followed by blade fragments (Table 3.61). This general lack of ironwork contrasts markedly with the quantities of pottery and animal bone recovered from the wells and may indicate that the wells were backfilled with midden material as opposed to special closure type deposits. It is interesting to note, therefore, that a potential stylus, pilum (see below), small cable fragment, handles and one half of a set of large tongs were found in these wells, in addition to a suspension ring and a lynch-pin (a second possible lynch-pin was retrieved from pit F.967 see below), and demonstrate that non-well related material was being deposited in these features. A significant find from well F.1236 is a substantial fragment of a steelyard arm. Found on numerous rural sites, albeit it in relatively small numbers, steelyard arms provide good evidence for the use of officially sanctioned weights and measures (Appleby in Evans et al. 2013). The recovery of the two biconical lead weights during metal-detecting further reinforce this interpretation and may provide proxy evidence for periodic market-like activity not dissimilar to that which most probably took place at Vicar s Farm, located less than 2km southeast of the site. At Vicar s Farm several biconical lead weights and a hook from a probable balance were also recovered. As commented upon by Crummy, it is difficult to assess the extent of either military or official involvement in the economy at Vicar s Farm, if any, but it is likely that both sites were supplying surplus produce to the urban community in Cambridge, if not further afield. The presence of several spearheads and a potential ballista bolt-head and arrowhead fragment should not be interpreted as direct military involvement in the organisation or running of the economy on the site. Military weaponry is frequently found on non-military sites, for example, the 23 spear and boltheads found during metal-detecting and excavation of the mausoleum and villa site at Bancroft, Buckinghamshire (Skinner 1994, 339), the examples from Mucking, Essex (Mould in Lucy et al. forthcoming), and the several arrowheads and spearhead found at Silchester, Hampshire (Richards 2000, 374). It should be noted, however, that the presence of these artefacts does not preclude periodic visits by the military, or even the disposal or recycling of ironwork on the site (see Timberlake, below). We should also not discount the possibility that the leaf-shaped spearheads and possible arrowhead fragment may have been used for hunting, and those found during metal-detecting may even be Post-Roman in date, for example the Medieval spearhead recovered from quarry pit F Nonetheless, as an artefact group these form a significant percentage of the identified material after the blades/knives and structural fittings (nails etc). Of the remaining ironwork the recovery of a horseshoe is of note as these are infrequently found Romano-British sites and debate continues as to whether such finds should be attributed to this period (see Challands in Evans et al 2013; see also the discussion below regarding the possible hipposandal fragment). The recovery of two crudely manufactured potential lynch-pins is of interest as these two objects (if correctly identified) are used to secure wheels to cart axles through the hub. The presence of several pieces of undiagnostic binding or strapping may also relate to the construction/dismantling of cart fittings, but at present this interpretation is 265

164 purely speculative. One object that may also be related to horse gear is a looped object recovered from ditch F.1169 (cat. no, 11116). This is provisionally identified as part of a slave shackle (Manning 1985) due to the shape and narrow cross-section of the metal; however, this object may be part of a bridle or snaffle bit. The non-cemetery copper alloy assemblage detailed above is very small in quantity but includes a relatively high number of personal items, notably the two finely decorated disc brooches, a cable bracelet, toilet implements and a substantially complete needle. Objects recovered from secure contexts suggest these were either chance losses or incorporated into back-fills, including the fluted-headed pin found in pit F The two brooches and nail cleaner found during metal-detecting, however, may represent post-depositional disturbance and loss of either inhumation or cremation burials either through ploughing or quarrying activity. The two brooches also hint at a community that had access to high status goods, although these could also be chance losses. The cemetery assemblage is in many ways less problematic. The recovery of hobnails is an expected feature of formal burials, with shoes and footwear frequently found as accompanying grave goods in cemeteries. The recovery of only one or two hobnails from a burial should also not be considered problematic, especially as several pieces of shoe leather have been recovered from waterlogged deposits on the site, some of which are sufficiently preserved to provide information on how the shoes were nailed (see leather report, this volume, by Mould). The recovery of single hobnails from across the site also attests to how these objects were lost through everyday use. In respect of other finds from the cemetery, the recovery of a copper alloy ring from inhumation F.501 may represent either an object placed near to the body, but most likely outside of the coffin, or even a toe-ring. This, however, must remain speculative and it is interesting to note how relatively few dress or jewellery items were found accompanying either the burials or cremations. Of particular interest were two burials, cremation F.865 and inhumation F.865. With respect to the former, the presence of a large number of nails may indicate the burial deposit was accompanied by or associated with a wooden box or container, although the lack of directly associated animal or pottery remains with the nails would suggest that if this was a box it contained perishable items (either food-stuffs or textile) that have not survived. An inhumation burial, F.849, was also probably accompanied by a box or casket, possibly with a key and also accompanied by a possible lynch pin and knife blade included as a grave goods. Cremations and burials with accompanying caskets or boxes are found in several cemeteries in Britain, such as at Butt Lane, Colchester (Crummy 1983) and Folly Lane, St Alban s, Herts. (Niblett 1999). The large nails found with inhumation F.865 raises two linked questions: are these coffin nails, or was the coffin constructed from recycled wood? In answering the second question first, if recycled wood was used for the construction of the coffin, this would have involved rather thick and relatively heavy planks or pieces. If correct, by default, these nails will have been reused, possibly still in situ when the wood was recycled. An alternative 266

165 explanation for these large nails is that the body was placed in a coffin-like container or on bier constructed from robust pieces of re-used timber. Again, this interpretation is speculative, but the nails found in this inhumation are much larger than reported coffin nails (see Manning 1985). Nonetheless, with the reported plank thickness estimated up to 20mm thick, these are somewhat thinner than the reported wood thicknesses for coffins excavated at Ville-Sur- Retourne, France (Stead et al. 2006, 95) where the wood used in coffin construction was between 20mm and 40mm thick. Interestingly, inhumation F.865 also contained a possible hipposandal fragment and a bracket-like iron fragment. If the identification of the hipposandal fragment is correct, this attests to the use of shod horses on metalled surfaces. The bracket is more challenging, as a function for this piece cannot be positively identified. Again, similar items were found in graves at Ville-Sur-Retourne, such as the 24 objects described as d-shaped ferrules from grave V2 (ibid., fig. 111). Numerous pieces of the assemblage (c. 70 objects), the vast majority of them iron, require x-raying to aid further identification and function (and to provide a permanent record of some fragile and degrading objects. Assessment of the iron assemblage would benefit from further detailed analysis following x-raying of the material, with the brooches also examined by an external specialist. Iron Age Coins John Talbot One definite and one possible Iron Age coins were recovered: <10664> sf.769 F.2025: Trinovantes/Catuvellauni. Copper alloy unit BMC , ABC 2694, VA 1820 Obverse: Head right showing part of pelleted border Reverse: Horse or ram left, above pellet-in-ring and small device of three lines. Below is inscribed VIR Diameter 17mm. Weight 2.17g The coin is generally attributed to Tasciovanus and the reverse is very close in style to BMC which is inscribed Tasci. The inscription VIR is assumed to relate to Verulamium. There are 26 records of the type in the Celtic Coin Index; seven came from the Harlow Temple site and the seven others with a provenance from Buckinghamshire, Hertfordshire, Bedfordshire and Northamptonshire. The approximate dating is 20BC 10AD (Haselgrove Phase 7). <10079> sf.121: Irregular copper alloy disc Diameter 13mm by 17mm (approx. 2mm thickness) Weight 2.62g One side of the coin has a horizontal line above which to the right is an S -shape. These two marks are evocative of an exergual line and a tail, but there are no significant traces of design to the left of the tail. The other side is blank. The irregularity in the shape of the disc appears to be due to the remains of a casting sprue. It is questionable whether this is a coin, it is too small to be a British Potin and British LIA struck bronzes do not show residual casting sprues. Continental bronzes do sometimes show sprues as can be seen in BMC volume III. 267

166 Roman Coins Nick Wells Seventy-two Roman coins plus one possible Roman coin were recovered during the excavation phase (Tables ; Fig. 3.48). The assemblage is rather unusual, though, and deserves some comment. The Romano-British period can be divided into two; a small group from the 1st/2nd century, and a much larger group from the late-3rd to the mid/late-4th. The first group, stretching from Domitian (81 96) to Antoninus Pius ( ), is not unusual, though interestingly, three of the uncertain 1st/2nd century dupondii/asses have been cut up clearly being prepared to act as blanks for copying. This is paralleled by one of the late-3rd century radiates and four of the uncertain 3rd/4th century copper alloy coins which have been worn and/or filed down to create similar blanks for copying. All-in-all this suggests that the site might have been a centre of production for either the prolific radiate or mid-4th century copies common on all Romano-British sites (Fig. 3.49). Period No Notes 1st/2nd Century 9 Including one late 1st century & three 2nd century. Late 3rd Century 14 4th Century 41 3rd/4th Century 8 Either a 3rd/4th century coin or Uncertain 1 late 17th century token. Table 3.62: Breakdown of Roman coin. Another interesting feature of this assemblage is that the number of 4th century coins outnumbers late-3rd century radiates by almost 3:1. This is unusual as one would expect at least as many radiates as 4th century nummi. In a similar vein, coins dating to before 330 usually a small proportion of the total 4th century assemblage comprise 36% of the 4th century total. This is unusually high and suggests major activity at the site in the early/mid-4th century. Only three post-348 (the date of a major reform of the coinage) coins were found, the latest being a SPES REIPVBLICE type (355 61). There were none of the common House of Valentinian (364 78) and House of Theodosius ( ) types, suggesting that the site had gone out of use by the 360s. 268

167 19th Century Quarries 1st-2nd century coins 3rd-4th century coins 0 metres 100 Figure Distribution of Romano-British coins

168 19th Century Quarries Archaeological features 1st to 2nd century 3rd to 4th century metres Figure Distribution of evidence for Romano-British coin copying

169 Cat. Feature SF Issuer Denomination Reverse Type Issue Date Mint Note Domitian As Unk Rome Subsoil Subsoil Furrow Subsoil Uncertain 1st/ 2nd Century Uncertain 1st/ 2nd Century Uncertain 1st/ 2nd Century Uncertain 1st/ 2nd Century Uncertain 1st/ 2nd Century Dupondius/As Unk Unk. Dupondius/As Unk Unk. Dupondius/As Unk Unk. Dupondius/As Unk Unk. As Unk Unk Antoninus Pius Dupondius BRITANNIA SC Rome Trajan Dupondius Unk Rome Subsoil 32 Antoninus Pius Sestertius Unk Rome Probus Radiate AEQVITAS AVG Rome Cut up and worn - probably to create a copy. Cut up coin for creating blanks for copies - originally probably a 1st/2nd century dupondius/as. Cut up worn 1st/2nd century dupondius/as - for creating copies. Heavily corroded Subsoil 85 Victorinus Radiate Unk Irregular Copy Subsoil 119 Uncertain radiate Radiate Unk Unk. Probable copy Subsoil 174 Divus Claudius II Radiate CONSECRATIO Eagle 271 Rome Subsoil 202 Victorinus Radiate Unk Unk. Possible copy.

170 Tetricus II Radiate Unk Unk Allectus Radiate PAX AVG C Mint Subsoil 443 Uncertain radiate Radiate Unk Unk Subsoil 561 Uncertain Radiate Radiate Unk Irregular Copy Topsoil Topsoil Topsoil (1) (2) NS1/ 133 H15/ A65 H16/ B52 Claudius II Radiate Unk Unk. Probus Radiate COMES AVG Lyon Victorinus Radiate INVICTVS Trier 1236 Claudius II Radiate Unk Unk. Obverse & reverse striated & worn - to create a blank for copying Uncertain radiate Radiate Unk Irregular Copy Subsoil Subsoil Subsoil Subsoil 324 Uncertain 3rd/ 4th Century Uncertain 3rd/ 4th Century Uncertain 3rd/ 4th Century Uncertain 3rd/ 4th Century Uncertain 3rd/ 4th century Uncertain 3rd/ 4th Century Uncertain 3rd/ 4th Century Unk. Unk Unk. Unk. Unk Unk. Unk. Unk Unk. Unk. Unk Unk. Unk. Unk Unk. Unk. Unk Unk. Unk. Unk Unk. Cut up blank, very worn. Worn and striated - a blank for a copy. Possibly not a coin - very thin. Worn and striated - a blank for a copy. Very corroded. Possibly a cast copy. Cut up worn coin to be used as a blank for

171 11378 Topsoil H14/ D Subsoil Subsoil 75 Uncertain 3rd/ 4th Century House of Constantine ( ) House of Constantine ( ) copying. Unk. Unk Irregular Copy. Nummus Nummus Subsoil 78 Constans Nummus VIRTVS EXERCIT Standard PROVIDENTIAE CAESS VICTORIAE DD AVGGQ NN Irregular Copy Unk Trier Subsoil 80 Constantinopolis Nummus Victory on prow Trier Subsoil 84 Constantine I Nummus CONCORD MILIT London Subsoil 86 Helena Nummus PAX PVBLICA Trier Subsoil 92 Magnentius Nummus Subsoil 94 Constantine I Nummus Subsoil 99 Constantius II Caesar Nummus SALVS DD NN AVG ET CAES SOLI INVICTO COMITI GLORIA EXERCITVS 2 standards Unk London Lyon Subsoil 100 Urbs Roma Nummus Wolf & twins Trier Subsoil 101 Helena Nummus SECVRITAS REIPVBLICE Trier Possible good copy. Possible good copy Subsoil 104 Uncertain 4th century Nummus Unk Unk. Possible copy Subsoil 105 Constantine II Caesar Nummus Subsoil 115 Constantine II Caesar Nummus Subsoil 107 Constantine I Nummus Subsoil 110 Constantine II Caesar Nummus PROVIDENTIAE CAESS GLORIA EXERCITVS 1 standard VICTORIAE LAETAE PRINC PERP VIRTVS EXERCIT Standard Trier Trier Trier London Subsoil 114 Constantinopolis Nummus Victory on prow Irregular Copy.

172 10075 Subsoil 117 Crispus Caesar Nummus Constans Caesar Nummus Constantine II Caesar Nummus BEATA TRANQLITAS GLORIA EXERCITVS 1 standard GLORIA EXERCITVS 2 standards London Irregular Copy Trier Subsoil 245 Urbs Roma Nummus Wolf & twins Lyon Subsoil Subsoil Subsoil Subsoil Subsoil 302 Constantine I Caesar Constantius II/ Constans Constantius I Caesar House of Constantine ( ) House of Constantine ( ) Nummus Nummus Nummus Nummus Nummus GENIO POP ROM Turreted VICTORIA AVGVSTORVM GENIO POPVLI ROMANI GLORIA EXERCITVS 2 standards GLORIA EXERCITVS 2 standards 307 London Lyon Unk Irregular Copy Irregular Copy Subsoil 304 Constantinopolis Nummus Victory on prow Irregular Copy Subsoil Subsoil 306 House of Constantine (348 64) House of Constantine ( ) Nummus SPES REIPVBLICE Unk. Nummus Subsoil 307 Crispus Caesar Nummus Subsoil 384 Constantine I Nummus Subsoil 387 Constantine I Nummus Subsoil 550 House of Constantine (348 64) Nummus GLORIA EXERCITVS 1 standard CAESARVM NOSTRORVM VOT X VICTORIAE LAETAE PRINC PERP SOLI INVICTO COMITI FEL TEMP REPARATIO Galley Unk Lyon Irregular Copy Arles Unk.

173 Constantinopolis Nummus Victory on prow Irregular Copy House of Constantine ( ) Nummus VICTORIAE LAETAE PRINC PERP Irregular Copy Theodora Nummus PIETAS ROMANA Trier Constantinopolis Nummus Victory on prow Trier Urbs Roma Nummus Wolf & twins Unk Topsoil Topsoil Topsoil EW02/ 50 H08/ B15 H12/ D40 Theodora Nummus PIETAS ROMANA Unk. Uncertain Tetrarchic Subsoil Crispus Caesar Nummus Nummus Unk Unk. Helena Nummus PAX PVBLICA Trier VICTORIAE LAETAE PRINC PERP Irregular Probably a GENIO POPVLI ROMANI type. Copy. Sample Constans Nummus VIRTVS AVG NN Trier Sample Table 3.63: Roman coins. Divus Constantine I Constantine II Caesar Constantine II Caesar Nummus Quadriga Trier Sample 97. Nummus Nummus BEATA TRANQVILLITAS GLORIA EXERCITVS 2 standards Trier Evaluation Unk. Evaluation

174 Worked Stone Objects Simon Timberlake The worked stone consists of fragments of shale bracelets, some partly shaped and natural stones found in a cremation and probably used as gaming pieces, rotary and saddle querns, anvils/ mortar stones and whetstones. Shale Bracelets Two fragments of hand-carved (and partly decorated [e.g. <452>]) round shale bracelets made of Kimmeridge Shale were found within a Romano-British inhumation burial and at the base of a box-lined Romano-British well. The shale bracelet finds from this site are interesting in terms of what it tells us about the geographical extent of the fashionable use of this material and the links and trade networks with the industry that produced them based near to Kimmeridge Bay in Dorset. A number of these shale bracelets have now been found at Iron Age Romano-British settlement sites within the Cambridge area (e.g. Trumpington Meadows (Patten 2012), Anglesey Abbey (Slater & Tabor 2012), and Babraham (Timberlake & Amour 2007)). <895> F.865 ([2498]) - Fragment of shale bracelet found with an inhumation burial in C2. This item was recovered from the general fill of the grave, and whilst it may be a grave good this cannot be conclusively demonstrated. Made of a dark grey black shale/mudstone (probably Kimmeridge Shale from nr Kimmeridge, Dorset), this 30mm long broken fragment of bracelet has a smooth/well polished surface and rounded oval x-section (7mm high and 6mm wide with an angular/flattened facet on its internal circumference) and weighs just 4g. This may have had an original diameter of between 60 70mm.The production of lathe-turned and hand-polished shale bracelets from the Blackstone Bed of the oil-rich Kimmeridge Shale began in the Late Iron Age from cliff sources at Kimmeridge and Brandy Bays on the Dorset coast ( Cunliffe 1978). This industry becoming more prominent during the Romano-British period with lathe-turned examples of bracelets becoming common from the 1st century onwards (Clark 1986). This example from North West Cambridge may well be hand carved and polished, and might usefully be compared with the similarly delicate Iron Age example from Trumpington Meadows. In contrast to this example was the thicker lathe-turned shale bracelet (probably also from Kimmeridge) recovered from Burial 45 at Babraham (Timberlake & Armour 2007). The finer Kimmeridge shale bracelets are commonly found broken, and must have been much more fragile. <452> F.608 ([1338.1]) - Another small (16mm long) fragment of a split and broken hand-worked bracelet (probably originally 6 7mm in diameter; weight < 1g), most likely made of lignite or cannel coal. This item might also come from the Kimmeridge area of Dorset. Although a very small fragment, part of a single ring and dot (cut) motif ornament (2.5mm diameter) can be seen on the outer circumference of this piece. Such ornamentation of similar size was found on the bone gaming counter from Iron Age Barleycroft (see Timberlake in Evans & Tabor 2013), as well as on a Late Iron Age Romano-British bone comb (object no. 366) from the Cambourne New Settlement, Cambridgeshire (Wright et al. 2009). However, it is difficult to reconstruct the appearance of this based on such a small surviving fragment. The probable diameter of the original bracelet was between 60 70mm. The object comes from a box-lined well on RB.2B. Possible Gaming Counters Four round partly shaped disc-like and natural rounded pebbles found associated with a cremation burial (total weight 63g). They were well stratified and had been deliberately placed in the cremation vessel in a tightly clustered group. The pebbles lay within the deposit of cremated bone, well below the dish that sealed this deposit. <236a> F.506 ([1326] sl.7) - A small round disc (dimensions: 26mm diameter and 2 3mm thick; weight 2g) made of a hard flinty chalk which has apparently been 276

175 ground down (and faceted) around the edges. This may have been fashioned to be used as a gaming counter. This was found associated with a cremation burial in C1 in RB.2B. <236b> F.506 ([1326] sl.7) - A still smaller round disc cut out and ground down from chalk (dimensions: 12mm diameter and 2 3mm thick; weight <1g). Found with <236a> (see above) <236c> F.506 ([1326] sl.7) - Two smooth oval-shaped flattened pebbles, which may (or may not) have been used as gaming counters as well. One of these is orange and a perfect flattened oval (42mm x 32mm x 7mm; weight 20g), whilst the other is a similarly flattened oval and is flesh coloured (mottled) with dimensions 50mm x 34mm x 11mm and weighing 40g. Both are pebbles of chalcedony which may have been collected from a beach (littoral) environment. Found with <236 a+b> (see above). The placing of these natural (little-worked or unworked) pebbles with this cremation burial, some or all of which may have been used as gaming counters or else as talismanic charms by the deceased, indicates an unusual, but not unique practice. Pliny refers to the game of ludus calculorum ( game of stones ) whilst another such Romano game would have been pentelitha ( five stones ; The Portable Antiquities Scheme ( records numerous finds of groups of pebbles thought to have been used as Romano-British gaming counters; one recent example being that of a round white quartz pebble from Lincolnshire (NCL-502AEO). However, records of these as finds with burials are rarer. Rotary and Saddle Quern Both rotary and saddle quern stone was recovered (Fig. 3.50). A total of kg of rotary quern stone was recovered from excavated Romano-British features on this site, the majority of which (54.86kg) was found as re-deposited stone within wells or watering holes (the largest amount of broken-up quern was recovered from well features F.2768 (RB 1) and F.1020 (RB.2B); respectfully 18.14kg and 20.34kg). All of the remaining quern was found redeposited within ditch fills, and to a lesser extent within backfilled quarry pits for gravel. In terms of the distribution of quern across the North West Cambridge landscape within archaeological features, the vast majority of this was concentrated in the area of RB.2B (51.17kg), most of the remainder of this being found within settlement RB.1 (28.08kg), with only a small amount (3.33kg) coming from RB.2A. Even when we add the relatively much smaller amount of rotary quern recovered from the 2011 evaluation phase, the picture doesn t change that much, the sum total for these different settlement areas being as follows; RB.1 (28.18kg), RB.2B (53.67kg) and RB.2A (4.64kg). Based on projected diameter (determined from the curvature of the circumference) these rotary hand mill querns may be divided up into the following categories: Small/ Beehive size ( mm diameter); Medium size ( mm); large size ( mm) and Millstones ( mm; Fig. 3.51). Except for the millstones (>600m diameter) which were massive and probably power-driven (King cited in Watts 2002; Appleby in Evans et al. 2013), this subdivision is to some extent arbitrary. Nevertheless, such a classification is still useful in helping to determine the scale of domestic milling activity taking place, perhaps also the number of households, the degree of mobility and potential sharing of mills (as suggested by the use of small or lightweight stones), or else the presence and frequency of millstones, which might indicate small-scale production rather than just self-sufficiency. Whilst it is possible that some of these querns were brought onto the site for use as rubble for the foundations of buildings, metalling roads, or to help revet or line the bottom of wells, the likelihood is that most were deposited as rubbish from the settlement following their use wear and breakage. More than likely many of these were first broken up by fire following their discard and secondary use within domestic hearths, and prior to their end use as the rubble lining of wells and other features. In terms of their overall site distribution, the following associations between the incidence of stone (lithological) types and size of querns has been noted (Table 3.64). 277

176 19th Century Quarries Archaeological features Niedermendig Millstone Grit Old Red Sandstone Puddingstone Other Sandstone metres Figure Distribution of Romano-British quern stones by weight, distinguishing four main stone types

177 Figure 3.51 Romano-British quern stone size (handmill/ millstone type). Small/beehive ( mm); Medium ( mm); Large ( mm) and Millstone ( mm) Figure 3.52 Romano-British quern stone provenance (quarry sites) Figure 3.53 Deposition by feature type for Romano-British rotary and saddle quern stone

178 Millstone Grit/ Carbonif RB.1 RB.2A RB.2B? unsized Small 1 1 Medium 2 1 Large RB Way 2 Old Red Sandstone? 1 Small 1 Medium 1 3 Large 3 5 Millstone size 2 Other sandstone (LGS?)? 1 Small Medium Large 2 Eocene conglomerate Small (Beehive) Niedermendig lava quern? Small Medium 1 Large 2 1 Table 3.64: Quern lithology and stone size (type) for each settlement area. Lithology and Provenance of Querns: Stone Typologies (Fig. 3.52) Old Red Sandstone (Devonian) 1.Quartz conglomerate Pale yellow-brown to pink quartz conglomerate (80 90% quartz) with some dark grey, green and red (hematite) lithic inclusions, little or no calcite or mica, and some feldspar. Possible provenance quarry site? near Ross-on-Wye, Hereford (see Shaffrey 2006). 2. Pebbly sandstone Pale brown sandstone with more rounded pebbles and 60% or less quartz. Uncertain provenance, but probably Wye Valley, Mendips or South Wales. 3. Gritty pebbly sandstone Pale brown gritty pebbly sandstone (Upper Old Red Sandstone?). Possibly Welsh Borders/ South Wales. 4. Coarse quartz grit with minor lithic inclusions. Unknown source. 5. Micaceous sandstone. Pale pink micaceous sandstone/ grit (Upper Old Red Sandstone?). Unknown source. 280

179 Millstone Grit (Upper Carboniferous) 6. Coarse pebbly quartz grit. Identity: Ashover Grit (or Chatsworth) from Melbourn or Duffield, Derbyshire, S.Pennines? (source info. K.Hayward in Evans forthcoming). 7. Coarse quartz grit with pink feldspar (arkosic). Identity: Ashover Grit (or Chatsworth) from Melbourn or Duffield, Derbyshire. 8. Micaceous quartz grit. Southern Pennines. 9. Medium grained micaceous sandstone. Southern Pennines? 10. Rough Rock coarse angular quartz grit-sandstone. Identity: Ashover Grit (or Chatsworth) from Melbourn or Duffield, Derbyshire, S.Pennines? (Hayward in Evans forthcoming) Coal Measures Sandstone (Upper Carboniferous)? 11. Micaceous sandstone Grey-yellow micaceous sandstone (similar to Pennant Stone). Unknown location. but possibly Southern Pennines. Other Sandstone (Lower Greensand?) 12. Ferruginous grit possible sources Eastern England or SE England? Conglomerate (Eocene). 13. Puddingstone grey to reddish round flint pebble clasts embedded within a silica-rich matrix and cement. Bradenham, Buckinghamshire, but most likely St. Albans or in particular Abbington Piggotts, Hertfordshire. Vesicular Lava (Tertiary) 14. Black vesicular basalt a dark grey-black coarsely vesicular hard basalt lava with frequent pyroxene (augite) phenocrysts. Niedermendig quarries, Andernach region of the Rhine. 15. Light grey vesicular basalt a slightly softer and more easily workable lava bed with small vesicles and inclusions of calcite and zeolites. Breaks up much more easily on weathering and burning. Niedermendig quarries, Andernach region of the Rhine. A total of 4.964kg of saddle quern (3 pieces) was recovered from the North West Cambridge excavations. This came from two Romano-British features on RB.2B, both of them wells, whilst another very small fragment (<2464>) was found within a quarry pit on RB.2A. In addition a fragment of discarded rotary quern (probably made of Lower Greensand ferruginous grit) from RB.1 seems to have been re-used as a saddle quern (this was also recovered from a well F.2768). As is typical of this artefact type, the querns appear to have been fashioned from small sarsen-type boulders of quartzitic sandstone or quartzite, most probably collected as erratics from the local terrace gravels. If one takes into account the saddle quern finds from the evaluation, the sum total of quern from the various different sites amounts to 11.82kg (RB.2B), 0.074kg (RB.2A), and 1.04kg (RB.1). This seems to show a clear bias in the distribution pattern towards RB.2B, and one should note that these forms of quern are typically Iron Age in character, though sometimes continuing into the 1st century AD on Romano-British sites; the latter reflecting the persistence of a conservative local tradition. During this period the on-the-spot creation of saddle quern appears to represent a very local utilitarian response to an immediate need. These are not items that travel, or are traded, in contrast to the manufactured rotary quern. <1436> - Small fragment of quern made from a split quartzitic sandstone (sarsen) boulder (probably Cretaceous-Eocene origin); dimensions 95mm x 85mm x 60mm (weight 0.54kg), found within a Romano-British well (F.564 [1444]) on RB.2B. This forms part of the rim of a shallow keeled quern with a flat and well-polished grinding surface (480 sq mm). The original quern may have been oval-shaped and around 250mm long. <1577> F.1402 ([4100]) - A burnt and cracked slab of quartzitic sandstone (sarsen), perhaps originally collected as a glacial erratic, which may have been used as a saddle quern (190mm x 170mm x 80mm (thick); weight 4.35kg). This was collected from the fill of a revetted Romano-British well (RB.2B). Whilst it is difficult to determine its original function, there appears to have been a subsequent use of this as building stone, this having been crudely dressed to a rectangular shape. 281

180 <2464> F.1917 ([5742]) - A small fragment from the edge of a burnt and broken-up flat-topped smooth sarsen-sandstone cobble, possibly part of a thin (25mm thick) saddle quern (dimensions of piece 45mm x 35mm x 24mm; weight 0.074kg) Some approximate, but useful comparisons can be made between the quernstone assemblage (including saddle quern) recovered from North West Cambridge and those from other Cambridgeshire rural Romano-British settlements. The figures within Table 3.65 show the approximate percentages (in terms of number of querns ) of the different lithologies (geological rock types or sources) represented at each settlement. North West Vicar s Langdale Cambridge Farm Hale Babraham Millstone Grit ORS LGS Puddingstone Niedermendig Sarsen (LGS) saddlequern Table 3.65: The lithological make-up (rock types) of rotary quern and saddlequern recovered from near-cambridge rural Romano-British settlements (here expressed as percentages of the site total). Not surprisingly this confirms that the proximity of these settlements to major road routes to the west, north and east seems to have been an influence upon the proportions of different quernstones supplied and also used. However, in the absence of any breakdown by settlement date, or by phase of use, such an analysis can only provide a very crude set of comparisons to work on. It seems more than likely that the road/trading network supplying Millstone Grit quern to Southern England came via Lindum and Durobrivae (Water Newton), which might explain the significantly higher % of these gritstone querns which are found at Langdale Hale. However, by the end of the 1st and early-2nd century this trade network is already supplying, and pretty much dominating the market for quern in and around Cambridge. Even so, the much higher proportion of Old Red Sandstone querns found at North West Cambridge compared to the other sites suggests that the periphery of another trading route from the southwest is reaching Cambridge via the Godmanchester Road (Corinium Ratae or Corinium Verulanium to Durovigutum (Godmanchester; Shaffrey 2006). Old Red Sandstone quern has also been found at other sites in Cambridge along the Huntingdon Road (e.g. at Orchard Court), and it could be that the easternmost limit of this supply system could be sites such as Vicar s Farm where we first begin to see this quern appearing in small amounts only at the beginning of the 3rd century AD (see Hayward in Lucas & Whittaker 2001). Babraham which lies close to the Romano-British road to Great Chesterford that crosses the Hertfordshire Chilterns (as well as the Icknield Way) not unexpectedly has a marginally higher amount of Hertfordshire puddingstone quern. This trade in puddingstone beehive quern continues from the Late Iron Age into Romano-British times, and certainly until the end of the 1st century AD (Peacock 1987). This Southern Cambridgeshire settlement s location near to the junction of the Via Devana from Colchester may also explain the much higher incidence here of imported Niedermendig lava quern originating from the Rhineland; this trade route crosses the North Sea landing at the port of Colchester (Camulodunum), thereafter supplying the Essex-Suffolk-Cambridgeshire hinterland, and still more importantly Londinium. 282

181 Cat no. Feature (context) Site Wt (kg) Size (mm)+ depth stone Upper/ lower stone Inner diam. Top/ Base (mm) Outer diam.(mm) [E= est.] Wear Notes Geology Origin 3335* 3467 (a) 3467 (b) 3467 (c) 3467 (d) 3467 (e) * 3467 (f) (a) 3399 (b) (a) 3360 (b) 2768 [9114] well 2768 [9316] well 2768 [9316] well 2768 [9316] well 2768 [9316] well 2768 [9316] well 2768 [9316] well 2768 [9112] well 2768 [9191] well 2768 [9190] well 2768 [9191] well 2768 [9191] 2768 [9187] well 2768 [9317] well 2513 [9174] ditch 2513 [9174] RB x 200 x L 28 / Lozenge hybrid (Type 2) MG coarse pebbly grit RB x 100 x U? 450 E 4 Flat top Type 1: with harp slight ferrug grit? RB x 95 x U? 450 E 4 part of 3467(a)?: re-used as saddleq? Derbys slight ferrug grit? RB x 70 x 40 U? 450 E 3 Flat top quern indet. MG Derbys RB x 120 x U 430 E Flat top Type 1c? ORS conglom Hereford (Ross)? RB x 170 x U 520 E 4 Flat top Type 3c RB x 200 x U 480 E 3 Flat top Type 1 or 3 RB x 50 x L 2 RB x75 x L 450 E 3 RB x 80 x U 500 E 4 Hybrid Type 4/6? (+wear grooves) Lower cake or hybid: Type 4/ 6 part of 3467 (d)? Flat top Type 1c: ORS pale yellow br qtz conglom ORS qtz conglom (banded) MG coarse grit MG coarse grit ORS conglom Hereford (Ross)? Hereford (Ross)? Derbys? Derbys? Hereford (Ross)? RB x 70 x 50 U? 250E 2 Flat top MG Rough rock Pennines RB 1 RB (x20 small frags) (x10 small frags)? 5 vesicular basalt? 5 vesicular basalt RB x 90 x U 3 RB x 95 x L 3 RB x 110 x U 9-12 E 3 Flat top Type 1a? +concent wear grv Lower cake: similar to 3400? same as 3360a poss the LS? MG micac qtz grit MG qtz grit or rough stone+ MG qtz grit or rough stone+ Niedermendig quarries at Andernach, Rhine Niedermendig quarries at Andernach, Rhine Pennines Pennines Pennines

182 3360 (c) 3360 (d) (a) 3408 (b) * [9174] ditch 2513 [9174] ditch 2514 [8046] ditch 2514 [9195] ditch 2514 [9195] ditch 2514 [8046] ditch 588 [1519] quarry RB x 90 x 37 U? 4 Flat top disc? ORS qtz conglom Hereford (Ross)? RB x 160 x U E 4 Beehive type RB x 110 x U 500 E 4 Flat top disc Type 2 Eocene puddingstnconglom Micaceous sstn U. Carbonif CM? Abbington Piggots, Herts.? Pennines or S Wales? RB x 65 x U? 2 Flat top with harp furrows MG coarse pink grit Derbys RB >20 small piece 5 NB burnt and weatherd vesicular basalt lava RB x 60 x 55? 420E 3 RB 2B 586 [1620] lined well RB 2B 586 [1574] RB lined well 2B 586 [1621] RB lined well 2B 1470 RB [4388] 2B quarry pit 564 [4194] well RB 2B 1402 [4100] RB revetted 2B well 1020 RB [3100] well 2B x 130 x U 95/ 45/ 28 Flat top with pecked pattern grind sfc (burnt) Beehive type Millstone Grit Eocene puddingstn conglom Niedermendig quarries at Andernach, Rhine Southern Pennines Abbington Piggots, Herts.? x 160 x U 40 E 400 E 5 Flat-top Type 1a ORS conglom Hereford (Ross)? x 50 x U 510E 3 Type 2 Flat top disc? ORS qtz conglom Hereford (Ross?) x 115 x U 510E 3 Type 2 Flat top disc? ORS qtz conglom Hereford (Ross?) x 130 x U 500 E (a) 110 x 75 x (b) 135 x 70 x re-fitting270 x 210 x x U + L U 120 E 400 E E 3 Disc Type 2a: with harp furrows Flat top: frags of L + U stones Flat top: Type 1a (possibly all same stone) MG (Rough Stone) vesicular basalt lava ORS qtz conglom Derbys Niedermendig quarries at Andernach, Rhine Hereford (Ross?) x 100 x L 80 E 3 Lower disc Type 6 MG mic qtz grit Pennines

183 * (a) 346 (b) [3695] well 1020 [3646.3] well 1020 [3646.2] well 1444 [4329] ditch 1109 [3200] well 1886 [5640] quarry pit 1117 [3217] ditch 1227 [3575] pit 1214 [3116] ditch RB 2B RB 2B RB 2B RB 2B RB 2B RB 2B RB 2B RB 2B RB 2B 906 [2599] pit RB 2B 952 [2725] RB ditch 2B 1236 RB [3721] 2B well 595 [5975] RB ditch 2B 595 [5975] RB ditch 2B 684 [1831] RB tree throw 2B x 180 x L kg 260 x 180 x x 250 x U U 30/ 40/ E? 260 E 400 E E 3 Lozenge Type 2b/ pecked pattern + worn grv Flat top disc Type 2b? NB poss millstone re-use as anvil Flat top disc Type 2b? millstone with poss rynd chas NB major re-use as anvil ORS pale brown pebbly sstn ORS? coarse quartz grit with minor lithic ORS? coarse qtz grit with minor lithic x 70 x U? 3 Type 5 Flat top disc? vesicular basalt lava Welsh Borders/ S.Wales? Niedermendg quarries at Andernach, Rhine x 70 x U 400 E 3 Flat top: Type 3? MG (Rough Stone) Derbys x 10 x U 520 E 4 Flat top Type 1a?: waterwn piece x U 550 E 3 Flat top Disc Type 2c? x 80 x U 520 E Flat top Type 1a: pick dressed top + scored grind surface x 95 x U 410 E 4 Flat top Type 3a? x 65 x U 500 E x 115 x U? 90 E Flat top Type 1b? Flat top disc Type 2b? reused as anvil? MG (Rough Stone) MG med gr micac sstn grit ORS pale brown gritty pebbly sstn (UORS?) ORS pale brown quartz conglom pale pink micac sstn w grit (UORS?)? micac sstn grit Derbys Pennines Welsh Borders/ S.Wales? Hereford (Ross)? Welsh Borders/ South Wales x 70 x U 3 Flat top Type 1? MG qtz grit sstn Pennines x 70 x L 500 E 4 Type 6 Lower cake MG felspathic qtz grit Pennines x 110 x L 550 E 2 Type 5/6 Lower Cake MG qtz grit sstn Pennines x 50 x L 290 E 4 Type 2 Lozenge? MG micac qtz grit Pennines

184 (a) 2985 (b) (b) 1619 (b) 1619 (c) 3196* [7505] well 2023 [6084] well 2023 [7462] well 2042 [7090] ditch 2056 [7295] box well 2056 [7295] well 2056 [7445] well 1236 [3599] quarry 1236 [3599] quarry 1236 [3599] quarry 2584 [8355] ditch 1979 [7065] ditch 1586 [5697] ditch 1786 [5331] quarry RB 2A RB 2A RB 2A RB 2A RB 2A RB 2A RB 2A RB 2A RB 2A RB 2A RB Way 2 RB 2A RB 2B RB 2B x 115 x U 550 E Flat top Type 1a MG micac qtz grit Pennines x 60 x U or L? 480 E 5 Flat top: weatherd rim frag vesicular basalt lava x 60 x 15 poss.frag burnt quern; no work sfc gritstone (MG?) Niedermendg quarries at Andernach, Rhine Pennines x 60 x L 500 E 3 Type 4? Hybrid style MG (Rough Stone) Derbys x 50 x 27 U or L 2 Flat topped MG coarse pink grit Derbys x 45 x U? 3 Flat topped peck pattern NB re-use as whetstone MG med gr sstn grit Derbys? x 90 x 40 U 300E 3 Flat top ORS conglom Ross (Hereford)? x 46 x U? 2 Flat top Type 1? NB re-use as whetston x 60 x 30 U 500E 3 Flat top Type 1b x 35 x 35 U 250E Beehive type (v small frag) MG med grit sstn ( Rough Rock )? MG med grit sstn ( Rough Rock )? Herts Puddingstn conglom x 115 x L 470 E 4 Type 5 Flat top disc? vesicular basalt lava 0.16 (x5 frags) 0.05 (>20 frags) quern may be 45mm + thick x 145 x L 460E 3 Table 3.66: Detailed catalogue and identification of rotary quern. 5 vesicular basalt lava 5 vesicular basalt lava Flat Top Type 2 Lozenge: with segment radial grooving Sandstone grit MG Rough Rock? Pennines Pennines Abbington Piggots, Herts. Niedermendgquarries at Andernach, Rhine Niedermendg quarries at Andernach, Rhine Niedermendg quarries at Andernach, Rhine S Pennines

185 19th Century Quarries Archaeological features Whetstone Anvil metres Figure Distribution of Romano-British anvil/mortar stones and whetstones

186 Anvils/Mortar Stones Six anvil or mortar stones (total weight c. 30kg) used for crushing, and in one case perhaps for forging iron (<1390>), were recovered from four different Romano-British features (mostly wells) located within RB.2 (Fig. 3.54). At least half of these mortars appear to have been made from re-used fragments of rotary hand quern or millstone, most of this re-use only having been identified during post-excavation analysis. Once again, the re-use of worn, broken and heat-fractured quern stone represents a very interesting aspect of the worked stone assemblage from this site. Such re-cycling may in fact be a lot more common than previously thought within the Romano-British settlements of the Cambridge hinterland. The much larger anvil stones are typically large cobbles or else boulders (up to 10kg in weight) collected from the local drift deposits such as the gravel terraces or gravel beds which lie close to the Traveller s Rest Pit. It seems that these have been used expediently as implements before being thrown into wells for use as metalled surfaces, or else simply as stone fill within the collapsing bottoms. Some of the mortar stones may well have been used for the crushing of seeds or foodstuffs, whilst the flatter anvils could have been associated with some craft function. The suggestion that <1390> may have been a smith s anvil has some support in the locally high incidence of iron smithing waste nearby, but even so, the find spot location is still very unlikely to reflect the proximity of the actual forge. <1389> - Re-used rotary quern stone fragment made from a coarse gritty Old Red Sandstone rock type found within F.1020 [3643.3] (Romano-British well), RB.2B; dimensions 260mm x 180mm x 110mm (> 5kg). This has at least two shallow mortar/ anvil depressions located on former grinding surface (c. 70mm diameter and 5mm deep). <1051> - Re-used rotary quern stone fragment made from a medium-grained micaceous sandstone grit (possibly from the Millstone Grit) found within F.952 [2725] (Romano-British ditch), RB.2B; dimensions 115mm x 115mm x 70 60mm (1.62kg). This has one possible flattish anvil surface on it (c. 70mm diameter). <1388> - Re-used large millstone fragment made from similar rock type as <1389> found in Romano-British well F.1020 [3646.2]. Dimensions 380mm x 250mm x 125mm (5 10kg). Has one large (or double) mortar hollow on old grinding surface; 170 x 80mm and c. 15mm deep. <3585> - Broken end of a heat-fractured split sandstone (sarsen) cobble found within a Romano-British well (F.996 [2851] SF 809) on RB.2B; dimensions 115mm x 70mm x 45mm (weight 0.58kg). This has an area of pecking on its flat surface, presumably where this has been used as some sort of anvil or hammerstone for crushing. <1577> F.1402 ([4100]) - A rounded sarsen boulder (a quartzitic sandstone erratic: 250mm x 200mm x 120mm (thick); weight c. 5kg) which appears to have been used expediently as some sort of anvil on its smooth upper surface. <1390> F.1020 ([3646.4]) - A large flat sarsen boulder found within the base of a well (RB.2B) which shows some possible use as an anvil stone on its upper surface (this bruised and also indented over an area of 130mm). In addition there is an iron pick mark on this surface which may be ancient. Dimensions: 280mm x 270mm x 90mm (thick); weight >5kg. The dark stain on the underside is manganese resulting from its burial in waterlogged sediments. Whetstones Some twelve whetstones or hone (polishing) stones weighing 2.57kg in total were recovered from eleven different Romano-British features (mostly wells but also including ditches, quarries and an inhumation burial) within three different settlement areas (Table 3.67; Fig. 3.54). The distribution of these doesn t appear to show any sort of significant association within the settlement contexts, except perhaps in re-deposition as stone rubbish fills (alongside quern and other worked building stone) within wells and quarry pits. The 290

187 exception to this perhaps is the whetstone made of non-local rock (schist) which it appears may have been placed with a human burial (F.848), perhaps as a personal item. Cat. No (b) Feature / F type/ context F.548 lined well [1381] F.1592 ditch [4727] F.1402 well [4100] F.614 well [1653] F.1236 quarry [3599] F.1236 quarry F.1547 quarry [4608] F.2056 box well [7295] F.1020 well [3102] F.1765 ditch [5277] F.848 inhumation burial [2378] Site RB.2B RB.2A RB.2B RB.2B RB.2A RB.2A RB.2B RB.2A RB.2B RB.2B RB.2B C RB.1 Table 3.67: Catalogue of whetstones. Description hone stone: v fine gr, polished + rectangular whetstone: fine gr, smooth + rectangular whetstone: fine gr, smooth + sub cylindrical whetstone: med gr, rectang pebbl, broken quern frag reused as coarse whetstone: square whetstone: rectang tablet whetstone: med gr + rectang/ cylindrical quern reused as coarse whetstone: square stone roof tile re-used as whetstone?:flat? whetstone: rectang pebble (broken) stone prepared for use as whetstone? whetstone/ polishing stone? Dimensions (mm) 130 x x (thick) 75 x 31 x 20 (thick) 80 x 25 x 18 (thick) 80 x 40 x 30 (thick) 43 x 46 x 27 (thick) 100 x x (thick) 80 x 23 x 17 (thick) 65 x 45 x (thick) 130 x 70 x 20 (thick) 68 x 32 x 23 (thick) 80 x 75 x 45 (thick) Wt (kg) x 85 x Geology slate/ volc tuf (Palaeoz UK) fine gr qtz sstn + slight micac (UK) fine gr crystal (hornblende) tuff (mid- France?) reddishgreen micac siltstn-sstn (Trias UK?) med gr Millstone Grit (U Carbonif, Pennines) fine-med gr micac flag sstn (Carbonif?) med gr micac yellow sstn (local LGS or Blackdown Greensand?) med-coarse gr grit (Millstone Grit; U. Carbonif) fine gr micac sandy limstn (Collyweston Slate) 0.15 quartz schist 0.56 quartz schist micaceous quartzitic siltstone Notes worked all edge + knife cuts worked on some edges worked (rounded) on all edges worked 2 narrowest edges worked hollow perhaps sharpening chisel? worked on faces + 2 ends: knives worked all edges/ faces worked around edges: knifes worked expediently on 1 face only possibly used on flat face frag broken off to use? surface of pebble used opportunist What is interesting about this particular assemblage is the mixture of whetstones made from non-local and imported whetstone-type rock types such as quartz-mica schist and various volcanic tuffs (both from Britain and perhaps also France (such as <1575> made of 291

188 hornblende tuff found within well F.1402) alongside those from classic Romano-British whetstone quarrying sites such as the Blackdown Greensand (in East Devon), those made from regionally local rocks such as Collyweston Slate (probably just re-used roof tile fragments collected from the North West Cambridge settlement) as well as re-used fragments of broken quernstone, in particular the finer Millstone Grit lithologies which make excellent sharpening stones for knives. This degree of re-use of quern for whetstone purposes has not been noted before within the Romano-British settlements of the Cambridge hinterland, yet it is not particularly surprising, and indeed might be more common than previously thought. To some extent this is a matter of recognition; some of this worked stone was not recognized as such when collected, just as the re-use of quern for whetstone was not initially detected. The similar-sized assemblage of whetstone recovered from the Romano-British settlement at Langdale Hale was composed mostly of local Lower Greensand (Appleby in Evans et al. 2013), whilst that from the nearby Camp Ground appears to have been more diverse, and of a higher quality. Most of the North West Cambridge whetstones appear to have been used for the sharpening of (probably small) iron knives (as suggested by the edge-worked areas and small knife cuts on some (e.g. <387>)), yet a worked hollow present on one of these (<1618>) might suggest that it was used for the sharpening of a small narrow chisel. Worked Antler and Bone Ian Riddler The small assemblage of objects and waste of antler and bone includes both late prehistoric and Romano-British material. The late prehistoric implements are limited to a large pointed blade and a scoop or scraper, object types that are reasonably common from the Bronze Age onwards, if not always in these specific forms. The large pointed blade is an object type more familiar from Bronze Age contexts than from those of a later date. The Romano-British material forms the majority of the assemblage and consists both of waste material (of antler and bone) and finished objects. A single piece of antler waste (1327) may be an unfinished handle, whilst a second object (2571) may have been produced as a handle but was subsequently adapted as a working surface, largely for the drilling of perforations with a knife point. Thus, although there are only a few objects and a small quantity of waste, some understanding can be obtained of the methods of a Romano-British workshop. Where the objects can be dated with any precision, they belong largely to the late Romano-British period. This is the case with all three of the bone pins. Two of the pins are common types whilst the third example (1568b) establishes a previously unrecognised, regional form of pin, found only in the Cambridge area and the outskirts of London. Metapodial tools, in contrast, are generally found in late prehistoric and early Romano-British contexts. The three examples belong to two common forms of the object type, which was used in textile manufacture. The remaining objects included three handles, two of antler and one of bone, the latter a comparatively rare form of object associated with large metal implements. An antler awl is probably also of local manufacture and could have been used as a punch, although it has a rounded point. Antler and Bone Working A small quantity of antler and bone waste, amounting to just two fragments, came from two separate contexts. One of the fragments is antler and the other is bone. A red deer antler tine 292

189 end (1327) has been separated from the remainder of the antler with the aid of a blade and a lateral notch has been crudely cut close to the terminal. The notch may possibly have been intended to decorate the tine end, much in the manner of grooved antler implements seen at Cadbury Castle, amongst other sites (Britnell 2000, fig ). It is more likely, however, that the intention was to cut away the end of the tine. The crude style of cutting is more functional than decorative and the tine appears to represent an unfinished handle. The single piece of worked bone waste consists of a fragment of cattle sized midshaft (530), cut from the side of the bone and facetted by knife or drawknife to produce a straight shaft of oval section at one end. In broad terms, it resembles bone pin waste but the shaft is much more substantial than those found on bone pins. Waste from the manufacture of bone pins is well illustrated by finds from Romano-British Cambridge (Alexander & Pullinger 1999, pl XXII). A larger fragment of worked cattle bone from Cambridge forms a good parallel for this piece (ibid, pl XXIII.183), both fragments of waste including an area of cortile tissue at one end, leading to a straight, facetted shaft. The size and shape of the waste is indicative of the manufacture of more substantial objects like bone keys, spoons or one piece handles. <1327> F.1020 ([2922] sl.482) - Complete antler tine end, separated from the remainder of the tine with the aid of a knife, with a lateral notch cut around the circumference close to the rounded terminal. No further modification. Length: 121.1mm Width: 23.0mm. Context dated to AD. <530> F.247 (F.548 in excavation phase; [716]) - Fragment of worked cattle-sized bone, possibly from the side of a radius, towards the distal end. The bone widens to a fractured, spatulate terminal with cortile tissue present on the inner face. Facetted by knife to a straight shaft of oval section, also fractured along its length. Length: 140mm Width: 30.8mm Context dated to 2nd 3rd century AD. Antler and Bone Objects Pins - Two of the bone pins belong to familiar and well established types, whilst the third is a rare form. A complete bone pin with a discoidal, nail like head set perpendicular to the shaft (1532) belongs to Greep Type B2.1 and Crummy Type 6 (Greep 1995, 1118; Crummy 1983, 24 25). One example is known from Romano-British Cambridge (Alexander & Pullinger 1999, pl IX.79). The type belongs to the late Romano-British period and there are few examples earlier than the 4th century. A complete pin with a pear shaped head and a swollen shaft (1568a) is a variant of Greep Type B1.4 and can also be placed in Crummy Type 3, another late Romano-British form (Greep 1995, 1117; Crummy 1983, 21 2). It is a common bone pin type, seen in some numbers at Romano-British Cambridge (Alexander & Pullinger 1999, pls. VIII IX). The third pin (1568b) is a rare form of late Romano-British date. It has a discoidal head set in line with the shaft and pierced by a perforation at the centre. It came from the same context as the late Romano-British bone pin (1568a) with a pear shaped head and should, therefore, be of a similar date. Its discoidal head is largely unparalleled in the Roman world, however, and is more commonly seen in post Roman contexts. A similar admixture of apparently Romano- British and Post Roman elements occurs with a bone pin from Earith in Cambridgeshire, which has a perforated discoidal head set above a swollen shaft. The head fits the Post Roman world whilst the swollen shaft belongs to the late Roman world (Riddler in Evans et al. 2013). A fragmentary third example of a bone pin with a perforated discoidal head came from Romano-British Cambridge, but is unstratified, unfortunately (Alexander & Pullinger 1999, pl IX.77). The pin has a notched head that allows it to be considered as a bone skeumorph of Cool s Group 9 Romano-British metal pins, which form a small group found largely in eastern England (Cool 1990, 160). The presence of a further example of a bone pin with a perforated discoidal head from a late Romano-British context at Shadwell in London (Gerrard 2011, fig. 90.4) confirms the dating of this group of pins. Their distribution echoes that of their metal equivalents and they appear to be a small group centred on the western part of East Anglia, the Shadwell pin forming an outlier to the south. <1532> F.1167 ([3402] sl.536) - Complete bone or antler pin with a shaft of circular section, widening below the discoidal head before tapering evenly to a rounded 293

190 point. Knife cut head with a flat apex. Highly polished throughout. Length: 92.9mm Width: 7.3mm. Context dated to AD. <1568a> F.1173 ([3540] sl.536) - Complete bone or antler pin, straight shaft of circular section, widening just below the head before tapering to a slightly facetted, blunt point. Head is pear shaped, with a flat apex. Length: 81.1mm Width: 9.1mm. Context dated to AD. <1568b> F.1173 ([3540] sl.536) - Complete bone pin with a lightly curved shaft of circular section, widening slightly from the head before tapering evenly to a bevelled point. Flat, discoidal head with central, knife cut perforation. Length: 110.2mm Width: 12.3mm Context dated to AD. Handles and Working Surfaces - Two handles have been produced from red deer antler tines. The smaller example <1738> has been cut from close to the end of the tine. The outer surface has been smoothed and the handle has been hollowed throughout. A slot at one end is intended for an iron blade, indicating that this is a handle for a knife. With a length of just 50mm, this is a small and delicate handle, seemingly at variance with the majority of Romano-British knives, which are substantial implements (Manning 1985, pls ). Small antler handles are known, however, from Late Iron Age contexts at Danebury and late Romano-British contexts at Ivy Chimneys (Sellwood 1984, fig ; Crummy 1999, 200, figs , , ). An example came from Romano-British Cambridge (Alexander & Pullinger 1999, pl XVIII.164). The larger handle <2571> was cut from a section of tine located close to the beam of the antler. It has been trimmed to a flattened oval section and smoothed on the outer surface; the centre has been hollowed throughout. It is a typical antler handle of Romano-British type, comparable with an example from Catterick (Wilson 2002, fig ). However, it may never have been used as a handle and at some point in the Romano-British period it was re used as a working surface. The two broad faces of the object are marked by numerous shallow indentations, produced either by the point of a metal awl or the tip of an iron knife. One face of the handle has fractured and only a few indentations remain, whilst the marks are spread across the centre of the complete side, where several lateral knife marks are also present. The adaptation of worked and unworked antler and bone to form working surfaces is a notable characteristic of the Late Iron Age and Romano-British periods in England. At Meare Village East an antler hammer head and a fragment of antler waste were used as surfaces into which numerous holes were drilled, as well as a trial of a single ring and dot motif (Coles 1987, 88, fig H5, 3.25H34). A cattle radius from Danebury includes a series of these motifs on one side of the bone and it was clearly used as a trial piece by a bone worker (Sellwood 1984, 395, fig ). A closer parallel is provided by a section of a horse innominate bone and a cattle sized radius midshaft retrieved from Romano-British contexts at Rectory Farm, West Deeping, which include clustered groups of puncture marks, also produced by an awl or a knife tip (James Rackham, pers.comm.). In this case the bones are not worked and they were chosen because of their sizeable flat surfaces. With all of these examples of indented bones the pitted marks are concentrated in clusters over a part of the available surface, which is generally flat and reasonably smooth. The indentations could stem from the use of awls in leatherworking, but might equally be the result of drilling holes with a knife point into objects of antler or bone. The indentations are seldom perfectly circular and some are rectangular in shape, suggesting that the latter alternative is the most likely. As such, they provide a further indication of antler and bone working practices on site. Medieval examples of working surfaces, interpreted as anvils used in metalworking, are known largely from the Continent and are also indented with sequences of rectangular or triangular indentations (Brios et al. 1995; Rodet Belarbi et al. 2002; Davis & Moreno García 2007, 85 88). <1738> F.1160 ([4066] sl.748) - Complete antler handle, formed from a section of tine, hollowed throughout. Outer surface smoothed and facetted by knife, and handle narrowed to a rectangular section close to the blade, with a vertical slot cut to receive the end of the blade. No iron staining visible. Length: 50.2mm Width: 16.5mm Context dated to AD 294

191 <2571> F.1806 ([5388] sl.1107) - Incomplete antler handle, produced from a section of an antler tine, lightly curved, with a flattened oval section. Hollowed throughout. Outer surface smoothed, both ends sawn laterally and lightly rounded. Subsequently used as a working surface, with pock marks on both broad faces (one of which has fractured in part) and also along the middle part of one of the narrow faces. Several lateral knife marks also visible on one broad face. Length: 96.8mm Width: 31.1mm. Bone Object - Both ends have been removed from a cattle femur <694> and two perforations have been neatly drilled close to the proximal end of the bone. It can be compared with several bones, including a horse femur and a cattle radius, from Dragonby in Lincolnshire (Taylor & May 1996, 357 9a, fig , ). These objects have also had their proximal ends removed, and up to seven holes perforate them, close to that end of the bone. The perforations are set either on one or both faces. It has been suggested that they were whittle tang handles for large metal tools, the tangs passing through the bones and secured at their ends (ibid, 359). This is plausible, although it is worth noting that these would have been substantial metal tools, probably held in both hands, for which the femur, as a heavy and well rounded bone, is eminently suitable as a handle. <694> F.739 ([2010] sl.239) - Fragment of a cattle femur, with both ends fractured away, unmodified except for the presence of two drilled perforations, 5.5mm in diameter, on the anterior face close to the proximal end of the bone. Length: 212mm Width: 58.3mm. Context dated to AD. Metapodial Tools - Three ovicaprid metapodial tools include two metacarpals and one metatarsal. The implements can be assigned to two different types within the series devised for Dragonby in Lincolnshire (Taylor & May 1996, ). A fragmentary example <919> lacks the distal end and has been perforated axially through the medial surface of the proximal end. It can be placed in Dragonby type 4 (Taylor & May 1996, 355). A complete example <2271> is little modified but there are two crimped indentations on the medial face, each of which is highly polished, allowing it to be placed in Dragonby Type 8 (ibid, 355). The third example <2572> is highly polished, with two light incisions on the anterior face and faint traces of transverse wear along both sides. It can also be placed in Dragonby type 8, on that basis. This tends to be the most common type of metapodial tool, with comparable examples from nearby at Cat s Water near Peterborough, Haddenham, Hinchingbrooke and Trumpington (Pryor 1984, 164, fig ; Evans & Hodder 2006, 210; Riddler forthcoming). The transverse wear visible on two examples stems from their use as bobbins. Wool or cordage would have been wound around them, causing the wear marks at either end of the bone and the high level of polish (Sellwood 1984, 392; Taylor & May 1996, 357; Coles 1987, ). It is not clear why some examples are perforated, either axially as here, or transversely through the centre of the bone. In this particular case <919> the distal end is missing, but the bone has suffered from modern damage. Axially perforated examples invariably retain their distal ends, so that a shaft of wood or some other material could only have been added to one end of the bone. <919> F.154 (F.1901 in excavation phase) - Fragmentary ovicaprid metacarpus, now lacking the distal end. Perforated axially through the proximal end; highly polished throughout. No further modification. Length: 96.5mm Width: 18.6mm. Context dated to AD. <2271> F.1765 ([5277] sl.169) - Complete ovicaprid metacarpus, lightly polished throughout with two crimped indentations on the medial face, each rounded and highly polished. Otherwise unmodified. Length: 110.6mm Width: 21.9mm. Context dated to 50BC 70AD. <2572> F.1697 ([5169] sl.169) - Complete unfused ovicaprid metatarsus, highly polished throughout, otherwise unmodified. Length: 128.5mm Width: 28.4mm. Context date to 50BC 70AD. 295

192 Antler Tine Implement - An antler tine end <183> has been facetted by knife to form a rounded point; it has fractured across the opposite end. It forms part of a large, awl-like implement, similar in some respects to Post Roman antler punches (Riddler 2003, 44, fig. 2). Large awls of antler or bone have also been found in Romano-British contexts, however, notably at Baldock, Bletsoe and Greenhithe (Stead & Rigby 1986, fig ; Dawson 1994, fig. 6.49; Detsicas 1966, 189, fig. 21.2; Riddler 2003, 42). <183> F.053 [129] - A near-complete antler tine implement, probably an awl, the end of the tine facetted by knife to a rounded point. Oval in section with the broad end fractured away. Length: 85.3mm Width: 22.2mm. Daub, Burnt and Worked Clay Simon Timberlake A total of 5.35kg of daub, burnt and worked clay was recovered from Romano-British features (including at least 1.18kg of fashioned clay brick or briquetage; Table 3.68; Fig ). A minimum of 17 different fired clay fabrics were recorded (including Prehistoric examples); the commonest being Types 3, 8, 4 and 9, with most being fairly well oxidised, grogged but with some organic inclusion, and in general typical of burnt and fragmented daub walling (in particular <2712> and <993> had impressions of large stick wattle). Some such as Type 3 were associated just with Romano-British features, and there are also suggestions of clay oven floors or bricks (<2494> <1868> <1911> <1961> and <1456>) as well as hearth bricks or clay pedestals (<1332>) which may be briquetage (i.e. Types 8, 15 & 16 ) found within well feature F.1020 Fabric types: 1) fine grained clay with few inclusions; oxidised exterior + grey reduced interior and porous 2) hard + well-fired light brown colour clay daub with small (<3mm) pink-white grog inclusions 3) pinkish-red to white or grey fine grain clay daub with some grog and flint inclusions and voids from burnt-out organic (chaff) as well as clay flow lines 4) sandy-silty red daub with small (<2mm) angular flint grit and grog inclusions, and a hard tile-like consistency 5) coarse grain silty red daub with numerous heterogeneous inclusions (1 7mm diam) incl flint, lithic, daub pellet and rounded chalk grains 6) pale cream-coloured soft silty daub, some with pinkish patches 7) yellow well-fired clay possibly a handmade brick 8) similar to Type 3 but coarser with large stone and straw inclusions 9) a coarse buff-pink coloured silty-clay daub with some mixing lines and largish inclusions of flint, chalk and grog 10) a conglomeratic gritty clay daub, well-fired with few inclusions 11) hard light brown-buff coloured clay with streaked mixing lines and few (grog and flint) inclusions 12) hard and homogenous buff-coloured clay with pinkish mottled streaks (no inclusions) 13) sandy brown grey daub (brown exterior) with angular burnt flint inclusions of diff sizes (2 10mm) 14) a mid to dark grey silty daub with v fine burnt-out organic, grit and rare charcoal inclusions 15) mid to dark grey brown medium coarse clay with burnt shell and grog inclusions 16) a hard and homogenous baked silty clay with no inclusions or voids 17) a roughly pink-grey streaky clay with occasional grit inclusions 296

193 19th Century Quarries Archaeological features Daub by weight (g) metres Figure Distribution of Romano-British daub by weight

194 19th Century Quarries Archaeological features Oven or Kiln material by weight (g) metres Figure Distribution of Romano-British clay oven or kiln floor fragments and possible kiln furniture by weight

195 A total of c. 1.8kg of clay oven or kiln floor fragments and possible kiln furniture (square poorly fired clay bricks and a cylindrical/cone-shaped support) was recovered from various re-deposited contexts within a number of ditches, quarry pits and a well (F.1020; Fig. 3.56). The 50mm diameter cylindrical support resembles to some degree briquetage debris, although another explanation would seems more likely here, given the distance of this settlement from the nearest Fenland salt-making sites. However, the use of this object within a pottery kiln is still conceivable, though the lack of any evidence for high temperature firing remains something of an issue here. The occurrence of square bricks (c. 100mm x 100mm; weighing c. 0.5kg) plus flat and right angle fired clay walling fragments might be interpreted here as being parts of the base of a kiln (either a pottery or corn-drying kiln) or bread oven, although the fabrication of these square to square/ half-round clay bricks might also be looked at as forms of kiln furniture. Various other flat burnt daub wall surface fragments within this assemblage seem much more convincingly to be parts of the walls of burnt-down buildings. This is certainly relevant in the case of the 226g of daub fragments collected from the fill of a beam-slot (i.e. F.900), although it should be said that other structural features such as postholes have provided little in the way of this material (< 30g). Cat no. Feature Context Site 2549 Wt (g) Description 1560 ditch 7019 RB.2A 24 burnt daub Fabric Nos. type pieces Type 3 + Type 4 2 Notes ditch 5948 RB.2A 44 external flat plastered clay surface 1586 ditch 5697 RB.2B 2 burnt daub Type 3 1 Type 3 1 part of oven, floor or brick? ditch 5072 RB.2B ditch 5073 RB.2B 4 burnt daub with fingerprint reduced fired daub Type 3 4 Type ditch 5245 RB.2A 12 daub Type lined well 7163 RB.2A 96 daub with large organic + fingerprint Type 3 pale col as little burnt ditch 5107 RB.2A 60 Type 3 1 pale col as little burnt ditch 5120 RB.2B 66 daub with fingerprint decoration? Type 3 4 pinkish slight more burnt 299

196 ditch 5157 RB.2B 34 burnt daub, some with uneven exterior Type ditch 5123 RB.2B 10 burnt daub surface 1733 ditch 5167 RB.2A 22 well-fired daub Type 4 1 Type ditch 5169 RB.2B 8 daub with fingerprint Type quarry 5217 RB.2B 6 burnt daub Type quarry 5075 RB.2B 18 oxidised pinkish daub exterior 2044 well 7112 RB.2A 18 burnt daub Type 3 6 Type 4 1 sandy and softer well 6085 RB.2A 534 part of daub walling with inter cavities and organic Type well 6084 RB.2A 638 part of daub walling with inter cavities and organic Type 8 + Type 6 13 impression of large stick (wattle?) c. 25mm ditch 5047 RB.2B 46 forms lip of clay structure 1670 quarry 4943 RB.2B 22 daub Type 3 2 Type 9 1 similar to <2494> ditch 4976 RB.2B 52 flaky daub surface 1654 quarry 4911 RB.2A 16 daub surface Type 3 12 Types similar to <2177> quarry 4922 RB.2A 18 part walling with stick impression 2073 ditch 7215 RB.2A 24 interior daub Type 10 1 Type ditch 4811 RB.2A 4 exterior daub 2040 ditch 7249 RB.2A 8 burnt daub Type 3 5 Types similar to <2087> 300

197 ditch 7323 RB.2B 410 large amorphous lumps + red ext Type quarry 5640 RB.2B quarry 7527 RB.2A 58 forms lip of flat surface clay flat ext surface daub 2087 ditch 7394 RB.2B 18 burnt daub 900 beam slot 2650 RB.2B 226 frags of surface Type 3 4 Type 3 3 similar to <2116> pale colour as less burnt Type 4 1 more sandy 35 Type (only 8 + one of most prob Type Type walling 13 13) material quarry 2599 RB.2B 144 some frags of uneven surface Type walling? ditch 2543 RB.2B 26 uneven external surface Type posthole 2068 RB.2B 40 abraded soft daub pieces Type lined well 1573 RB.2B 166 amorphous lumps burnt daub 962 ditch 2769 RB.2B 28 burnt daub 952 ditch 2725 RB.2B 6 burnt daub Type 14 5 Type 10 3 Type well 2922 RB.2B 1336 (a) half-round hearth brick or hearth lining (100x100x45mm thick; 540g) (b) clay pedestal or support (50mm diam; 186g) a) Type 15 b) Type WC objects include a hearth brick + clay stand (similar to pan support saltmaking); also poss bread oven or briquetage 301

198 (c) wall or oven lining (610 g) c) Type well 2922 RB.2B 64 part-vitrified hearth lining 1020 well 3099 RB.2B 78 burnt daub Type 9? 5 Type 9? NOT iron smithing, but poss traces of salt slag ditch 4169 RB.2B quarry 4231 water-abraded flakes detached from daub surface RB isolated feature 24 burnt daub 1487 quarry 4432 RB.2B 20 uneven surface Type 4 15 Type 9 + Type 6 8 Type ditch 4376 RB.2B 82 uneven lumps burnt daub Type quarry 4498 RB.2B 128 flat and 3x90 angle clay surface (70x50x22) 1167 later infill 3402 RB.2B 58 burnt daub RB.2B 10 burnt daub ditch 4173 RB.2B 32 burnt daub WC: could be part of Type 3 3 brick (sim <2116>) Type 4 + Type 9 2 walling fill? Type 4 1 Type ditch 4602 RB.2B 158 square/ rectang clay brick 70mm wide x35mm deep Type 3? 7 WC: clay tile brick (sim to <1868>) quarry 4724 RB.2B quarry 4660 RB.2B 54 flat surface + edges of bricks? sooted daub frags Type 3? 13 Type 9 13 as above or sim <2494> 302

199 ditch 3599 RB.2B 76 extern surface daub Type 6 + Type 14? 2 walling? posthole 2712 RB.2B 10 RB doubleditched 515 posthole 1234 system 10 abraded lumps burnt daub waterworn lumps daub quarry 3308 RB.2B 10 clay extern surface Table 3.68: Catalogue of daub, burnt and worked clay. Type 13? 7 Type 11? 4 Type 3 1 similar <1961>? Burnt Stone Simon Timberlake Altogether some 20kg of burnt stone was recovered from some 65 Romano- British features (Tables ; Fig. 3.57). Whilst some of this may represent residual Prehistoric material, it was generally found some distance from identified Prehistoric features and the low quantities of Prehistoric pottery and worked flint from the same areas suggests that much of this material was burnt during the Romano-British period. The largest quantities from Romano- British features came from a quarry F.1747 (3.29kg) and ditch F.952 (2.05kg), in neither instance is there an obvious Prehistoric source for this material. Some of these burnt and cracked cobbles may have been collected from the gravels and used to surround small fireplaces, whilst some may represent stones that were accidentally burnt. Cat. No. Feature F type/ date + site ditch RB.2D ditch RB.2B ditch RB.2A ditch RB.2B ditch RB.2B ditch RB.2B Context No pieces Total Wt (kg) Size (mm) Geology etc. lamin flag sstn + white sstn fine gr soft siltstn quartzite micac qtz sstn (sarsen)+ siltstone fossilif volc tuff periglac weathered Jur estuarine series sstn ditch RB.2B med gr sstn ditch fine gr lamin RB.2B sstn ditch RB.2A lamin sstn ditch lamin siltstn Notes build stn? 303

200 RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2A ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2A ditch RB.2A ditch RB.2B ditch RB.2A ditch RB.2A ditch RB.2B ditch RB 1 ditch RB 1 Ditch RB.2B ditch RB.2B lined well RB.2A lined well RB.2A lined well carstone carstone soft brown sstn pale yellow sstn grey-white sstn oval pebl Pal sstn coarse ferrug sstn white qtz sstn lamin soft fine sstn lt brwn micac sstn red-brown qtz sst grey brown sstn sstn grit carstone red sstn fissile sstn/sltstn pale sstn pebble flaggy sstn/ sltstn white sstn yellow sstn yellow + pink sstn Millstone Grit + basalt lava micac qtz sstn (similar to <1967>) carstone, sstn + qtz sstn brown qtz sstn Millstone Grit micac sstn + metaquartzite thin piece flaked from block thin flaked pieces ALL are quern frags >WS possible frag of saddlequern (1.618 kg) both pieces of rotary quern >WS 304

201 RB.2B (Bunter) pit RB.2A same- - same pit sstn RB.2B greywacke quarry fine gr+ micac RB.2B sstn quarry RB.2B spotted sstn quarry RB BF quarry RB.2B gritstone quarry RB.2B BF quarry RB.2B med gr sstn quarry 80 lamin sstn RB.2B 130 siltstn quarry white fine mic RB sstn Way quarry dolerite RB.2A (picrite) quarry RB.2B ferrug sstn quarry RB.2B carstone Quarry RB Way andesitic rhyolite (refitting pieces) quarry RB.2B micac qtz sstn quarry 50 var type qtz RB.2B 130 sstn Tree throw RB.2B micac qtz sstn (similar to <1967>) NB stone Well similar to RB.2D <1967> well weathered RB 2 Barnack build stn? well weathered RB.2A boulder sstn well pale fissile RB.2B sstn well oolitic RB.2B Barnack? Table 3.69: Catalogue of burnt stone recovered from excavated Romano-British features. 305

202 Cat. No. Feature F type/ date + site Well RB.2B Well RB.2B Well RB.2B Lined well RB.2B (see <409>) Quarry pit RB.2A Quarry pit RB 1 Well RB.2B Cremati on RB.2B C1 Ditch RB.2B Quarry RB.2B Ditch RB.2A Ditch RB.2B Ditch RB.2A Ditch RB.2B Ditch RB.2B Cremati on RB.2B C1 Context 2957 ENV <54> 2852 ENV <52> 2956 ENV <53> 1574 ENV<3> 4924 ENV <191> 8148 ENV <304> 3040 ENV <61> 1404 ENV <42> 2254 ENV <22> 4676 ENV <181> 4647 ENV <180> 2764 ENV <49> 1884 ENV <8> 3086 ENV <64> 3189 ENV <69> 1275 ENV <33> No piece s Total Wt (kg) Size (mm) Geology etc. fissile brown micac sstn frag hard sstn/ siltstn cobble round cobble of pale sstn pale sstn fissile micac fine gr sstn volc tuff/ chert? pale white hard sstn + pink micac sstn brown sstn sstn sstn granodiorite var sstn sstn calcin flint sstn sstn Table 3.70: Catalogue of burnt stone recovered from environmental samples from Romano- British features. 306

203 19th Century Quarries Archaeological features Burnt Stone by weight (g) metres Figure Distribution of Romano-British burnt stone by weight

204 Iron Slag Simon Timberlake A total of 76.51kg of iron smithing slag was recovered (Table 3.71; Fig. 3.58). This included at least 120 complete or fragmentary smithing hearth bases (SHBs) along with another c. 20 proto-shbs. Typically most of the larger SHBs were between mm in diameter (weighing between kg) and plano-convex to rounded-irregularly ellipsoid in shape (sometimes with a round to conical base), the smaller SHBs generally being more irregularly formed, and between 70 90mm in diameter. The relatively high magnetism of some of these suggests the presence of significant amounts of free iron (remelted hammerscale and forge fragments); this was also a lot more evident within the associated slag smithing lumps (SSL) and other accreted smithing material, some of which included visible fragments of partly-forged iron pieces. The variable nature and composition of these smithing hearth slags (fayalite-wüstite-iron) suggests that both forging and welding activities were probably taking place here; the latter associated with the liberal addition of fine quartz sand as a flux to remove the iron oxide scale and to produce cleaner and more effective welds (Bayley et al. 2001, 14). Yet other categories of debris found included fired clay hearth lining (HL), sometimes part-fused to the underside of the generally much denser SHBs, and also vitrified hearth lining (VHL), the latter being fragments of clay wall (and perhaps tuyere) which have become detached and fused (i.e. frothy and glassy) where these have fallen into the hottest parts of the fire. Meanwhile the presence of slightly hollowed-out fragments of vitrified clay and also blast-type impressions on some of the surfaces of the SHBs seems to confirm the use of clay tuyeres (air pipes) with internal diameters of around 40 50mm. Apart from the unusually high iron content of some of the smithing slags some characteristic (and perhaps diagnostic) aspects of this assemblage included evidence for the size of these round-bottom hearths (typically with bases mm diameter), and also the suggestion that these were shallow and dug straight into the ground (i.e. without much or any clay lining at all); the evidence for this being the accreted and part-fused sand or chalky marl to which many seem to be directly attached. The recovery of burnt Romano- British pot or tile from the bottom of these SHBs has helped to confirm the dating of this ironworking (smithing) activity. None of this material though appears to have been found in situ, all having been dug out of the ground and disposed of into ditches, pits, quarry pits, and perhaps most frequently of all within the bottoms of (abandoned) wells. The largest collection of slag from a single feature was recovered from well F.1236 on RB.2B (14.78kg), with another well on the same site (F.1020) producing a further 9.74kg, and the later infill of F.1167 another 13.58kg. However, some of the largest accumulations recovered from the Romano-British ditches amounted to just 1 2kg (the largest came from F.1442) although given that the percentage sampling of these would only have been around 5 10%, the amount per feature would probably have been many times higher. Much more significant was the distribution of slag between sites, and perhaps also between different settlement phases. More than 97% of the iron smithing slag came from RB.2B, but more specifically from the central-eastern half of the sub-rb.2c, from features dating from the mid 2nd to the mid-4th century AD, and most likely to the latter half of this period (mid-3rd to mid-4th 308

205 century AD). Just 3% of the slag came from the rest of the North West Cambridge site, most of this (2.2%) coming from a small group of features close to the northwestern edge of the adjacent settlement RB.2A, with only a tiny amount from one or two other features, such as those associated with RB Way 3. This unequivocal distribution pattern seems to suggest a focus for this ironworking activity within a 500 metre radius of feature F.1020 on settlement RB.2C. Given the high level of truncation across this area, it seems very unlikely that any evidence for shallow hearths, raised floor level(s), anvil benches, and traces of ephemeral workshop (i.e. shallow posthole) structures will have survived. Just as significant, though not at all surprising, is the complete absence of ironworking slag or other metalworking evidence from any of the prehistoric North West Cambridge features The ferrous concretions collected during the course of excavation (from both prehistoric and Romano-British features) seem all to be examples of a broken-up hard-compacted iron pan (goethite (FeO.OH)) deposit formed (pre-archaeologically) at the migrating water-table interface between the shallow sands and gravels and overlying subsoil. It appears likely that much of this was formed locally beneath the peats and gleyed peaty soils laid down across the lower waterlogged areas of this North West Cambridge landscape; since which time this horizon has been disturbed, broken-up and redistributed by agriculture. The mineral grade of this seems far too poor for use as a bog iron ore (i.e. significantly less than 50% Fe content), therefore it is unlikely this would ever have been sought or used during the Romano- British period whilst better quality Jurassic iron ores were still available to the northeast of the Fenland area within the vicinity of Durobrivae (Water Newton). More importantly, no evidence whatsoever of iron smelting has been found at North West Cambridge. The initial assessment of this large assemblage of iron-working slag (alongside that of the hammerscale recovered from environmental samples and grid sampling survey supports the conclusion all of it is Romano-British in date, and the product of secondary iron smithing work (i.e. the welding and forging of bar iron to fabricate and repair iron tools). The probable location of this industry lies close to the central-eastern half of RB.2C, although the truncation of the east side of this settlement area by Post- Medieval quarrying may have had the effect of appearing to shift this focus further to the west. Essentially here we are looking at a re-distribution assemblage of ironworking waste which appears to have been discarded into the fills of pits, quarry pits, and in particular into ditches and abandoned wells, but also into the overlying later infilling. Likewise the distribution of hammerscale reflects this primary or secondary dispersal of waste from the smithies into ditches and pits, but probably not the locations of the smithies themselves. However, the coincidental distribution of fine hammerscale and coarse smithing waste within the very same areas/ features implies that the former ironworking area lies close by (probably within a 500m radius of the densest concentration of slag), and that unlike the building stone it has not been brought onto the site from elsewhere for use as backfill or as rubble foundations. 309

206 19th Century Quarries Archaeological features Slag (Qty) Smelting Hearth Bases metres Figure Distribution of Romano-British iron slag by count and number of smithing hearth bases

207 Further analysis and comparison of this assemblage with that from other sites will be required, particularly in terms of assessing the importance of ironworking here at North West Cambridge during the later Romano-British period. It seems likely, however, that production at this site probably exceeds that of any other near-cambridge settlement, suggesting that this may well have been a net supplier of iron goods to the hinterland of this Romano- British town. Cat. No. Feature/ context/ site SF no. No. piece Wt (g) Magnet. (scale 0 >4) Iron smith slag * * * * Notes SHB (85x110x35mm) with solid flat top surface + free Fe slight irreg pl-convex SHB (120x110x45mm) SHB (130x105x40mm) +impress ch SHB (140x100x45mm) with 2 layers * large lump free Fe within SHB * proto SHB (35x55x20mm) * SSL with melted iron * broken up SSL with ch impres * irreg SHB(60x95x40mm) + ch * SSL (1 with free iron?) * edge frag of SHB * SSL with accret hearth base * proto SHB (35x30x20mm) * protoshb (45x35x5mm)wthrd * frag SHB with VHL attached * frag SHB with VHL attached * irregular SHB (65x85x45mm) * SSL * * md * * slag smith lump (SSL) with free iron? SHB (110x85x40mm) with two layers and cone base: weathered smithing hearth base (SHB) 80x60x30mm: weathered proto SHB (75x50x40mm) + free Fe * VHL weathrd coal cinder (Post Med) 271 F.525 [1254] * frags of x1shb(110mm diam) 317 F.561 [1440] half of pl-convex SHB * ditch RB.2B (crystalline) 80mm + diam 3534 F.579 RB.2B * proto SHB (20x25x10mm) F.586 [1573] concret x6 goethite concret + x1 free * well RB.2B 3 free iron (re-melted) Fe F.667 [1846] ditch RB.2B F.738 [2021] quarry RB.2B F.845 [2760] ditch RB.2B * * concret 2 - slag small frags of an SHB with Ferich layer a vesicular to frothy vitrified slag part of a chalk-rich clay hearth lining? * x1 SSL (50mm) 311

208 F.864 [2446] ditch RB.2B F.866 [2453] ditch RB.2B F.901 [2585] ditch RB.2B F.906 [2899] pit RB.2B F.921 [2639] ditch RB.2B F.925 [2654] ditch RB.2B F.928 [2661] pit RB.2B F.946 [2698] ditch RB.2B F.952 [2725] ditch RB.2B F.952 [2724] ditch RB.2B F.953 [2728] ditch RB.2B F.958 [2754] ditch RB.2B F.961 [2764] ditch RB.2B F.962 [2769] ditch RB.2B * SSL * SSL * concret * * * concret 2 slag 4 free Fe 0 concret 1 2 slag irreg SHB with blow hole (85x100x50mm) * x1 SSL + x3 concretion * * x4 frags of 1 SHB (90x80x45mm) + concret x1 irreg partial pl-convex SHB (140mm diam; 934g) SSL (some with melted iron) + proto SHB+ VHL+ concret x3 goethite concret + x1 slag droplet + x1 re-melted Fe x7 goethite sstn/gravel concretions + x1 SSL + 1 proto SHB(50x40x25mm) * x1 SSL * pl-convex SHB (100mm diam) weathrd * SSL conglomeratic with fl * (slag) 4Fe * frag of SHB with VHL attach (est 90mm dia)+ SSL+ Fe obj x1 goethite concret + x3 SSL + x2 free iron 1113 F.962 [2769] ditch RB.2B * x1 VHL (30mm) 1449 F.962 [3310] (5) + 0 * Fe objects / Fe melted in slag 1121 F.963 [2771] ditch RB.2B * SSL 1135 F.967 [2778] quarry RB.2B 4 400? 1 2 SHB 1 SSL 4 free Fe * x1 large corroded Fe object + x1 SSL + x2 SHBs (120x120x45mm + 80x70mm) 1143 F.972 [2795] ditch RB.2B * SSL 1453 F.972 [3312] * SSL w free melt Fe + concret 3568 F.997 ditch SHB (100x90x35mm ) with * RB.2B charcl and chalky clay lining 1171 F.998 [2861] ditch RB.2B Fe 1 slag 0 concret 1177 F.998 [2865] * F.1001 [2855] ditch RB.2B F.1003 [2870] ditch RB.2B * * x1 SHB (75x50x30mm) + x12 SSL (largest 100mm) + x2 pieces of (re-melted) free iron + x1 goethite (N) concret frags of prob x1 irreg SHB (80mm diam?) + SSL x1 v large pl-convex SHB (150x150x70mm) with 2 layers and blast h + 2 small SHB (80mm diam) + SSL 1198 F.1003 [2871] * SSL 1186 F.1003 [2869] * VHL 312

209 F.1010 ditch RB.2B F.1010 ditch RB.2B F.1010 [2963] ditch RB.2B * * Fe 2 3 slag * SSL 60x60x30mm with charcoal inclusion SHB (115x80x40mm) with charcoal incl x1 free iron (corroded) + x1 proto SHB (60x55x40mm) 1219 F.1010 [2963] * VHL in shape pl-convex SHB 1216 F.1012 [2899] * SSL 1225 F.1015 [2907] quarry RB.2B * SSL F.1017 [2912] ditch RB.2B F.1018 [2916] ditch RB.2B F.1020 well RB.2B F.1020 well RB.2B * * convex SHB (110x95x50mm) with unfired clay lining + irreg SHB(90x100x45mm) +SSL+F parts of x2 SHB (110x120x 40mm +120x120x45mm) + SSL without clay lining * proto SHB (70x60x25mm) * 1242 F.1020 [2922] * large SHB and hearth base with a conical underside + blowing depess interior (170x130x50mm) + piece of free iron in loose slag frags of 3 small SHBs (80x90 x22mm + 70x60x20mm + 60x50x40mm) + SSL 1241 F.1020 [2922] * frags fired clay lining + VHL 1248 F.1020 [2930] * SHB (75x60x23mm) 1346 F.1020 [3100] * x1pl-c SHB(120x95m)+SSL 1338 F.1020 [3099] F.1020 [2922] (1)+2+0 * frags pl-convx SHB (120mm; 776g) +x2 small ireg SHBs + pl-cnvx SHB(90mm)+SSL x3 lg pl-convx SHB(120mm) + 3 SHB(90m)+SSLwithHL 1199 F.1031 [2954] ditch RB.2B ? decalcify SSL? 1288 F.1057 [3031] frags of irreg-round SHB * ditch RB.2B (70x70x40mm) 1296 F.1059 [3045] quarry RB.2B * SSL or frags of SHB + ch x1 pl-cnvx SHB(130x100mm) F.1059 [3046] * frag SHB (90mm) 1310 F.1069 [3064] half of pl-convex SHB * ditch RB.2B (110mm+)with free Fe +SSL 1313 F.1070 [3068] ditch RB.2B to 4 * SSL with free melted iron 1359 F.1078 [3103] x1 pl-convex SHB (120mm) * pit RB.2B +irreg SHB + SSL 1357 F.1078 [3104] * small pl-convex SHB (80mm) 1430 F.1090 [3277] * x1 ellips SHB (150x100mm) 1378 F.1095 [3143] ditch RB.2B (x2) * SSL with ch 1395 F.1103 [3175] * x3frags pl-convx SHBs + SSL 1401 F.1108 [3189] x1 broken SHB (dense * ditch RB.2B crystalline) 1457 F.1109 [3308] * SSL w free melted Fe + Fe obj 1304 F.1122 [3249] quarry RB.2B * SSL 1414 F.1114 [3207] x2 irreg SHBs(90mm) +7SSL * pit RB.2B +melt end tuyere +TL in VHL 1420 x1 perfect pl-convex SHB F.1117 [3216] * (120mm; 1258g) +x7 irreg ditch RB.2B SHBs (90mm) with ch + SSL + 313

210 conc 1445 F.1136 [3306] pit RB.2B * SSL (one weathered) 1465 F.1138 [3319] small frag pl-convex SHB ditch RB.2B SSL 1514 F1152 [3406] ditch RB.2B * melted Fe? 1510 F.1152 [3420] frag of Fe bar- from smithing? 1733 F.1160 [4066] 1 small pl-cvx SHB (100 mm) * ditch RB.2B 2 frags SHB +4SSL+ 2VHL 1543 F.1167 [3404] x3 large round SHBs (130x * 100x50mm + 100x90x50mm) infill RB.2B and x3 small irreg SHBs (80 x4 large irreg-round and plconvex SHBs ( mm 1533 F.1167 [3402] * diam) + x5 small SHBs (65mm di) + SSL + VHL(70g) x4 large pl-convex to convex SHBs (c.120mm diam), some 1538 F.1167 [3403] * with clay lining + x4 small SHBs (85mm dia) + VHL(x1) 1526 F.1167 [3401] (5) * x3pl-cvx SHBs (c.110mm) + x5 half SHBs + x8 small frags SHB + 11 SSL + x2 melted Fe 3569 F.1168 well SSL + vitrified hearth lining * RB.2B (VHL) 1551 F.1168 [3544] * SSL+VHL(square edge hearth 1500 F.1189 [3465] * VHL with incl of melted Fe 1260 F.1214 [3116] small pl-convex SHB * (90x90x25mm) and parts of 3 ditch RB.2B other small SHB + iron 1361 F.1214 [3115] * SSL F.1236 [3721] well RB.2B F.1236 [3712] well F.1236 [3599] well F.1236 [3848] well F.1278 [3720] ditch RB.2B F.1292 [3756] quarry RB.2B F.1391 [4064] ditch RB.2B F.1442 [4310] ditch RB.2B * * (3=25%) (0=75%) * * * x3 pl-convex SHBs (largest 140x120mm)+ x4 smaller SHBs+SSL+VHL lump +Fe x1broken SHB (orig 120mm diam) + x3 small conical-irreg SHB + proto SHB +SSL + Fe x1 SHB (130mm) + x6 small SHB(65mm) +56 SSL+4VHL (incl FL+Rom pot)+17 Fe obj x5 SHBs (120mm; 700g) a (90 100mm) +frag SHB+SSL x1 proto SHB +x6 large/ small SSL+ HL+VHL(x3) + Fe * x2 SSL + x1 concretion x1 pl-cvx SHB (100x80mm) * large frag SHB + 3SSL 2 parts of SHB * (100x100x45mm) with ch F.1442 [4376] 4 SHBs (largest concav-round * 90x90x35mm; smallest ditch RB.2B 80x60x40mm) + 3 proto-shb frag of SHB (Fe-rich) with 1833 F.1442 [4380] * chalk base + SSL 1829 F.1442 [4377] * SSL part decalcif (light) 1806 F.1442 [4308] * pl-cvx double SHB (120mm) 1815 F.1444 [4329] 65mm sq frag of SHB (25mm * ditch RB.2B thick) with calcined fl 1846 F.1470 [4388] SHB (120x105x40mm) with * quarry RB.2B charcoal 1900 F.1543 [4593] globular/irreg to cone shape * SHB (110x90x70mm) with gas ditch RB.2B ves + ch 314

211 2566 F.1601 [7014] ditch RB.2A * SSL from top of SHB 2182 F.1729 [5150] ditch RB.2B * frag SHB with free iron 2192 F.1731 [5157] refit pieces of proto SHB * ditch RB.2B (50x40x10mm) 2267 F.1765 [5257] * small frag of SHB 2373 F.1821 [5487] 1 conical proto-shb * well RBWay3 (40x35x40mm) + burnt stone 2308 F.1866 [5642] * SSL 2510 F.1999 [5970] quarry RB.2A * SSL 3577 F.2023 well SHB (120x95x35mm) with * RB.2A small charcoal incl 2847 F.2044 [7505] plano-convex SHB * well RB.2A (130x120x45mm) with blow h 2974 F.2056 [7281] SSL (largest 80mm diam) with * well RB.2A gas voids 2939 F.2081 [7247] quarry RB.2A corroded iron frag 3017 F.2083 [7323] globular shape SHB * ditch RB.2B (110x90x60mm) 2960 F.2086 [7271] * SSL 3096 F.2134 [7537] quarry RBWay * proto SHB (55x45x15mm) Table 3.71: Catalogue of slag and ironworking debris, not including hammerscale. Romano-British Hammerscale Simon Timberlake Some 2043 pieces of hammerscale from iron smithing were recovered from 100 small soil samples (approx. 250 litres soil) taken from a 50m long subsoil transect (Transect E) sampled beneath the now-removed concrete farm road which formerly crossed the University Farm in a northeast to southwest direction (Fig. 3.59). This transect was sampled as it provided the only surviving subsoil strip not removed by machining during the initial stripping process. The purpose of this sampling was to try and help determine if there was any evidence for in situ iron smithing within the near vicinity. This distribution of hammerscale (the hammerscale/soil concentration) has now been plotted as a contour gradient, and as such this may useful for comparing with the distribution plot for iron smithing slag, the plan of archaeological features (such as ditches) which either underlie or abut this road, and the current model regarding the focus and whereabouts of the Romano-British ironworking activities. A 50m-long strip of subsoil beneath the removed concrete road consisting of two parallel transects (northwest and southeast) several metres apart was sampled at 1m intervals (total 100 samples) by removing c. 2.5 litres of soil sample from the centre of each 1m2. These were bagged and returned to the CAU where they were processed as environmental samples, wetsieving and collecting the non-organic (mineral) residues within the >4mm (4mm-10mm), 2-4mm and <2mm size fractions (JH). The residues were then air dried and the magnetic (or magnetised) material within each of them collected and removed using two powerful handheld magnets. These samples were then re-bagged and forwarded to Simon Timberlake for examination under a x10 illuminated magnifying lens. Individual pieces of platy hammerscale and spheroidal hammerscale were separated and counted in a glass petri dish 315

212 using a camel hair brush. The remaining composition of these magnetised residues (i.e. the ferruginous concretions, goethite/ magnetite grains, micro-fragments of iron slag, fragments of iron objects, burnt pot or clay, burnt stone, burnt flint, and magnetised charcoal or burnt soil) was then visually assessed and recorded. The number of pieces of spheroidal and platy iron hammerscale collected from each sample (residue) are recorded in an excel spreadsheet within the archives; Samples representing the southeastern transect (0 50m) and Samples and (0 50m) representing the northwwestern one. It can be seen from this that there are at least two peaks in hammerscale abundance corresponding to one larger anomaly at A (Samples m, 30 33m, northwest transect, c. 500 pieces and Samples , 33 36m, southeast transect, 126 pieces) and, both probably representing the same dispersal, smaller ones at B (Samples , 9 12m northwest, 73 pieces) and C (Samples , 42 45m northwest transect, 155 pieces). In broad terms, these hammerscale concentrations correspond well with two positive anomalies (of between (SI x 10-5) recorded in the topsoil magnetic susceptibility survey within Area 1 that crosses the route of the road, confirming to some degree that it is probably the hammerscale distribution which is significantly influencing the pattern of this anomaly map. Interestingly, the moderately high readings at this point probably just reflects the volume of soil with generally dispersed hammerscale and other residual magnetised material present within the fills of a number of ditches. In other words, this pattern probably just confirms the secondary distribution of iron smithing waste, either as intentional or unintentional deposition, throughout the fills of the deeper buried features within this moderately truncated Romano-British settlement landscape. Nevertheless, a much higher magnetic susceptibility anomaly (of up to 65 x 10-5) picked up some 25 30m to the northwest of (A) might indicate a much more localised concentration of hammerscale, suggesting that the original source of all this (i.e. a since truncated smithing floor) could lie somewhere in this direction (note also the slag (smithing hearth base (SHB) finds found within this same area and towards the eastern edge of this RB.2C settlement). Comparison of the counts of platy hammerscale (resulting from forging activity) against spheroidal hammerscale (resulting from forging and sometimes welding activity using a flux (Bayley et al. 2001)) shows little evidence of spatial separation across this area, confirming once again the likely secondary deposition of this material, yet at the same time serving to demonstrate that the activity this represents is on the whole quite typical of smithing, in particular the day to day repair and re-forging of tools, with some, but no main emphasis on the manufacture of large welded iron objects. Tthe magnetisation of the remnant mineral/ organic material present within the sample residues is more likely just a function of the accidental firing of naturally occurring iron-rich ferruginous concretion and iron-impregnated soil, charcoal and stone and flint present within and around smithing/domestic hearths and/or kilns. This material would have been distributed in the self-same way throughout the fills of these ditches as a result of rubbish backfill, natural slumping in of banks and the washing-in of fine sediment, and perhaps also vertically, with particles migrating through the topsoil and subsoil layers as a result of ploughing and earthworm activity. 316

213 0 metres th Century Quarries Archaeological features Plate Spheroidal metres Figure Distribution of Romano-British hammerscale from sample transects

214 In conclusion, there are a number of distinct (both rising and falling) concentrations of hammerscale present within the subsoil sampled along this 50m transect underlying the concrete road. These anomalies correspond quite well with existing magnetic susceptibility plots as well as with a number of large buried features such as northwest-southeast Romano-British ditches possessing deep fills, some of which also seem to be associated with finds of iron smithing slag such as SHBs. The distribution of hammerscale does not appear to represent a primary distribution on the floors of or around the sites of smithies (i.e. as dumped material), but rather as secondary deposition of fine material washed along the ditches from the original sites of rubbish backfill or the slow infill with material left lying around the surface. The findings support the idea that the original smithy floors probably lay some 50m to the north and east of here. Moreover, the ubiquitous presence of hammerscale within this soil suggests that we are unlikely to be looking at slag brought in from elsewhere for use as rubble fill, but instead that it represents the product of light industrial activity taking place on the settlement itself, albeit in rather large amounts. Wood and Timber Richard Darrah The wood and timber recovered consisted of the lower portions of the in situ linings and revetments of five wells, plus occasional items from the fills of several wells and ditches (Fig. 3.60). Material survived where features had been cut sufficiently deeply into the underlying gault clay that fluctuating water tables have not affected the wood. With the exception of the in situ linings there was generally very little wood or timber present, indicating that little of this material was dumped in features presumably because the bulk of unwanted material was utilised as firewood. Material has been recorded, drawn where appropriate and sampled for species identification and dendrochronology. Species identification has not yet taken place; as a result all species identifications in this report other than oak are provisional. Wells It should be noted that the majority of well linings had decayed entirely, so that those where wood survives are a minority. Based on overall feature morphology principally the fact that wattle lined wells typically have a distinct deeper circular shaft whilst box-lined wells require larger construction holes to facilitate working within the hole it can be suggested that in total there were nine wattle-lined (one definite and eight probable) and box-lined (two definite and probable) well shafts. There is no evidence for the use of cask-lined wells, although these might not be readily distinguishable from wattle-lined wells. Although some stone was found in the base of three wells this does not appear to be structural or even to form a base for the well structure, instead the stone was simply placed during construction to provide a firmer and more reliable footing. The well linings would have all been over 1.0m deep and including likely superstructures some were probably in the region of 2.5m. Only the lowest c m of the well linings were preserved. They varied in their construction from carefully planned robust structures of oak to simple wattle lined structures. Two of these consisted of a combination of vertical stakes and horizontal boards on edge of between m square whilst the wattle-lined well was c. 0.9m in diameter. A fourth example appears not to be the well shaft proper, but some temporary revetments linked to the construction process. Finds of wood from well fills were minimal, suggesting that floating objects had been removed. Several of the wells had had pieces of poor quality round wood in the fills, it is unclear if this material was deliberately dumped in them or if it represents the upper portions 318

215 of structures that inadvertently collapsed in during dismantling or decay when the wells were abandoned. RB Well 03 (F.564) - Five tangentially sawn oak boards with surviving lengths of up to 0.88m formed the sides of this well constructed in the mid-2nd century. At least one of these boards had been reused as it had a redundant 8mm diameter peg hole near its end, this peg hole did not coincide with the impression of the well upright on this timber. Two uprights survived, one made from a split section of a sawn oak board, the other carefully shaped into an octagonal cross sectioned point from an 1/8th radially split Ash trunk. The shaping of this piece was the best example of skilled axe work from the site. These boards were up to 0.25m wide, as the well had a maximum surviving depth of 1.9m and was probably c. 2.3m deep this implies approximately nine courses of boards to reach the surface, with perhaps twelve courses if there was an additional low superstructure. RB Well 11 (F.1020, 1148) - The shaft of the wattle-lined well (F.1148) which was probably constructed in the 4th century was heavily deformed, as excavated it was an oval m in diameter, indicating that originally there was probably a c. 0.9m diameter shaft. The wattle lining was built of fifteen sails (uprights) of 40 50mm diameter round wood and with horizontal rods less than 25mm diameter. Half of the sails were Oak, these were felled as the spring vessels were either forming or just formed. This suggests that the wood was collected and the well built in late spring or early summer. The diameter of the Oak sails varied between 40 and 50mm, their ages varied between five and eleven years suggesting that this wood represents draw felling specifically for the construction of the well rather than cutting on a coppice cycle. Although the most of these pieces were ideal clean straight stems two sails plus one piece from the fill were overgrown knotty wood suggesting a shortage of supply of the ideal material. Several short split pieces of timber and one sawn piece of oak had been used to revet the sides while the wattle structure was being built, none of these survived with any clear evidence of their previous use. It is likely that the wattle structure was woven in situ rather than being prefabricated and lowered into position. The main evidence for this is that the points of the sails had been driven a considerable distance into the undisturbed natural Gault clay under the structure, typically c m. It is improbable that these could have been driven in more than c. 0.2m if the wattle structure had already been constructed. It should be noted that the complete wattle lining would probably have weighed kg, so that constructing it in situ may have been simpler than manoeuvring a prefabricated structure into position. The surviving shaft was 0.7m deep; to reach the contemporary ground surface it would probably have had to be a minimum of c. 1.2m deep and may have had some form of superstructure. There was also some wood associated with an earlier phase of the well (F.1020). This 0.32m long and 0.10m diameter piece ([8695]) that has both ends cut by an axe in a similar fashion is probably a firewood log rather than an off-cut. Whilst it could represent an accidental inclusion it may have been deliberately placed to provide a firm footing during construction. RB Well 12 (F.1402): The lining of this well shaft does not survive, although the form of the feature indicates that it was probably a wattle-lined structure. The failure of this structure to survive when other wood does suggests that it was constructed from species more prone to decay. The surviving material is in situ but represents two phases of temporary revetment that were utilised during the initial construction of the well. These probably date to the 3rd century. The first phase of revetment consisted of two horizontal boards on edge ([4246] and [4249]) held in position by four roundwood uprights ([4247], [4248], [4250] and [4251]). These were set around the southeastern portion of the well construction pit and spanned around 1/8th of the circumference. Based upon observations during excavation this revetment was probably required as a stopgap measure to prevent collapse/erosion caused by particularly high flow of groundwater through a sandy 319

216 seam in the natural. One of the uprights ([4250]) included a reused tangentially sawn board with a shallow lap (8mm deep and 45mm wide) on its edge that had been split up to create a 40mm wide stake. The on edge boards were a radially split oak board ([4246]) and a reused part of the seat of an oak stool ([4249]) 0.5m wide, discussed further below. The second phase of revetment which is stratigraphically later but probably effectively contemporary consisted of more horizontal boards on edge ([4097] and [4099]) and upright half round stakes ([4219] [4222]) with large pieces wood/timber and stone subsequently laid horizontally in front of them. These were set on the western side of the well construction pit and spanned around a quarter of the circumference. This revetment was probably required as a stopgap measure to prevent collapse/erosion of the relatively poorly consolidated fills of an earlier feature to the west which formed the edge of the pit. The boards on edge are outer tangentially split boards, one of which had been cut to length with an axe with a 71mm wide blade ([4099]) had been cut to length with an axe with a 71mm wide blade. The horizontally laid pieces included a horizontal halved reused ash trunk ([4098]) 1.62 m long by 0.17m diameter that is discussed further below. There were also four pieces of tree trunk ([4216] and [4217]), although it is possible that some may come from a single tree. These were 0.25m, 0.22m, 0.15m and 0.09m diameter and m long. Several showed sign of insect and woodworm attack, one had overgrown knots and another was sinuous. None had been worked apart from having axe marks at the end and they all appear to be pieces that would have been of little other use apart from as firewood. No nails were found in any of the well timbers and overall these temporary revetments are a much more haphazard construction than the other well linings. This is presumably because they do not represent the well lining proper but just temporary construction related expedients. RB Well 15 (F.2056) - This square well structure consisted of on edge horizontal oak boards with carefully shaped cleft oak uprights at the corners and connecting half round cleft Ash poles and was constructed in the 3rd century. The first stage in the construction was that the four carefully shaped cleft oak uprights were driven into position. These were then connected by 50mm diameter Ash poles nailed horizontally across their inner faces with 90mm long nails. The Ash poles were nailed less than 80mm from their ends and had been predrilled with two 5mm diameter holes side by side to stop it being split by the 6x6 to 6x8mm square nails. The position of the Ash braces at m above the base of the points suggests that this reinforcing was near the bottom of the well. These Ash poles held the structure square and the radially cleft unworked Oak boards were then placed on edge round the outside of the Oak corner posts. There was no evidence on the complete oak boards that they had been nailed in position. All the boards and uprights from this well had been made from a single oak tree. Several of the boards had the remains of the felling cut on their ends, as the highest quality section of the trunk is directly above the felling cut this means that the best pieces of this tree were being used for the well lining indicating that the oak tree was felled and split up specifically for the construction of this well. The fast growth rate of up to 6mm per year may occur in trees growing widely spaced in woodland (based upon personal observation in Bradfield Woods). A distinct feature of this tree was longitudinally short medullary rays. These short rays suggest that this straight trunked tree had been stressed, which would fit with it being a hedgerow tree with a large crown. It is possible that the trunk was not much longer than the length of the longest 1.24m board; although the grain at the top ends shows no evidence that the boards had been split into the crown. This suggests that the trunk may have been double this length at c. 2.5m. The steps in the split surface suggest a difficult tree to split because of the short rays. This length of the trunk had been split into approximately 24 pieces that were m wide. This would have provided a lining, including four pieces used for uprights, of m deep. If 320

217 there was a second upper portion of the trunk this would have provided a similar amount of planking, giving another potential 1.2m of lining. The maximum surviving depth of this well was 1.7m; the original shaft must have been c. 2.0m to reach the contemporary ground surface and it should be remembered that the wells should ideally be boarded above ground level to stop surface water and waste entering the well. It therefore seems likely that the entirety of the 2.5m useable trunk of this approximately 60 year old tree was used to construct the lining for the shaft of this well. Although no complete axe marks survived from the felling cut, the axe used was narrow bladed with a width of <75mm. This corresponds well with the complete axe blade width from another axe of 71mm from a timber in RB Well 12 (F.1402), suggesting that the felling axe used by farmers was comparatively narrow. RB Well 16 (F.2044) - The only surviving in situ timbers in this well, which was constructed in the 3rd century, were two roundwood stakes. Given their location these probably relate to temporary constructional revetments similar to RB Well 12 (F.1402). Many Romano-British wells from East Anglia were made from carefully sawn planks, including examples at Scole (Ashwin & Tester forthcoming; Darrah 1996) and Great Holts Farm (Darrah 2004). It appears that at Scole in particular the wells in the centre of the settlement were built by specialised carpenters to a relatively standardised style (Rogerson 1977), with square corner posts, elegantly curved horizontal braces backed with horizontal planks on edge. In some cases these well linings were constructed from parts of a single tree, and so may have been sawn especially for the construction of the well. In contrast some outlying wells at Scole were made from cleft timber or reused materials, and did not conform to the style found nearer the centre of the settlement. These are more likely to have been constructed by the sites inhabitants with whatever materials were available to them at a particular point in time. The wells from North West Cambridge essentially fit this latter model; of wells built as one-off structures by a site s inhabitants with whatever material was available as opposed to those built by specialised craftsmen who followed a standardised pattern over a long period of time. Wooden Items Part of a chair seat and a structural timber from RB Well 12 (F.1402) have already been mentioned, other notable items include a lathe-turned furniture finial, a fragment of a coopered stave and two boards, possibly from a costrel (a container for liquids with loops through which a belt or cord may be passed for easy carrying). The lathe-turned furniture finial was recovered during the evaluation ([714] F.247 renumbered as F.548 (RB Well 05) in the excavation phase; Evans & Newman 2010, 56 57) and was deposited in the 2nd 3rd centuries. This piece had been turned on a lathe out of a 1/8th radial section and visual examination suggests that it may have been made from Holm Oak, a large evergreen oak native to the Mediterranean region first introduced to the United Kingdom in the 16th century. If this is correct then this would indicate that this is an import from the Mediterranean. One end had been turned down to 10mm diameter cylinder that had broken across. This rod was probably the means of attachment to the furniture. The other end had two non-central nail holes; these were definitely not the metal centre of the lathe on which the turned piece rotated. Around this end was an area of turning or carving that either represented turning with multiple centres of freehand carving, the end with the nail holes was square. This square end and nail holes suggests that this piece may have been finished with a metal cap. 321

218 19th Century Quarries Archaeological features NWC12 <003> [3321] F Wood Type Well lining Well revetment Wooden object metres NWC09 <108> WD25 F centimetres Figure Distribution of Romano-british features that contained wood plus illustrations of selected items

219 A fragment of a coopered stave was recovered alongside a few woodchips in the fill of a ditch (F.1138 [3321] <003>). This comes from a high quality small diameter mm coopered vessel. The fragment survives a broken and burnt piece. The piece of stave is backed (rounded on the outside) and hollowed (curved on the inside to match the curve on the outside) over its incomplete length of 78mm tapering from 25mm on its outer face to 21.5mm on its inner face. It was clearly coopered, as the sides show the compression where the adjacent staves have pushed against them and was probably from a straight sided vessel. It has a delicately rounded lip, suggesting that it could be a drinking vessel rather than a tub or bucket. Most Romano-British coopered vessels were oak or silver fir, this fragment is neither and was clearly the work of a skilled craftsman unlike the two pieces from a vessel in RB Well 12 (F.1168) that were much cruder. The burning was on the outside of the stave and was not the typical charring found on some modern coopered vessels. The broken and burnt state of the stave indicates that the vessel was already damaged before deposition. The fill of RB Well 12 (F.1168) contained two approximately circular boards, both 130mm diameter and 21mm and 23mm thick respectively. These were probably deposited in the 4th century. One was made from Oak and was not quite circular with an uneven double chamfer round its edge. This may have been the base of a coopered vessel or hollowed trunk vessel. The similar size of the second piece, which was made from a different species, with an offcentre irregular hole may have been the other end of the same vessel. If the two pieces do come from a single vessel then they represent the two ends of a costrel. The temporary constructional revetments of RB Well 12 (F.1402), which probably date to the 3rd century, contained two interesting reused pieces There was a broken oak plank that is part of a hollowed stool seat (<038> [4249]). This radially faced Oak seat was a very different quality from the rest of the Oak found on the site. The rays were very long up the length of the trunk and the piece may have come from a tree growing in conditions where the trunk was not stressed. If this seat was made from a tangentially sawn plank the full width of the tree it would probably have had to be reinforced to function, as there was no evidence for this it is likely to have been made from a radial board from a 0.8m or larger diameter tree. Such a tree would not be local, which may be confirmed by dendrochronology. The sides of the seat were all rounded and a fine notch had been cut along each side into the top of this curve. The hollowing of the seat was carefully done and no tool marks survived, except where the flat surface had been reused as a cutting board. These cut marks did not occur within the hollowed area, suggesting that this was a seat and not a shallow bowl on legs. One of the most interesting pieces of woodworking from the site was a repair that had been carried out on a large reused timber also from RB Well 12 (<041> [4098]). This 1.62m length of 170mm diameter Ash had been split in half down its length and then had two through mortises cut near its ends and a circular 50mm diameter hole cut between them. One of these mortises, which was cut 270mm from the wider end of the trunk, was 125mm long by 70mm wide. The trunk had split at this end and a repair had been carried out by cutting a channel across the split on the flattened rounded face of the trunk. This channel had slopping sides so that a dovetailed cross sectioned lath could be hammered in into the channel. This was also nailed in place on either side of the split. Although certainty is impossible there is nothing in the standard or technology of the woodworking of this piece to indicate that it was not originally part of a nearby structure at the site. There was no use of a saw and it is functionally competent but relatively rough and ready work; the tenons are not regular and the finish on the inside of the tenons and circular hole was adequate but not carefully finished. The function of this piece is not obvious, but the mortise hole that has split is much bigger than would be necessary to join two pieces of wood together. It may therefore have been more robust for some other reason; the need to repair the split suggests that movement may have been involved. There was no wear-evidence to indicate that anything had rotated in the circular hole in the centre of the piece, it may have held a stationary piece of wood with a number of facets on it. Although it is impossible to determine precisely what this piece is, it is clearly a structural timber that was purpose made for a particular role and had been repaired whilst in use. 323

220 Woodworking Evidence Three distinct sets of woodworking are present in the assemblage; firstly the remains of portable objects made by skilled craftsmen operating elsewhere. This includes the skill of the lath turner working on the furniture finial who used several turning centres to achieve oval sections, the backed and hollowed stave from a coppered vessel presumably for drinking less than 150mm diameter with a carefully shaped rim and the well-shaped stool seat. These all attest to long distance trade, which may be confirmed and clarified through species identification and dendrochronology. Then there are the locally made simple but effective pointed, cleft and hewn well parts and objects associated with them. These can be sub-divided into two groups, defined principally by the use of saws. Although sawn planks were used in well construction at the site several of these pieces were clearly reused, additionally all the sawn timbers were slower grown than the cleft timbers and there were no saw marks on any of the cleft timbers. This suggests that the sawing took place elsewhere and the inhabitants of the site utilised re-used sawn planks. The woodworking that appears to be directly linked to the sites inhabitants takes the form of trees split up and cleft with axes. Despite the lack of evidence for skilled rather than functional woodworking by the site s inhabitants, the axe blades recorded were not only sharp, but their corners were not rounded suggesting good quality well maintained axes of approximately 70mm diameter. The few 4-inch nails used had been hammered in cleanly in what must have been awkward and slippery conditions within a well. Timber had been stopped from splitting by pre-drilling with double holes when hammering in nails through near the ends. Careful sectioning of the oak corner posts along the nail shanks revealed no evidence pre-drilling of these. The repair to structural timber ([4098]) whilst it was in use suggests knowledge of simple and effective methods of repairing splits in timbers. From the remains we can identify that the sites inhabitants possessed a tool kit of well maintained axes with narrow blades; one was 71mm wide, an 18mm wide chisel and drills of approximately 5mm and 8mm, although it is not clear whether these drills were augers or gimlets. The sites inhabitants appear not to have possessed saws, but these were in use elsewhere locally and had a saw rate of 3 4mm per stroke. The technology of the site s well construction can be summarized as utilizing either reused materials or locally grown timber. Woodland Resources Although the sample was small a few long pieces of round wood and halved round wood were recorded with the branch free growth habit associated with multiple stemmed trunks. It is likely that these represent the product of coppiced stools or pollards. In RB Well 11 (F.1148) the sails were selected for size rather than being the even aged products of a periodic cutting of a panel of coppice stools. The 1.62m length of slightly sinuous stem used for ([4098]) was also branch free, suggesting a small 0.17m diameter trunk rather than branch wood. Several outer tangential splits from m diameter trunks of both oak and other species were used; this timber was generally fast grown and was likely to have come from the more open conditions of hedgerows or steam edge trees. Two oak trees of over 0.45m diameter had been split up radially for use in the wells; their use suggests not only the importance of wells but also the availability or protection of trees so that they were able to reach this size. The pollen analysis (Boreham this volume) provides indications of the tree species growing locally. This appears to broadly correspond with the preliminary visual identifications of species, although these need to be confirmed. Dendrochronology and Species Identification In total eighteen dendrochronology samples were identified, although several of these are duplicates where two samples from the same piece of wood are required to obtain the full sequence. This reduces the number of effective samples to 13. In several instances there are multiple dendrochronology samples from a single feature. The purpose of some of the 324

221 dendrochronology samples is not to date features per se, but to either determine how old a piece of wood was when it was reused/deposited or in some instances to identify the ultimate geographical origin of a piece of wood. In total 84 samples were identified as requiring microscopic species identification, there has however been an element of inadvertent double sampling and additionally some material comes from poor groups. The number of species identification samples could realistically be reduced to 64 with no appreciable loss of information beyond possibly losing a few species from the overall site list. Although the quantity of wood and timber recovered was not large it is significant for a number of reasons. Most Romano-British discoveries of wood and timber from come from military or urban contexts and this group from a rural context is therefore important. Additionally, discoveries of Romano- British wood and timber are rare locally, although a well with three tiers of tangentially sawn dovetailed plans resting on foundation of sawn timber including a reused structural timber was recently excavated at the nearby NIAB site (Luke et al. 2013, 7). Some of the skilfully made artefacts provide important evidence for long distance trade, whilst the well linings constructed by the site s inhabitants provide vivid insights into a number of aspects of their environment and lifestyle. A number of the items should be illustrated and species identification and dendrochronology of selected items should take place. Ceramic Building Material Grahame Appleby A total of 1123 fragments of brick and tile weighing 92.09kg were recovered during fieldwork (Fig. 3.61). 51 fragments (5799g) were identified of post- Roman attribution, nineteen of which (523g) are considered intrusive. Where possible, tile form was identified, although unclassified pieces constitute the single largest category, 38% (409) of the total (40.27kg; 46.68%). Forty-seven fragments (2849g) were recovered from post-roman contexts which includes 44 unclassified fragments considered to be of probable Romano-British origin. The fragments varied from large unabraded pieces to small fragments; no complete examples were recovered. An assessment of the fabric used in the manufacture of the tiles revealed that they were predominately made from well prepared clays with inclusions ranging from small shell and or flint fragments to relatively large, but infrequent, angular small pebbles. The visual assessment further revealed that the clay used for the manufacturing of tegulae and box-flu was markedly less sandy than that used for making imbrex roofing material. The clay was fully converted to ceramic and the majority of tile displayed surface colours ranging from oxidised orange to deep red, with larger (i.e. thicker) fragments revealing reduced, grey interiors. Several fragments were manufactured from buff, marl-type clay and thin sectioning of these may reveal the source of the clay used for their manufacture. In addition, four Collyweston (stone) tiles were recovered, all from Romano-British features. Analysis of the distribution of identified tile fragments (662 pieces) by feature type showed that nearly 86% of the material (569 fragments) was recovered from well contexts (Table 3.72), although only accounting for 52.17% (24kg) of this sub-assemblage. Interestingly, the well assemblage (Table 3.73) was almost equally divided between tegulae and imbreces, with only small numbers of box-flu and pilae; the remaining material from the wells was unclassified. 325

222 19th Century Quarries Archaeological features All tile by weight (g) Box Flue (g) Pilae (g) Sesquipedalis (g) metres Figure Distributions of Romano-British ceramic building material by weight, showing all material, box flue, pilae and sesquipedalis

223 Collyweston Imbrex Pilae Sesquipedalis Tegulae Box-flu Total Unc. ttotal Beam slot (159g) (159g) (266g) (425g) Ditch (344g) (1032g) (156g) 2645g) (4925g) (429g) (9531g) (10.46kg) (19.99kg) Furrow 4 4 (808g) (808g) Hollow 4 4 (79g) (79g) Land drain 2 2 (65g) (65g) Later infilling (252g) (147g) (824g) (378g) (1601g) (904g) (2505g) Occupation soil 1 1 (52g) (52g) Pit (300g) 925g) (1079g) (4374g) (6678g) (9178g) (15.86g) Posthole 1 1 (193g) (193g) Quarry pit (209g) (530g) (668g) (2630g) (4037g) (3889g) (7926g) Subsoil 6 6 (182g) (182g) Well (9794g) (1062g) (2430g) (8992g) (1726g) (24kg) (14.2kg) (38.2kg) Total Table 3.72: Tile quantification by form and feature. Due to the small quantity of tile present in individual features and specific feature types, such as the 49 identified fragments from 37 ditches, and a total of 192 pieces from 86 ditch features, only informative elements of the assemblage from non-well features are considered in the discussion; summaries by specific feature are available in the archive. The one exception to this is pit F.1997, which has a clay lining with fragments of tile fragments and tesserae. The bulk of this material was unclassified, with the remainder dominated by tegulae, with some imbrex and pilae (Table 3.73). 327

224 Feature Imbrex Pilae (470g) Sesquipedalis 1 (1105g) Tegulae Box-flu Unk. Total 1 (137g) 1 (103g) 6 (253g) 1 (214g) (162g) (377g) (770g) (3213g) (1512g) (1008g) (470g) (1056g) (919g) (65g) (2238g) (341g) (649g) (1731g) (142g) (2831g) (5517g) (3178g) (450g) (267g) (2892g) (1726g) (1936g) (288g) (624g) (1216g) (288g) (52g) 1997 Table 3.73: Quantification of tile from wells and lining of pit F (925g) (866g) (4374g) (9001g) 6 (253g) 3 (1712g) 1 214g) 1 (103g) 1 (162g) 24 (4360g) 476 (2520g) 22 (2445g) 1 (65g) 11 (3228g) 61 (10.2kg) 70 (10.45kg) 27 (2128g) 2 (340g) 40 (15166g) Quantitatively, the tile assemblage from North West Cambridge is (including the 225 fragments weighing 23.35kg from recovered during the evaluation) one of the largest assemblages from the Cambridge environs (excluding the Romano-British town). To qualify this statement, less than 50 fragments of brick and tile were recovered from the Addenbrooke s environs, with a similar lack of material from for the whole of west and southern Cambridge. The notable exceptions to this pattern are the assemblages recovered from Vicar s Farm (730, weight 5.6kg; inc. 37 tegulae, seventeen imbrex, five box-tile, 83 brick; Monteil in Lucas 1999) and the various Babraham Research Campus excavation assemblages (799, weight 99.71kg; inc. 176 tegulae, 82 imbrex, 80 box-flue, 94 floor-tile; Anderson 2005, 2007a, 2007b, 2012). Examining the data by form, the almost equal number of tegulae and imbreces present in the assemblage is to some extent unsurprising. Used in the construction of roofs each tegulae would be covered on each vertical and flanged side half an imbrex, the latter used to ensure that water did not penetrate between the roof tiles. This apparent ratio does, however, need to be treated with a degree of caution as a ratio of 2:1 tegulae:imbrex should be expected if a complete assemblage of unbroken and fragmented material was recovered; it is interesting to note that the assemblages from Vicar s Farm and Babraham do match this ratio. Taking the fragmentation of individual 328

225 elements into account, the ratio of tegulae to imbrex at these three locations may thus reflect different building histories on each site. The single imbrex fragment from beam-slot F.920 and eleven unclassified fragments from F.900, and the complete lack material from the other sixteen beam-slots identified across the site suggest these buildings were either carefully dismantled and the material recycled or reused or the beam-slot buildings at North West Cambridge were roofed with wooden shingles or thatch. In contrast, the ratios of these two tile forms at Vicar s Farm and Babraham may reflect the abandonment and wholesale collapse or demolition and clearance of buildings with little thought to the systematic recovery of reusable material. The North West Cambridge assemblage also revealed one slightly unexpected result the high and roughly equal number of tegulae and imbreces from well F.1168, possibly representing a collapsed rood structure associated with the well or deliberate dumping of roofing material from a nearby building. A significant failing of a number of studies of Romano-British brick and tile is the assumption that the presence of tegulae and imbreces represent the nearby presence of a high status building or villa, even where assemblages are very small, for example the quantity recovered at Knobb s Farm, Cambridgeshire, where a villa has been identified and confirmed on less than 60 fragments (Anderson 2008, 2010 for quantities and type). What is more equivocal is the recovery of box-flu, pilae and sesquipedalis heavy floor tiles placed on the top of a column of pilae at North West Cambridge. These three tile forms are direct evidence for a hypocaust system, and seven fragments clearly show extensive sooting (three fragments provisionally identified as imbreces may also be flu or chimney fragments, although the identification of chimneys and external flues in Romano-British buildings remains an area of debate). As such, this material demonstrates that a villa-type building (either in form, function or construction technique, e.g. temple) or bath-house, or both, were located nearby, possibly near to well F.2044, from which four fragments of box-flu were retrieved; one fragment measured 330mm long by 130mm wide, with a square-shaped profile and largely complete entry/exit hole for flu gases and two other fragments from this feature had traces of plaster adhering to them. The unabraded nature of the assemblage and retrieval of 39 pieces of tesserae, 35 fragments of plaster, opus signanum and window glass during the evaluation phase of excavation (Herring & Appleby in Evans & Newman. 2010, 111, 113) certainly lends further support to this interpretation. Due to the site-wide distribution of the assemblage, drawing firm conclusions about the number, scale and function of the buildings that were roofed and erected incorporating this material is slightly problematic. Nonetheless, the presence of a building of some pretension in the immediate vicinity with a hypocaust system and tessellated floors is highly probable. Further analysis of the assemblage, for example distribution plotting and improved identification of type (e.g. pilae versus tegulae/sesquipedalis, imbrex versus antifex not identified) may help reduce the potential number of zones within the site where this building may have been located. 329

226 Building Stone Simon Timberlake Stone Tesserae Tesserae were recovered from only one feature during the excavation; none were recovered from this area during the evaluation although Site VII produced 39 stone and tile tesserae associated with other high status building debris (Appleby in Evans & Newman 2010, 113). Some of these tesserae show evidence of wear and polish on their upper (floor) surfaces, as well as evidence for a still earlier phase of mortar attachment and/or covering with red painted wall (or floor) plaster. Following this first use, the tesserae then appear to have been reconstituted within a floor where they were embedded (i.e. cemented) within a plain white mortar. <2506> F.1997 ([5975]) - Some 34 small broken-up cubes of stone ranging between 10mm x 7mm x 5mm and 15mm x 12mm x 10mm in size, and 6g and 24g (each) in weight (total weight 0.55kg) were utilised in the lining of a pit associated with RB.2A. With the exception of one or two pieces which are made of a hard chalk rock (clunch), all of these are composed of the same stone; a hard bioclastic shell debris and partly-oolitic limestone, most probably extracted from a better cemented horizon within the Lincolnshire Limestone of North Cambridgeshire/ Southern Lincolnshire. Stone Roof Tiles A total of 12.89kg of stone roof tiles, mostly of local Collyweston Slate was recovered from the fills of Romano-British wells, ditches and quarry pits on RB.2B, with a smaller amount of tile from the fills of features located on RB.2A (Table 3.74; Fig. 3.62). An analysis of the fragments suggests that the tiles (slates) were typically small; the width of one size at least being c. 135mm and probably mm long (and between 10 20mm thick) with single holes punched through the middle of the top edge to take an (iron?) nail. Collyweston Slate is a sandy fissile limestone with a slightly micaceous content which was quarried from the Lower Lincolnshire Limestone (Middle Jurassic) beds near Collyweston or else Easton on the Hill, Northamptonshire, from Romano-British to Post-Medieval times. A much smaller amount of slaty (fissile) limestone rocks probably originating from the nearby Blisworth Limestone seems also to have been used. It seems likely that many tons of Collyweston Slate would have been used in Romano-British buildings near to Cambridge, though the costs of transporting these from the quarry sites some 50 60km away (perhaps by boat along the River Nene and then down through the Fens via Car Dyke, or else by road from Water Newton via Godmanchester) suggests that these would only here have been used on relatively high status buildings. More than likely the occurrence of these at the rural farmstead settlement(s) of North West Cambridge represents a re-use of building material once associated with nearby villas or the high status Romano-British settlement at Girton, or alternatively this material was dumped here within wells, quarries and ditches following its replacement at the aforementioned site(s). 330

227 19th Century Quarries Archaeological features Building stone (kg) Stone roof tile (kg) metres Figure Distribution of Romano-British stone roof tiles and building stone by weight

228 Cat. No. Feature no./ SF Context Site location RB.2B RB.2B infill RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2B RB.2A quarry Description uncut/ unfaced frag of roof slate x7 frags of roof slate incl 1 large x2 frags of roof slate x1 split frag of roof slate x4 frags split roof slate x2 frags roof slate (burnt) x2 frags of roof tile without faced edged small corner frag of tile with two cut edges frag slate with one edge preserved + nail hole x3 split frags roof slate; one with nail hole on break c. 10mm diameter corner of slight weathered tile thin split frags of burnt slate small corner frag of burnt slate corner of burnt slate x5 frags of roof slate; one with corner and all slight Dimensions (mm) 125x 100 x 13 (thick) 140 x 160 x 16 + smaller 23 thick largest 125 x 110 Wt (kg) x x 75 x x 130 x x 45 x x 100x largest corner frag 100 x x 60x x x 45 x x 30 x (a)120 x 95 x 15 and 55 x 50 (b)100 x 80x 11 (a) 0.53 (total) (b) Geology Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate black greywacke Notes some slight burnt may have trace stain of lichen break of pick hole (hourglass) for nail 332

229 RB.2A well RB.2A well RB.2A well RB.2A well RB.2A well RB.2A ditch RB.2A well RB.2A RB.2A RB.2A quarry RB.2A ditch RB.2B well RB.2B well RB.2B quarry burnt 0.22 slate (Carb?) (a) x4 frags of slate; one weathered, one with edge (b) x3 frags of rough roofing slate roughly cut piece of roof tile (a) roof slate with one edge (b) rough roof slate x1 burnt + x1 weathered small frag x1 weathered and burnt frag x5 frags of split roof slate weathered frag roof slate x2 adjoin frags of corner of roof tile x1 frag weathered roof tile x1 frag weathered split roof tile frag of roof slate x2 corner frags roof slate with worked (chipped) edges x9 frags broken roof tile incl prob whole top edge of one slate x10 frags split roof slates fairly (a) 80 x 40 x 12 to 130 x 65 x 16 (b) 140 x 80 x 16 AND 115 x 95 x x 130 x 19 (a)105 x 80 x 15 (b) 130 x 70 x 20 (a) 0.6 (b) (a) 0.26 (b) x 40 x x 55 x largest 140 x 110 x (total) 60 x 45 x x 70 x x 85 x x55 x x 50 x x 85 x x 90 x (width top) x 85 x 16 largest 110 x 80 x 18; smallest 35 x (a) Collyweston Slate (b) peletoid bioclast oolite Blisworth Lmstn? Blisworth Lmstn (fissile Gt Oolite?) (a) Collyweston (b) Blisworth Lmstn? Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate amongst burnt building stone etc. slight burnt slight burnt 333

230 Table 3.74: Stone roof tiles. RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2B ditch RB.2A ditch RB.2B quarry RB.2B ditch RB.2B quarry weathrd: one broken along small nail hole (4 5mm diam) x2 large frags of roof slate; one v weathrd x1 corner frag of roof slate x7 frags of roof slate incl split ones x1 frag split roof slate x3 frags split slate x1 small frag roof slate x1 v weathrd and waterworn slate x2 small split corner frags roof slate x1 small frag of corner of slate (burnt) 140 x 115 x x 90 x x 100 x largest 70 x 85 x x largest 100 x x 30 x x70 x x 70 x x 53 x Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate Collyweston Slate some burnt and fracture Column Fragment or Stone Post-pad A crudely dressed disc-shaped fragment of masonry recovered from a RB Well 12 (<1577> F.1402 ([4100]; dimensions = mm x 50mm ([thick]; weight > 5kg). The imperfectly circular shape of this was achieved by the crude vertical chiselling of this from a split slab of local fossiliferous sandstone; possibly the Kellaways Rock (Callovian) of the local Lower Oxford Clay (as suggested by the presence of the serpulid fossil Genicularia vertebralis), and most likely this had been collected locally as a glacial erratic. The lower(?) surface of this disc is uneven and slightly convex, and has been barely shaped by chiselling, whilst the upper(?) surface is similarly uneven and slightly concave, and is unshaped. Given the crudity of this work it is difficult to be certain that this was ever intended as a column base; an alternative explanation being that it was used as a post pad or else as a pedestal for a wooden column, the dimensions and chisel-cut nature of this being quite similar to the example from Langdale Hale (Appleby in Evans 2013). 334

231 Building Stone Over 20kg of very crudely dressed stone was recovered which might best be described as unfaced walling material, all of this dumped as re-deposited stone within features of Romano-British date on Sites RB.2A and RB.2B (Table 3.75). This former building stone consisted of a fairly diverse assemblage ranging from local sarsen (quartzitic sandstone) boulders collected as glacial erratics to (possibly re-used) Barnack Stone (quarried near Stamford) which may have been transported across the Fens to Romano-British Cambridge via the Car Dyke. Other identified stone includes probable Ketton Stone (also Lincolnshire Limestone) and Corallian Limestone which may have its origins closer by, such as from outcrops near Upware. A small proportion of this building material consists of pieces which had evidently been brought here as re-used partly-dressed stone. As well as for walling, some of this material may have been brought to site for the purpose of revetting the wells, or perhaps for providing steps or platforms into their bases. More than 50kg of building stone consisting of large fragments exhibiting a certain amount of chisel dressing and moulding (perhaps being elements from high status buildings or funerary monuments) was recovered from the base of RB Well 16. All of these appear to have been carved from Barnack Stone (Lincolnshire Limestone), a stone quarried at Barnack near Stamford from the Romano-British to Medieval period, and most likely transported across the Fens by boat along the Car Dyke. An additional five un-diagnostic fragments (c. 20kg) of crudely shaped but not faced Barnack Stone material were discarded on site. Additionally, there were some large unshaped cobbles of quartzitic sandstone (2) and quartzitic siltstone (1; c. 15kg) in the same deposit, these probably represent glacial erratics from the local gravels. All of this material had been deliberately placed in the base of the well, presumably as a temporary expedient during construction to provide a firm/dry footing. A few of the dressed stone faces show evidence of square pin holes for iron connecting or locating pins. These would have been set in lead, and would have helped fix slabs of dressed stone in place either for (funerary?) monuments or for building facings (e.g. slabs C and D). One of these dressed slabs (D) may have formed part of a stela, one perhaps associated with a funerary monument. Cat. F. Context Site location Description Dimensions (mm) Wt (kg) Geology Notes RB.2B well frag crude dress masonry made of sarsen stone 190 x 170 x 80 >5 quartzitic sandstone perhaps glacial erratic? Perhaps previous use as saddlequern? RB.2B quarry squareshaped worked block 285 x 210 x 90 4 Barnack Stone cone-shaped hole 60mm x 50mm (deep) pick-cut on 335

232 upper surface + other RB.2A well polygonalshape block with chisel imprint 280 x 280 x 170 Barnack stone perhaps orig discarded due to weathering RB.2A well crudely shaped but otherwise unworked square slab 250 x 220 x 70 Upper Jurassic? bioclastic limestone RB.2A well crudely shaped square/ polygonal block 300 x 230 x 110 Barnack stone RB.2A well x6 burnt and broken up frags largest 120 x 70 x (total) pink ooidal limestone poss Ketton stone (U Lincs Limstn) RB.2A ditch x3 small frags of weathrd oolitic lmstn pink ooidal limestone poss Ketton stone (U Lincs Limstn) RB.2B well x2 small frags of weathrd oolitic lmstn one poss a quoin largest 70 x 65 x ooidal lmstn (Ketton Stone?) RB.2B ditch x1 rough cut but broken block of shelly limestone 120 x 110 x Corallian Limeston (Callovian) with fossil heavily burnt: reused as a hearth stone for cooking on Table 3.75: Catalogue of building stone (walling material). 336

233 <A> F.2044 [7117] - A large dressed square block of Barnack Stone (310mm x 190mm x 245mm (thick)) with traces of chisel marks on at least one face. Probably an example of facing stone or else a quoin from a relatively high status (stone) building. Recovered from the fill of a well within the RB.2A. <B> F.2044 [7117] - A large chisel-dressed moulded slab of Barnack Stone (380mm x 305mm x 110mm (thick)) with a carved stepped (30mm wide) edge along the narrower side, presumably part of the base of a substantial stone structure, supporting either columns or a wall. <C> F.2044 [7117) - A large chisel-dressed slab of Barnack Stone with a rounded (but now damaged) edge along the longest side (dimensions: 500mm x 308mm x 160mm). Unidirectional chisel marks can be seen adorning the top? of the stone, although the underside is uneven and unworked. A square notch (40mm x 30mm x 30mm deep) has been cut into the middle of the top leading edge (interior) of this stone on its upper face; presumably this once held an iron connecting pin (in lead?) which used to help anchor this with others inside of a large stone edifice. <D> F.2044 [7117]. - The largest of the chisel-dressed slabs of Barnack Stone (760mm x 540mm x mm [thick]). Whilst this has no obvious moulding on it, the faces have been very carefully worked to obtain a smooth and shaped finish to the slab, with special attention to the chiselling direction around the margins. It seems plausible this might originally have been intended as a stelae or upright slab, yet it appears to have been unfinished, and was subsequently used as massive masonry, with an oval-shaped hole (105mm x 30mm x 55m) picked into the top middle of the flattest face, perhaps to take a pin to anchor it to an adjacent slab. Animal Bone Vida Rajkovača The assemblage generated the raw count of fragments and a total weight of over 230kg of bone (232,634g), recovered from 841 contexts (Fig. 3.63). The faunal record detailed here includes the hand-recovered material, with remains from heavy residues discussed at the end of this report. The assemblage was analysed in its entirety with a view to providing a clear account of the informative potential of the assemblage (in accordance with English Heritage guidelines published in 2013). The assessment aims to provide an overall view on how much data is present by area and phase, both in terms of physical quantification of faunal data and its interpretative potential. Contexts ranged in date from the earliest prehistory and through to the modern era, with prehistoric and Romano-British occupation leaving more substantial quantities of material culture (Table 3.76). At this stage, the entire pre-roman occupation will be considered as prehistoric. The earlier material, recovered from a range of isolated and settlement features, the ringditches and enclosures roughly amounted to just over 11% of the entire assemblage. The Romano-British contexts, on the other side, generated over 4000 specimens, taking the total to c. 84% of the entire assemblage. Despite there being a clear quantitative difference, the bone quality and preservation did not vary so much between different phases of occupation. In addition to the bone recovered from these excavations, more bone was recovered from 2009 evaluations that derive from the same sites (Table 3.77). 337

234 19th Century Quarries Archaeological features metres Figure Distribution of Romano-British animal bone

235 Phase Frag count % of assemblage total Prehistoric Romano-British Post-Medieval and Modern Unphased Total 4747 Table 3.76: Breakdown of animal bone by phase. Taxon PE3 RB.2A RB.2B RB way 3 Cow Sheep/ goat Sheep Goat 1 1 Pig Horse Dog 1 1 Red deer 1 1 Sub-total to species Cattle-sized Sheep-sized Mammal n.f.i Total Table 3.77: Number of Identified Specimens for all species from 2009 evaluations, the material which derives from the same sites investigated during the 2012/13 excavations. The zooarchaeological investigation followed the system implemented by Bournemouth University with all identifiable elements recorded (NISP: Number of Identifiable Specimens) and diagnostic zoning (amended from Dobney & Reilly 1988) used to calculate MNE (Minimum Number of Elements) from which MNI (Minimum Number of Individuals) was derived. Identification of the assemblage was undertaken with the aid of Schmid (1972), and reference material from the Cambridge Archaeological Unit. Most, but not all, caprine bones are difficult to identify to species however, it was possible to identify a selective set of elements as sheep or goat from the assemblage, using the criteria of Boessneck (1969) and Halstead (Halstead et al. 2002). Ageing of the assemblage employed both mandibular tooth wear (Grant 1982, Payne 1973) and fusion of proximal and distal epiphyses (Silver 1969). Where possible, the measurements have been taken (Von den Driesch 1976). Sexing was only undertaken for pig canines, based on the bases of their size, shape and root morphology (Schmid 1972, 80). Withers height calculations follow the conversion factors published by Von den Driesch & Boessneck Taphonomic criteria including indications of butchery, pathology, gnawing activity and surface modifications as a result of weathering were also recorded when evident. The bone preservation ranged from moderate to quite good (Table 3.78). Surface erosion was recorded on 7.7% of the material. The assemblage was highly fragmented, however, with a small number of 35 measurable elements. Canid was more common than rodent gnawing, although both were recorded on 6.7% of the bone. One aspect of taphonomy, however, clearly prevailed within the Romano-British period and that was butchery. Of 404 recorded specimens, 357 came from RB.2A (90) and RB.2B (267). A more detailed look at preservation recorded for the main sub-sets showed similar percentages (Table 3.79). 339

236 Preservation Context % Frag. % Good Quite good Moderate Quite poor Poor Total Table 3.78: Preservation categories for the animal bone assemblage as a whole: breakdown by context and fragment. PS1 PS2 RB.1 RB.2A RB.2B Preservation No. % No. % No. % No. % No. % Good Quite good Moderate Quite poor Poor Total Table 3.79: Preservation categories for the main animal bone sub-sets: breakdown by context and fragment. Relative to the vast quantity of pottery, the animal bone waste was not as abundant. This may either be due to the difficult recovery conditions, or, more likely, that the bone was deposited elsewhere. The excavations had not revealed any midden-like features or spreads, and it is possible that the bone waste was cleared away from the settlement before being incorporated into occupational layers. Results Pending further fine-tuning of the phasing within the Romano-British occupation and results from other material types, the Romano-British bone will be considered according to area and, in some cases, feature-type. Within this phase, cattle were by far the preferred species, often accounting for more than all other species combined (Table 3.80). Horse is unusually high, represented with numbers similar to sheep/ goat. The fauna from the two main settlement areas was overwhelmingly biased towards domestic sources of food, although a sporadic find of a cervid or a bird indicate these wild resources occasionally contributed to their diet. The ratio between isolated teeth and mandibles can be used to give a gross indication of the fragmentation state of an assemblage and a potential for further study. If we look more closely at the number of loose teeth for cattle and sheep/ goat, these accounted for between 13 27% of the identified count (Table 3.81). According to these numbers, the fragmentation is quite high. Within RB.1 ditches F.2513 and F.2514 generated a combined total of 40 specimens, a figure which amounts to almost half of the bone from the sub-set. Although the recovered bone from these ditches exhibited an overall moderate preservation, a portion of cattle-sized ribs, observed during excavation, had degraded to the point where it was visible, but could not be recovered. Another interesting find was the rack of ribs, possibly sheep, recovered from under the cremation pots (F.1958) from C3. 340

237 RB.1 RB.2A RB.2B Taxon NISP %NISP MNI NISP %NISP MNI NISP %NISP MNI Total NISP Cow Sheep/ goat Sheep Goat Pig Horse Dog Dog/fox Red deer Roe deer Fox ?Hare Rabbit Chicken Mallard ?Pheasant Galliformes Sub-total to species, order or family Cattle-sized Sheep-sized Rodent-sized 1 1 Mammal n.f.i Bird n.f.i Total Table 3.80: Number of Identified Specimens and Minimum Number of Individuals for all species from three Romano-British settlements; the abbreviation n.f.i. denotes that the specimen could not be further identified. RB.2A RB.2B Species % of total loose % of total NISP loose mandibles mandibles NISP for the subset teeth for the sub-set teeth Cow Sheep/ goat Table 3.81: Number of loose teeth for the two main domesticates, and the percentage of the total NISP count per species. Pit F.964 contained a well-preserved and near complete dog skeleton, with an estimated shoulder height of only 23cm. This is quite small, and the curved limb bones seem to indicate dachshund-sized dog. Despite the Romano-British date for the feature, the general condition of bone could be taken to suggest the deposit is of later or even more recent date. Of other points of interest within the Romano-British phase, butchery marks stand out in particular, as the majority of marks were recorded from RB.2. Of the assemblage s 404 butchered specimens, 357 were from RB.2 (88.4%). Marks from the entire butchery chaîne opératoire were recorded on a wide range of skeletal elements, and the techniques and the crudeness in some instances were characteristic of Romano-British material. Marks consistent with gross disarticulation and portioning carcasses into smaller joints were particularly common. 341

238 Taxon NISP %NISP MNI Cow Horse Sub-total to species 14 Cattle-sized 2 Sheep-sized 2 Total 18 Table 3.82: Number of Identified Specimens and Minimum Number of Individuals for all species from Romano-British double-ditched system F.513; the abbreviation n.f.i. denotes that the specimen could not be further identified. Apart from the bone, which came from two discrete dumps, the double-ditched boundary was in general sterile with only other finds being eight pottery sherds (Table 3.82). The prevalence of larger domesticates from peripheral features is not surprising as larger amounts of bone waste are tidied away from areas of immediate domestic activity. Perhaps unsurprisingly, the investigation of Ways 1 3 did not generate much bone (Table 3.83). The range of species broadly reflects that recorded from the rest of the assemblage. RB Way 1 RB Way 2 RB Way 3 Total Taxon NISP %NISP MNI NISP %NISP MNI NISP %NISP MNI NISP Cow Sheep/ goat Pig Horse Dog Sub-total to species Cattle-sized Sheep-sized 9 9 Total Table 3.83: Number of Identified Specimens and Minimum Number of Individuals for all species from Romano-British Ways 1 3; the abbreviation n.f.i. denotes that the specimen could not be further identified. Additionally, some animal bone was recovered from the heavy residues of processed samples, the list of all species is given in Table Although the range of domesticates recorded from this sub-set is similar to that from the hand-recovered material, other small vertebrates were also present such as mouse and vole. These two species and amphibian remains are most likely part of the background fauna and tell us more about the environment rather than animal use. Due to the fragmentation it is uncertain whether the mouse and vole specimens belong to the house mouse and water vole or the field varieties; however, these species are most abundant in open ground with a dense cover of vegetation. Of the total of 912 specimens, 138 (15.1%) were recorded as either charred or calcined. This either implies a degree of waste management was put into action or otherwise burnt material was inadvertently incorporated into occupation layers from domestic contexts. Fragment of a red deer antler tip came from well F With the exception of a tibio-tarsus fragment recovered from the same feature, tentatively assigned to the duck family, it is recommended a small number of bird and fish elements is identified to species level. Same could be proposed for the micro-fauna, although we are not certain this would generate noteworthy results. 342

239 Taxon RB.1 RB.2A RB.2B Way 1 Way 3 Total NISP Cow Sheep/ goat Pig 4 4 Dog 2 2 Cat 1 1 Red deer 1 1 Mouse sp. 1 1 Frog/ toad Vole sp. 5 5 Anseriformes 1 1 Sub-total to species Cattle-sized Sheep-sized Rodent-sized Mammal n.f.i Bird n.f.i Fish n.f.i Total Table 3.84: Number of Identified Specimens for all species from Romano-British contexts; the abbreviation n.f.i. denotes that the specimen could not be further identified. Period Number of Volume Specimens Specimens processed (l) per litre Prehistoric Romano-British Table 3.85: Bone density from heavy residues number of specimens per litre of soil. Although seemingly having produced more bone than the prehistoric contexts, samples collected from Romano-British features contained less bone per litre (Table 3.85) of soil. This is an important observation regarding both sub-sets fragmentation levels, especially if we look at the ratio of bone from the heavy residues compared to the hand-recovered material (Table 3.86). Whilst hand-recovered prehistoric bone was found in similar quantities to that from the heavy residues, the hand-recovered Romano-British material was over four times as abundant as that from the heavy residues. Another main difference between the hand-recovered and sieved bone was the burnt material. Looking at combined figure for both periods for the sieved material, of 1263 specimens, 233 were recorded as charred or calcined, a number which corresponds to 18.5% of the assemblage. Hand-recovered bone showing any signs of burning was very sparse, with 28 specimens or 0.5% of the assemblage. Period Number of specimens Number of specimens hand-recovered heavy residues Ratio Prehistoric : 1 Romano-British : 1 Table 3.86: Ratio of hand-recovered bone versus bone from the heavy residues. At this early stage, prior to any analyses of prehistoric pottery and before any further adjustments in the phasing within the Romano-British occupation, we are only able to assess the assemblage s potential very superficially. Despite seemingly small numbers from prehistoric contexts, their recovery from such an expansive area means we could study changes in this landscape over a longer period of time. Despite the clear evidence for the Iron Age occupation noted during the evaluation phase, the apparent paucity of Middle to Late 343

240 Iron Age remains from the excavations must be addressed. This is especially important if we were to establish whether the Romano-British occupation had its origins during the later stages of prehistory. The basic quantification and the broad characterisation of the assemblage showed the Romano-British occupation was extensive, and perhaps even had a level of specialisation as indicated by the potentially high horse component. The dominant cattle cohort is in keeping with the period, indicating the site was Romanised. The lack of pig remains is quite remarkable, although perhaps not unexpected according to some. It has been suggested that the pattern of high pig tends to be common on Late Iron Age sites which continue into the Romano-British period, with high pig numbers (Fifield 1988, 150). The insignificant pig cohort, coupled with the site s lack of Middle to Late Iron Age occupation, could confirm the site was abandoned during this period. In general, the faunal signature is ordinary. The site has so far failed to produce anything which would be interpreted as high-status, and that is exactly its greatest potential. The site gives us an opportunity to study the origins and the development of rural settlements on the outskirts of Cambridge, and the relationship between the Cambridge and the surrounding areas during the earliest phases of Romanisation. Oyster Shell Jacqui Hutton The assemblage of oyster shells recovered and recorded consisted of the European Flat Oyster family (Ostrea edulis Linnaeus), numbering 294 pieces in total (Fig. 3.64). The majority could be identified as either left or right valves with a remaining 10% un-measureable; there were 147 left valves and 117 right valves with only six sets positively matched. The condition of the shells prohibited any further positive matching. The assemblage from the whole site was fairly small and was recovered from 78 features in total. The shells were recovered from either Romano-British (90%) or Post-Medieval/Modern features (10%). Those recovered from Romano-British contexts were then broken into feature type categories, the majority being recovered from the ditches (Table 3.87). Feature Type Number of Features Romano-British ditches 46 Romano-British quarry pits 3 Romano-British pits 2 Romano-British wells 8 Romano-British postholes 3 Post-Medieval ditches 5 Post-Medieval quarry pits 3 Table 3.87: Feature types with oyster shell. The oyster shells were all examined by eye with all identifiable traits recorded into a database. No attempt was made to accurately date the individual oyster shells at this stage of the assessment. The oyster shells in this case were 344

241 washed before examination. The majority of the valves could be identified as either left or right and all valves were measured where possible. Any evidence of parasitic activity, anomalies or intentional damage was also duly recorded. In total, there were 294 individual oyster shells: 147 were left valves, 117 were right valves, with a remaining 30, which were unidentifiable. Only six sets were matched; these were from F.1020 (well), F.845 and F.958 (ditches), F.968 (quarry pit) and F.579 (subsoil). The general condition of the oyster shells ranged from poor to good. Any wear or damage is a good indication as to whether the shells were primary, secondary or tertiary deposits. However, the edges of the shells were worn, either due to taphonomic and/or cleaning processes. This made the matching of valves difficult and only a very small percentage was accurately matched. Approximately 42% of the assemblage was worn and flaky, which suggests that a certain degree of movement has taken place. The condition of the natural geology could also have an impact on the condition of the shell and also including how they are collected and processed and any mechanical damage from excavation and processing must be taken into account. The majority of the oyster shells were recovered from Romano-British contexts, primarily from the ditches, although only in fairly small numbers (below ten individual shells). 10% were recovered from Post- Medieval/Modern features, but many of these probably derive from underlying Romano-British features. See Table 3.86 for a breakdown on the feature types and number that contained oyster shells. The larger assemblages came from the wells. There were two Romano-British wells that contained the majority of the oyster shells( F.1168, 2044). The sixteen shells recovered from F.1168 were in fairly good condition, and had evidence of ligaments and Polydora ciliata. In well F.2044, 93 shells were recorded and although these were in fairly good condition, it was not possible to successfully match up any sets. Evidence of cuts and/or notches was on ten of the shells, all of them on the left valve. Marine polychaete (marine worms) was evident on 22 shells, on both the right and left valves; 21 had Polydora ciliata and a further one with bore holes. Evidence of the sponge Cliona celata was recorded on one valve. The epibiont organisms (an organism that lives on another organism) are an important indicator of local and regional environmental and can sometimes be used to trace where the oysters were originally cultivated (Claassen 1998). Marine polychaete (Marine annelid worms) are the most visible signs of infestation and on such example is Polydora ciliata, and as can be seen in Table 3.88, this was the most prevalent on the shells. Another infestation noted on one of the valves was Cliona celata, a sponge. The bore holes that were evident on the valves are from predatory marine gastropod molluscs. A number of valves had evidence of notches and cut marks, which indicate that they were for food consumption. These were recorded on shells that had clear evidence of notches, there could be more but due to the condition of some of the shells, this cannot be verified. Table 3.89 below illustrates where they were found. 345

242 19th Century Quarries Archaeological features metres Figure Distribution of Romano-British oyster shell by MNI count

243 Identifiable Traits Left Valve Number Right Valve Polydora Ciliata Cliona Celata 1 0 Bore Holes 7 2 Ligaments attached 2 10 Oysters Attached 8 1 Notches/Cut marks 19 5 Chambered 24 4 Table 3.88: Shells with identifiable traits. Feature Feature type No. F.960 Ditch 1 F.845 Ditch 2 F.1003 Ditch 2 F.1076 Ditch 1 F.1908 Ditch 1 F.2040 Ditch 3 F.2051 Ditch 1 F.7070 Pit 2 F.1020 Well 1 F.2044 Well 10 Table 3.89: Features with notched shells. The assemblage of oyster valves recorded from this site has provided evidence of infestations and notch marks. Although the majority of the assemblage appeared to be discarded shells in somewhat small numbers, the shell deposits in the wells are worth further study especially in aging them. With further examination, such as ageing, it could be deduced whether the oysters were specifically cultivated and perhaps with the study of the infestations in closer detail, a point of origin could be suggested. With the evidence of the notches on some of the valves, the assemblage possibly derives from human food waste. Human Bone Natasha Dodwell Sixteen inhumations and twenty-five deposits containing cremated human bone which have been dated to the Romano-British period were. In addition to the burials, six fragments of disarticulated bone were also recovered. The results of this assessment are presented by site/area as defined by the site director, primarily in tabular form. 347

244 19th Century Quarries Archaeological features Cremation Burial Stray Human Bone metres Figure Distribution of Romano-British cemeteries, burials and stray human bone

245 No. inhumations No. cremations No. disarticulated bone (contexts) Site RB Site RB.2A Site RB.2B (4) Site RB.2C 12 (13) 0 0 Site RB.2D Table 3.90: number/type of burials from each site. Numbers in brackets refer to material found in evaluation phase. All of the material has been scanned; with inhumations the completeness of the body was recorded and the presence of skull, innominate and dentition noted as well as the potential for taking metrical data including stature. The degree of surface erosion/abrasion was also recorded using the criteria in Brickley and McKinley (2004, fig. 6). For the cremated bone all extraneous material was removed and the bone passed through graded sieves to give an objective sense of how fragmented the material is. All of the bone was quickly scanned to assess approximate age of individuals, to identify any multiple burials and any burnt or unburnt animal bone. Site RB.1 Site RB.1 had a small cremation cemetery (No. 5), consisting of five burials and an additional seemingly isolated burial. A fragment of adult-sized femur (134g) was recovered from a ditch, F.2524 [8867]. With the exception of F.2666 where lots of the fragments are blue/black in colour (indicative of lower temperatures or a poorly maintained pyre) bone fragments are predominantly a buff white colour characteristic of complete oxidisation. Feature Burial type Location 2606 unurned Isolated 2663 unurned 2664 unurned 2665 urned Cemetery 5 Cemetery 5 Cemetery 5 Weight of cremated Age bone (g) 5- >10mm total 10mm adult adult adult Sub adult/adult urned Cemetery adult urned Cemetery adult Table 3.91: cremation burials from Site RB Max bone Comments length (mm) plough damaged Plough damaged Plough damaged Heavily plough damaged Plough damaged Plough damaged Site R.B.2A The human bone at Site RB.2A relates to Cemetery 3. Five urned cremation burials associated with at least one ring ditch and one penannular ditch are the earliest funerary activity in this area. These monuments then continue as a foci for four inhumation burials. In addition a disarticulated adult left fibula, SF 799 was identified; it could derive from disturbance of the inhumations, although it was found some distance from them. 349

246 All but one of the cremation burials are heavily truncated. All of the bone is buff white in colour. An unburnt neonate humerus was recovered from burial F.2005 and green staining suggestive of contact with a copper alloy object was noted on a burnt bone from F Animal bone (both burnt and unburnt) was identified in two, possibly four features. Feature Burial type 824 urned 910 urned 1958 urned 2005 urned Location 2ndary burial in R.D Centre of R.D Weight of cremated Age bone (g) 5- >10mm total 10mm adult adult adult adult in centre of adult urned penannular ditch Table 3.92: cremation burials from Site RB.2A Max bone Comments length (mm) 41.2 Heavily truncated. Animal 38.3 Heavily truncated.? animal bone. Cu alloy staining Relatively intact. Four vessels. Animal bone 77.8 Heavily truncated. Neonate humerus (unburnt) Truncated.?animal Feature Skeleton Age Skull Pelvis Joint measurement Dentition Body Erosion grade 1931 [5811] adult no no no no <5% [6824] adult no yes yes no 75% [7045] immature yes yes yes yes 75% [7298] immature yes no no no 50% 1-3 Table 3.93: inhumation burials from Site RB.2A. Site RB.2B Disarticulated human bone was recovered from three ditches in this area; these are in addition to the two fragments of human skull recovered from the surface of an unexcavated ditch in Trench 245 in the Evaluation. Although this site is primarily Romano-British the pottery indicates that it originated in the Late Iron Age c. 50BC-50AD and the assumption is that the human bone probably relates to this early phase. The most significant observation made is the evidence of polishing on the end of the femur suggesting that it may have been used as a tool. Cat Qty. Wt Slot elements Context Feature (g) No Neonate femur l. adult femur, prox & mid shaft with evidence of working/polishing Neonate scapula & humerus Table 3.94: disarticulated human bone from Site RB.2B. 350

247 Site RB.2C The human bone from this area relates solely to a single small inhumation cemetery (Cemetery 2). There are 14 graves, one of which, F.658 contained no bone but is child-sized and one inhumation, F.76 was studied in the evaluation phase. In addition two loose teeth were recovered from the subsoil and are thought to indicate the remains of an additional, ploughed out inhumation. Skeleton F.631 had been decapitated and the skull had been placed between the legs. Feature Age Skull Pelvis Joint measurements Dentition Body Erosion grade 76 adult Studied in the evaluation phase 500 adult y n n y % 501 adult y y y y 75% adult y y y Y 75% immature 2 teeth N/A only 629 adult y y n y 75% adult y y y y 75% adult y y y y 75% ?immature n n n n No bone 666 adult y y y y 75% adult y y n y 50% adult y y y y 75% adult y y n y 75% adult y y y y >75% 2-3 Table 3.95: inhumation burials from Site RB.2C RB.2D Human bone recovered from this area relates principally to Cemetery 1, which appears to start as cremation cemetery and then develop into a mixed cremation/inhumation cemetery. There are twelve cremation burials and two inhumations in the cemetery (plus one further grave with no bone). In addition two isolated cremation burials were recorded. Feature Burial type 502 urned 503 urned 504 urned 505 urned 506 urned 507 urned 508 urned 535 urned 542?unurned Location Cemetery 1 Cemetery 1 Cemetery 1 Cemetery 1 Cemetery 1 Cemetery 1 Cemetery 1 Cemetery 1 (slight outlier) Cemetery 1 (slight outlier) Age Weight of cremated bone (g) 5- >10mm total 10mm adult Subadult/adult Subadult/young 29 7 adult adult Adult & immature adult adult Adult & immature adult Max bone Comments length (mm) Untruncated.?animal bone V. heavily truncated V. heavily truncated V. heavily truncated Untruncated 83.3 Untruncated. Lots of poorly fired bone untruncated Untruncated. Animal bone Truncated. Animal bone 351

248 549 urned Cemetery 1 (slight outlier) 638 urned isolated 823 urned 869? 936 urned Cemetery 1 Cemetery 1 Cemetery 1 Table 3.96: cremation burials from Site RB.2D. adult Adult &?immature adult Subadult/adult 0 1 adult Untruncated. Animal bone 73.56?untruncated.? Animal bone Untuncated. Animal bone V truncated/disturbed Untruncated. Cu alloy staining on bone The burials at this site were less disturbed than those identified in Sites 1 and 2a; nine appear to have all of the bone that was originally deposited. The vast majority of the bone fragments are a buff white colour although several deposits (F.502 F.505,F.507) have more poorly fired fragments. Animal bone was identified in four, possibly six burials. Two, possible three duel burials were recorded. Skull Pelves Joint measurements Dentition Body Erosion grade 849 adult yes yes no yes 75% Adult?M no no no no <25% Older yes no no no <25% 5+ subadult/adult 867?immature no no no no No bone N/A Table 3.97: Inhumations from RB.2D. The inhumations are generally in a very poor fragmentary condition and this combined with the preservation of the cortical bone will inhibit the recognition of pathological conditions. However, the majority have the necessary diagnostic elements surviving which will allow aging and sexing methods to be applied. In addition a basic inventory of all surviving elements and dentition (the majority of teeth are loose) needs to be made for the archive. The vast majority of bone was buff white in colour indicative of complete oxidisation. The weights of bone recovered from the cremation burials range from g with most features containing between g of bone (although this figure includes truncated / disturbed cremation burials and double burials). Bone weight range No. of Burials 1-100g g g g+ 4 Table 3.98: Cremation bone weight frequencies. Both adults and immature individuals were identified as were dual burials and burnt and unburnt animal bone. It will be necessary to analyse all of the cremated material in more detail to refine this information. For those deposits where the excavator felt that there was little or no truncation (i.e. all of the bone that had originally been deposited survived) it is recommended that they are sorted by body part (skull, axial, upper limb, lower limb) as recommended by McKinley (2004) 352

249 Amongst the disarticulated material the polished femur F.1729 is of particular interest and is reminiscent of the worked human bone recovered at Trumpington Meadow (Patten & Lucy forthcoming); it deserves detailed recording and further analysis (including use-wear analysis. Plant Remains Anne de Vareilles 175 bulk soil samples, equating 1990 litres of soil, were processed using an Ankara-type flotation machine (Fig. 3.66). The flots were collected in 300µm aperture meshes and the remaining heavy residues washed over a 1mm mesh. The flots and heavy residues were left to dry naturally indoors prior to analysis. J. Hutton sorted the greater than 4mm fractions of the heavy residues by eye. Dry flots were scanned under a low power binocular microscope (6x 40x magnification) by the author, in order to assess the quantity and quality of plant macro-remains. Four waterlogged samples were wet-sieved through a stack of sieves ranging from 4mm to 300µm, and analysed under the same microscope. Sixteen waterlogged samples remain unprocessed. Nomenclature follows Zohary & Hopf (2000) for cereals and Stace (1997) for all other flora. RB.1 RB.2A Settlement 2B-D Routeways C3 ringditch Dry samples Volume of soil (L.) Waterlogged samples but processed as though dry 6 (1 analysed) Table 3.99: Environmental samples. 14 (3 analysed) 0 0 Both carbonised and waterlogged plant macro-remains were recovered. The condition of plant remains varied widely across the site, from features with very occasional and heavily abraded grains and seeds, to those with dense assemblages of pristine delicate plant parts. 51% of the dry samples contained no charred plants at all. 41% had cereal remains, but only 10% of those (4% of the 175 total) contained more than 50 components (cereal grains and chaff; Fig. 3.67). Waterlogged conditions also varied. Seeds did not survive well in all of the features from which waterlogged wood was retrieved, and it is evident that there has been a gradual decline in the level of the watertable. Rich assemblages of well-preserved waterlogged seeds were retrieved from the deeper wells, where carbonised plant remains were also recovered. Whereas the preservation of uncharred seeds was very much dependant upon propitious waterlogging conditions, the ubiquity and concentrations of charred plant remains appears to reflect a genuine pattern of activities across the site. Mollusc shells were only found in a selection of samples and tended to represent Planorbids and other similar species that live in shady, frequently wet zones such as deep ditches. Intrusive rootlets and modern seeds, namely chenopods, were present throughout and represent a light level of bioturbation/soil movement. 353

250 19th Century Quarries Archaeological features metres Figure Distribution of all processed Romano-British environmental samples

251 RB.1 - The corn-dryer F.2567 contained the most charred plant remains. It had less than ten cereal grains (apparently all spelt Triticum spelta), about fourteen glume bases (hulled wheat chaff) and a few wild plant seeds representing arable weeds. The paucity of remains in a corn-dryer where one might expect accidental charring of grains and chaff to occur, is not altogether surprising. Burnt waste was probably cleaned out and deposited outside the dryer in pits or ditches, or simply scattered on the ground. Wild grasses, dock (Rumex sp.), black bindweed (Fallopia convolvulus) and scentless mayweed (Tripleurospermum inodorum) were all likely arable weeds and their presence suggests crops were dried before all weeds had been removed. The latter were presumably sieved and picked out along with glume bases after the final threshing. Vitrified charcoal and hard, porous, clinker-like carbon was found in a few samples, suggesting a level of activity involving intensive fires. These materials should be identified in more detailed to confirm whether they are of mineral or organic origin, and mapped with the distribution of hammerscale finds to locate areas of smithing and/or other industries. Five waterlogged samples were obtained from a pit, a ditch and a well. Seeds from samples in pit F.2756 and well F.2768 [9114] were not numerous and poorly preserved. The other three samples were floated and the flots left to dry, though these could be sorted to get a general indication of environmental surroundings. The other two samples from well F.2768 had many crowfoot and fool s water-cress (Apium nodiflorum) seeds, both indicative of damp to wet mud, possibly from within the well itself. This description supports the entomological evidence which points to slow-flowing, stagnant or even temporary presence of water (see Smith, below). Sample Feature Context Volume (l) Preservation Charcoal volume Charcoal ID (<2mm) Cereal type Poor Low Yes Most charcoal from heavy residue Poor Low No nothing Poor Low Yes Most charcoal from heavy residue Poor Low Yes Spelt Mod. Low Yes Cereal grain Mod. Low No Spelt Good Low Yes Cereal chaff - gb. ++ gb. Other crop Weed' seeds Note Most charcoal from heavy residue - 1 knotgrass + c. 14 glume b. Grasses, Rumex, F.convolvulus, spike-rush & Tripleurospermum was WL. 1 charred culm node 355

252 Mod. Low Yes Spelt & Frags + + gb Mod. Low No Mod. Low Yes N/A 0 + Grasses and 1 Rumex. Hard, porous clinkerlike lump. 2 Aphanes/ Alchemilla sp.? all charcoal from heavy residue only modern rootlets Poor Low No nothing Poor Low No Many vitrified specks of charc Poor Low No Spelt - nothing Poor Low No nothing Poor Low Yes nothing Mod. Low Yes Spelt + - gb. 1 pulse wild? Was WL but preserv isn't great. 3 grains Mod. 0 was WL but preserv isn't great N/A 0 nothing Poor Low No nothing Poor Low No nothing Poor Low No nothing Poor Low Yes nothing Poor Low Yes Frags + All carbon vitrified and 'bloby' Poor Low No nothing Poor Low No Good Low Yes Good Low Yes - gb. a little charcoal and nematode egg sacks WL. Planorbids and LOTS of Batrachium seeds. WL. LOTS of A.nodiflorum. All charcoal from heavy res Poor Low No nothing Table 3.100: Scanned environmental samples from Site RB

253 19th Century Quarries Archaeological features 0 Charred cereals Charred weed seeds Charred other crop No charred remains Waterlogged plants 100 metres Figure Distribution plot of processed Romano-British environmental samples with waterlogged plant remains, charred cereals, charred weed seeds, charred other crops and samples with no charred remains

254 RB.2 - Archaeobotanical remains were not that common. Charcoal volumes were mostly low (<5ml per flot) and only 45% of samples contained any cereal remains. Wild plant seeds are associated with the crops, almost never occurring in grain-free samples. 55 samples had burnt grain and seeds from scattered cereal processing waste whilst only seven samples contained large assemblages of well-preserved burnt grain and chaff, presumably intentionally discarded into features shortly after carbonisation took place. These seven samples came from seven distinct features from across settlement 2, suggesting that no specific area was reserved for cereal processing. Another rich sample should be added to this category: sample 548 from F.122 taken during the site s evaluation phase ( Evans & Newman 2010). The seven rich samples vary in composition from grain and arable weed seed dominated assemblages, to those with hundreds of spelt glume bases, some weed seeds and fewer grains. The burnt plants all represent post-storage cereal processing waste; however, their different compositions propose various specific stages and should be further explored with detailed analysis. Spelt was practically the only cereal type found. Hulled barley (Hordeum vulgare sensu lato) occurred sporadically, apparently in far fewer numbers than is usually seen on Romano- British sites in Cambridgeshire (this statement will be verified after detailed sorting of the samples). Other crops include peas and other pulses, only seen in six samples as one or two specimens. Exotic fruit and spices were not noted. The most common wild plant seeds were from medics or clovers (Medicago/Trifolium sp.), grasses, docks and stinking chamomile (Anthemis cotula). The latter is an indicator that spelt was grown on damp, clay-rich soils. Three samples were taken from the ring-ditch of C3. Two spelt grains, a single hulled barley grain and a large quantity of well-preserved charcoal were found in context [7310]. Four wild plant seeds were dispersed across the three samples, and represent typical arable weeds, such as knotgrass (Polygonum aviculare) and stinking chamomile. It would seem that loose cereal processing waste lay scattered beyond the inhabited areas. Three samples were taken from Way 1 and four from Way 3. Unsurprisingly, only two cereal grains and one wild seed were found, both in well F Crops are not expected to have become burnt or indeed to have been processed in any way along route-ways. Context [5488] of well F.1821 did not hold any waterlogged plant remains. Four of the waterlogged samples were analysed. Preservation in pit F.1114 and wells F.1020, F.2056 and F.1148 was excellent. The pit contained the fewest seeds, suggesting it was filled at a faster rate. The numerous nettle (Urtica urens and U.dioica), thistle (Carduus/Cirsium sp.), greater burnet-saxifrage (Pimpinella major), rushes (Juncus spp.) and henbane (Hyoscyamus niger) seeds show that the pit was dug in fertile, seasonally damp and disturbed grassland. The more water tolerant species may have grown within the pit. Willow (Salix sp.) and elder (Sambucus nigra) grew nearby, along with the occasional bramble. The three wells contained similar assemblages to each other and that of the pit, all indicative of open, disturbed fertile grassland. The evidence supports that gained from the insects in F.2056 which suggests herbivores may have grazed nearby, creating fertile disturbed grassland. Brambles and elder were rare, and no other shrubs/trees were noted. Seeds and grains found charred in the dry samples were also found in the wells, as both carbonised and waterlogged specimens (grains were not waterlogged), pointing to a little processing waste from agricultural activities. Cremations - 25 Romano-British (54 samples, 535 litres of soil) cremations were bulk-sampled and processed using an Ankara-type flotation machine. The flots were collected in 300µm aperture meshes and the remaining heavy residues washed over a 1mm mesh. The flots and heavy residues were dried indoors prior to analysis. J.Hutton sorted >2mm finds and cremated bones from the heavy residues by eye. Dry flots were separated through a stack of sieves; fractions were sorted and macro remains identified under a low power binocular microscope (6x 40x magnification) by the author. 100% of all flots were sorted. All environmental remains are listed in Tables

255 All archaeobotanical remains are charred. Charcoal was infrequent and quite heavily fragmented. Seeds and grains were rare, especially in the Romano-British cremations, but well-preserved, suggesting that their absence is real and not a factor of adverse physical and chemical conditions other than the displacement of finds through modern truncation. There is no evident difference between plant remains found within vessels and those from surrounding, often truncated, contexts. Modern, intrusive rootlets were present in all samples, indicating a low level of bioturbation that may have displaced plant macro-remains. All of the cremations, except for F.2606, were interned in vessels which indicates a degree of post-burning interference that may explain why charcoal was so sparse. Other plant remains may also be lacking for the same reasons. The most common seeds in the cremations were cereal grains: hulled barley (Hordeum vulgare sensu lato), hulled wheat (Triticum dicoccum/spelta) and a single find of bread wheat chaff (hexaploid T. aestivum sl.). Low occurrences of grain are common in Romano-British features, especially from densely occupied sites, and one should not automatically assume that cereals were included as food offerings. Wild plant seeds were also sparse and mostly representative of wet/damp environments: blinks (Montia fontana), great-fen sedge (Cladium mariscus) and other sedges (Carex spp.). No obvious differences were noted between cremations within and those outlying the cemeteries. Cremations F.823 and F.2663 contained small (<4mm across) lumps of burnt soil. As was mentioned above F.2606 did not contain a vessel. It did, however, contain more charcoal and more seeds than all the other RB cremations, and does in fact compare well to the Bronze Age cremations. F.2606 appears to represent a different style of cremation, unique within the other Romano-British ones. Romano-British cremations from North West Cambridge were buried in pots with few plant remains, including charcoal. Excavations however, found that they had suffered heavy physical disturbances so that unknown quantities of plant remains may be missing. Cereal grains were quite common (not in absolute quantities but in overall presence) and it remains possible that these were food offerings. On the other hand, loose charred grains are usually prolific across densely occupied Romano-British settlements. F.2606, with no vessels and an archaeobotanical assemblage more similar to those from the Bronze Age cremations, stands out as being atypical within the Romano-British cremations. Figure 3.68 Cereal (grains + glume bases) concentrations in samples from different excavation areas 358

Former Whitbread Training Centre Site, Abbey Street, Faversham, Kent Interim Archaeological Report Phase 1 November 2009

Former Whitbread Training Centre Site, Abbey Street, Faversham, Kent Interim Archaeological Report Phase 1 November 2009 Former Whitbread Training Centre Site, Abbey Street, Faversham, Kent Interim Archaeological Report Phase 1 November 2009 SWAT. Archaeology Swale and Thames Archaeological Survey Company School Farm Oast,

More information

North West Cambridge, University of Cambridge Archaeological Evaluation Fieldwork

North West Cambridge, University of Cambridge Archaeological Evaluation Fieldwork North West Cambridge, University of Cambridge Archaeological Evaluation Fieldwork Christopher Evans & Richard Newman With contributions by Katie Anderson, Grahame Appleby, Emma Beadsmoore, Lawrence Billington,

More information

39, Walnut Tree Lane, Sudbury (SUY 073) Planning Application No. B/04/02019/FUL Archaeological Monitoring Report No. 2005/112 OASIS ID no.

39, Walnut Tree Lane, Sudbury (SUY 073) Planning Application No. B/04/02019/FUL Archaeological Monitoring Report No. 2005/112 OASIS ID no. 39, Walnut Tree Lane, Sudbury (SUY 073) Planning Application No. B/04/02019/FUL Archaeological Monitoring Report No. 2005/112 OASIS ID no. 9273 Summary Sudbury, 39, Walnut Tree Lane, Sudbury (TL/869412;

More information

Chapter 2. Remains. Fig.17 Map of Krang Kor site

Chapter 2. Remains. Fig.17 Map of Krang Kor site Chapter 2. Remains Section 1. Overview of the Survey Area The survey began in January 2010 by exploring the site of the burial rootings based on information of the rooted burials that was brought to the

More information

Chapter 2: Archaeological Description

Chapter 2: Archaeological Description Chapter 2: Archaeological Description Phase 1 Late Neolithic, c 3000-2400 BC (Figs 6-9) Evidence of Neolithic activity was confined to pits dug across the southern half of the site (Fig. 6). Eighteen pits

More information

Test-Pit 3: 31 Park Street (SK )

Test-Pit 3: 31 Park Street (SK ) -Pit 3: 31 Park Street (SK 40732 03178) -Pit 3 was excavated in a flower bed in the rear garden of 31 Park Street, on the northern side of the street and west of an alleyway leading to St Peter s Church,

More information

7. Prehistoric features and an early medieval enclosure at Coonagh West, Co. Limerick Kate Taylor

7. Prehistoric features and an early medieval enclosure at Coonagh West, Co. Limerick Kate Taylor 7. Prehistoric features and an early medieval enclosure at Coonagh West, Co. Limerick Kate Taylor Illus. 1 Location of the site in Coonagh West, Co. Limerick (based on the Ordnance Survey Ireland map)

More information

Colchester Archaeological Trust Ltd. A Fieldwalking Survey at Birch, Colchester for ARC Southern Ltd

Colchester Archaeological Trust Ltd. A Fieldwalking Survey at Birch, Colchester for ARC Southern Ltd Colchester Archaeological Trust Ltd A Fieldwalking Survey at Birch, Colchester for ARC Southern Ltd November 1997 CONTENTS page Summary... 1 Background... 1 Methods... 1 Retrieval Policy... 2 Conditions...

More information

THE PRE-CONQUEST COFFINS FROM SWINEGATE AND 18 BACK SWINEGATE

THE PRE-CONQUEST COFFINS FROM SWINEGATE AND 18 BACK SWINEGATE THE PRE-CONQUEST COFFINS FROM 12 18 SWINEGATE AND 18 BACK SWINEGATE An Insight Report By J.M. McComish York Archaeological Trust for Excavation and Research (2015) Contents 1. INTRODUCTION... 3 2. THE

More information

Cambridge Archaeology Field Group. Fieldwalking on the Childerley Estate Cambridgeshire

Cambridge Archaeology Field Group. Fieldwalking on the Childerley Estate Cambridgeshire Cambridge Archaeology Field Group Fieldwalking on the Childerley Estate Cambridgeshire 2009 to 2014 Summary Fieldwalking on the Childerley estate of Martin Jenkins and Family has revealed, up to March

More information

Greater London GREATER LONDON 3/606 (E ) TQ

Greater London GREATER LONDON 3/606 (E ) TQ GREATER LONDON City of London 3/606 (E.01.6024) TQ 30358150 1 PLOUGH PLACE, CITY OF LONDON An Archaeological Watching Brief at 1 Plough Place, City of London, London EC4 Butler, J London : Pre-Construct

More information

SALVAGE EXCAVATIONS AT OLD DOWN FARM, EAST MEON

SALVAGE EXCAVATIONS AT OLD DOWN FARM, EAST MEON Proc. Hants. Field Club Archaeol. Soc. 36, 1980, 153-160. 153 SALVAGE EXCAVATIONS AT OLD DOWN FARM, EAST MEON By RICHARD WHINNEY AND GEORGE WALKER INTRODUCTION The site was discovered by chance in December

More information

An archaeological evaluation at 16 Seaview Road, Brightlingsea, Essex February 2004

An archaeological evaluation at 16 Seaview Road, Brightlingsea, Essex February 2004 An archaeological evaluation at 16 Seaview Road, Brightlingsea, Essex February 2004 report prepared by Kate Orr on behalf of Highfield Homes NGR: TM 086 174 (c) CAT project ref.: 04/2b ECC HAMP group site

More information

Cambridge Archaeology Field Group. Fieldwalking on the Childerley Estate, Cambridgeshire. Autumn 2014 to Spring Third interim report

Cambridge Archaeology Field Group. Fieldwalking on the Childerley Estate, Cambridgeshire. Autumn 2014 to Spring Third interim report Cambridge Archaeology Field Group Fieldwalking on the Childerley Estate, Cambridgeshire Autumn 2014 to Spring 2015 Third interim report Summary Field walking on the Childerley estate of Martin Jenkins

More information

3. The new face of Bronze Age pottery Jacinta Kiely and Bruce Sutton

3. The new face of Bronze Age pottery Jacinta Kiely and Bruce Sutton 3. The new face of Bronze Age pottery Jacinta Kiely and Bruce Sutton Illus. 1 Location map of Early Bronze Age site at Mitchelstown, Co. Cork (based on the Ordnance Survey Ireland map) A previously unknown

More information

ARCHAEOLOGICAL EVALUATION AT BRIGHTON POLYTECHNIC, NORTH FIELD SITE, VARLEY HALLS, COLDEAN LANE, BRIGHTON. by Ian Greig MA AIFA.

ARCHAEOLOGICAL EVALUATION AT BRIGHTON POLYTECHNIC, NORTH FIELD SITE, VARLEY HALLS, COLDEAN LANE, BRIGHTON. by Ian Greig MA AIFA. ARCHAEOLOGICAL EVALUATION AT BRIGHTON POLYTECHNIC, NORTH FIELD SITE, VARLEY HALLS, COLDEAN LANE, BRIGHTON by Ian Greig MA AIFA May 1992 South Eastern Archaeological Services Field Archaeology Unit White

More information

Peace Hall, Sydney Town Hall Results of Archaeological Program (Interim Report)

Peace Hall, Sydney Town Hall Results of Archaeological Program (Interim Report) Results of Archaeological Program (Interim Report) Background The proposed excavation of a services basement in the western half of the Peace Hall led to the archaeological investigation of the space in

More information

Archaeological sites and find spots in the parish of Burghclere - SMR no. OS Grid Ref. Site Name Classification Period

Archaeological sites and find spots in the parish of Burghclere - SMR no. OS Grid Ref. Site Name Classification Period Archaeological sites and find spots in the parish of Burghclere - SMR no. OS Grid Ref. Site Name Classification Period SU45NE 1A SU46880 59200 Ridgemoor Farm Inhumation Burial At Ridgemoor Farm, on the

More information

An archaeological watching brief and recording at Brightlingsea Quarry, Moverons Lane, Brightlingsea, Essex October 2003

An archaeological watching brief and recording at Brightlingsea Quarry, Moverons Lane, Brightlingsea, Essex October 2003 An archaeological watching brief and recording at Brightlingsea Quarry, Moverons Lane, Brightlingsea, Essex commissioned by Mineral Services Ltd on behalf of Alresford Sand & Ballast Co Ltd report prepared

More information

ROMAN AND MEDIEVAL ACTIVITY IN THE UPPER WALBROOK VALLEY: EXCAVATIONS AT MOORGATE, CITY OF LONDON, EC2, 1997

ROMAN AND MEDIEVAL ACTIVITY IN THE UPPER WALBROOK VALLEY: EXCAVATIONS AT MOORGATE, CITY OF LONDON, EC2, 1997 ROMAN AND MEDIEVAL ACTIVITY IN THE UPPER WALBROOK VALLEY: EXCAVATIONS AT 12 18 MOORGATE, CITY OF LONDON, EC2, 1997 Graham Bruce, Dominic Perring, Tim Stevens and Melissa Melikian SUMMARY In January and

More information

New Composting Centre, Ashgrove Farm, Ardley, Oxfordshire

New Composting Centre, Ashgrove Farm, Ardley, Oxfordshire New Composting Centre, Ashgrove Farm, Ardley, Oxfordshire An Archaeological Watching Brief For Agrivert Limited by Andrew Weale Thames Valley Archaeological Services Ltd Site Code AFA 09/20 August 2009

More information

Changing People Changing Landscapes: excavations at The Carrick, Midross, Loch Lomond Gavin MacGregor, University of Glasgow

Changing People Changing Landscapes: excavations at The Carrick, Midross, Loch Lomond Gavin MacGregor, University of Glasgow Changing People Changing Landscapes: excavations at The Carrick, Midross, Loch Lomond Gavin MacGregor, University of Glasgow Located approximately 40 kilometres to the south-west of Oban, as the crow flies

More information

Undley Hall, Lakenheath LKH 307

Undley Hall, Lakenheath LKH 307 ARCHAEOLOGICAL EVALUATION AND MONITORING REPORT SCCAS REPORT No. 2010/005 Undley Hall, Lakenheath LKH 307 E. Muldowney SCCAS January 2010 www.suffolkcc.gov.uk/e-and-t/archaeology Lucy Robinson, County

More information

Bronze-Age and Romano-British Sites South-East of Tewkesbury: evaluations and excavations

Bronze-Age and Romano-British Sites South-East of Tewkesbury: evaluations and excavations From the Transactions of the Bristol and Gloucestershire Archaeological Society Bronze-Age and Romano-British Sites South-East of Tewkesbury: evaluations and excavations 1991-7 by G. Walker, A. Thomas

More information

Phase 2 Urban consolidation AD

Phase 2 Urban consolidation AD Phase 2 Urban consolidation AD 1250-1350 The second recognised phase of activity at Rådhuspladsen corresponded approximately to the High medieval period (c. AD 1250 1350), and saw consolidation of the

More information

17 Phase 5. High and Late medieval features and activities AD

17 Phase 5. High and Late medieval features and activities AD 17 Phase 5. High and Late medieval features and activities 1200 1550 AD 17.1 Results This time phase is based on all findings that can be placed in the High and Late medieval period 1200 1550 AD based

More information

Bristol & Gloucestershire Archaeological Society

Bristol & Gloucestershire Archaeological Society Chris Hayden, Rob Early, Edward Biddulph, Paul Booth, Anne Dodd, Alex Smith, Granville Laws and Ken Welsh, Horcott Quarry, Fairford and Arkell's Land, Kempsford: Prehistoric, Roman and Anglo-Saxon settlement

More information

Tell Shiyukh Tahtani (North Syria)

Tell Shiyukh Tahtani (North Syria) Tell Shiyukh Tahtani (North Syria) Report of the 2010 excavation season conducted by the University of Palermo Euphrates Expedition by Gioacchino Falsone and Paola Sconzo In the summer 2010 the University

More information

SERIATION: Ordering Archaeological Evidence by Stylistic Differences

SERIATION: Ordering Archaeological Evidence by Stylistic Differences SERIATION: Ordering Archaeological Evidence by Stylistic Differences Seriation During the early stages of archaeological research in a given region, archaeologists often encounter objects or assemblages

More information

Suburban life in Roman Durnovaria

Suburban life in Roman Durnovaria Suburban life in Roman Durnovaria Additional specialist report Finds Ceramic building material By Kayt Brown Ceramic building material (CBM) Kayt Brown A total of 16420 fragments (926743g) of Roman ceramic

More information

Grim s Ditch, Starveall Farm, Wootton, Woodstock, Oxfordshire

Grim s Ditch, Starveall Farm, Wootton, Woodstock, Oxfordshire Grim s Ditch, Starveall Farm, Wootton, Woodstock, Oxfordshire An Archaeological Recording Action For Empire Homes by Steve Ford Thames Valley Archaeological Services Ltd Site Code SFW06/118 November 2006

More information

A Sense of Place Tor Enclosures

A Sense of Place Tor Enclosures A Sense of Place Tor Enclosures Tor enclosures were built around six thousand years ago (4000 BC) in the early part of the Neolithic period. They are large enclosures defined by stony banks sited on hilltops

More information

An archaeological evaluation at the Lexden Wood Golf Club (Westhouse Farm), Lexden, Colchester, Essex

An archaeological evaluation at the Lexden Wood Golf Club (Westhouse Farm), Lexden, Colchester, Essex An archaeological evaluation at the Lexden Wood Golf Club (Westhouse Farm), Lexden, Colchester, Essex January 2000 Archive report on behalf of Lexden Wood Golf Club Colchester Archaeological Trust 12 Lexden

More information

Fieldwalking at Cottam 1994 (COT94F)

Fieldwalking at Cottam 1994 (COT94F) Fieldwalking at Cottam 1994 (COT94F) Tony Austin & Elizabeth Jelley (19 Jan 29) 1. Introduction During the winter of 1994 students from the Department of Archaeology at the University of York undertook

More information

THE RAVENSTONE BEAKER

THE RAVENSTONE BEAKER DISCOVERY THE RAVENSTONE BEAKER K. J. FIELD The discovery of the Ravenstone Beaker (Plate Xa Fig. 1) was made by members of the Wolverton and District Archaeological Society engaged on a routine field

More information

FURTHER MIDDLE SAXON EVIDENCE AT COOK STREET, SOUTHAMPTON (SOU 567)

FURTHER MIDDLE SAXON EVIDENCE AT COOK STREET, SOUTHAMPTON (SOU 567) Roc. Hampshire Field Club Archaeol. Soc 52,1997, 77-87 (Hampshire Studies 1997) FURTHER MIDDLE SAXON EVIDENCE AT COOK STREET, SOUTHAMPTON (SOU 567) By M F GARNER andj VINCENT with a contribution byjacqueline

More information

St Germains, Tranent, East Lothian: the excavation of Early Bronze Age remains and Iron Age enclosed and unenclosed settlements

St Germains, Tranent, East Lothian: the excavation of Early Bronze Age remains and Iron Age enclosed and unenclosed settlements Proc Soc Antiq Scot, 128 (1998), 203-254 St Germains, Tranent, East Lothian: the excavation of Early Bronze Age remains and Iron Age enclosed and unenclosed settlements Derek Alexander* & Trevor Watkinsf

More information

A visit to the Wor Barrow 21 st November 2015

A visit to the Wor Barrow 21 st November 2015 A visit to the Wor Barrow 21 st November 2015 Following our exploration of Winkelbury a few weeks previously, we fast forwarded 12 years in Pitt Rivers remarkable series of excavations and followed him

More information

TIPPERARY HISTORICAL JOURNAL 1994

TIPPERARY HISTORICAL JOURNAL 1994 TPPERARY HSTORCAL JOURNAL 1994 County Tipperary Historical Society www.tipperarylibraries.ie/ths society@tipperarylibraries. ie SSN 0791-0655 Excavations at Cormac's Chapel, Cashel, 1992 and 1993: a preliminary

More information

Control ID: Years of experience: Tools used to excavate the grave: Did the participant sieve the fill: Weather conditions: Time taken: Observations:

Control ID: Years of experience: Tools used to excavate the grave: Did the participant sieve the fill: Weather conditions: Time taken: Observations: Control ID: Control 001 Years of experience: No archaeological experience Tools used to excavate the grave: Trowel, hand shovel and shovel Did the participant sieve the fill: Yes Weather conditions: Flurries

More information

A R C H A E O L O G Y. N o. 8 NORTH WEST CAMBRIDGE. The Traveller s Rest Sub-site. Assessment Report. Christopher Evans

A R C H A E O L O G Y. N o. 8 NORTH WEST CAMBRIDGE. The Traveller s Rest Sub-site. Assessment Report. Christopher Evans NORTH WEST CAMBRIDGE A R C H A E O L O G Y The Traveller s Rest Sub-site Assessment Report Christopher Evans N o. 8 NORTH WEST CAMBRIDGE ARCHAEOLOGY University of Cambridge 2013 Excavations - The Traveller

More information

Excavation of Iron-Age and Roman Occupation at Coln Gravel, Thornhill Farm,Fairford, Gloucestershire, 2003 and 2004.

Excavation of Iron-Age and Roman Occupation at Coln Gravel, Thornhill Farm,Fairford, Gloucestershire, 2003 and 2004. From the Transactions of the Bristol and Gloucestershire Archaeological Society Excavation of Iron-Age and Roman Occupation at Coln Gravel, Thornhill Farm,Fairford, Gloucestershire, 2003 and 2004. by Dan

More information

An archaeological evaluation in the playground of Colchester Royal Grammar School, Lexden Road, Colchester, Essex

An archaeological evaluation in the playground of Colchester Royal Grammar School, Lexden Road, Colchester, Essex An archaeological evaluation in the playground of Colchester Royal Grammar School, Lexden Road, Colchester, Essex February 2002 on behalf of Roff Marsh Partnership CAT project code: 02/2c Colchester Museum

More information

1 The East Oxford Archaeology and History Project

1 The East Oxford Archaeology and History Project 1 The East Oxford Archaeology and History Project EXOP TEST PIT 72 Location: Bartlemas Chapel, Cowley Date of excavation: 6-8 November 2013. Area of excavation: 0.8m x 1.2m, at the eastern end of the chapel.

More information

Lanton Lithic Assessment

Lanton Lithic Assessment Lanton Lithic Assessment Dr Clive Waddington ARS Ltd The section headings in the following assessment report refer to those in the Management of Archaeological Projects (HBMC 1991), Appendix 4. 1. FACTUAL

More information

This is a repository copy of Anglo-Saxon settlements and archaeological visibility in the Yorkshire Wolds.

This is a repository copy of Anglo-Saxon settlements and archaeological visibility in the Yorkshire Wolds. This is a repository copy of Anglo-Saxon settlements and archaeological visibility in the Yorkshire Wolds. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/1172/ Book Section:

More information

2 Saxon Way, Old Windsor, Berkshire

2 Saxon Way, Old Windsor, Berkshire 2 Saxon Way, Old Windsor, Berkshire An Archaeological Watching Brief For Mrs J. McGillicuddy by Pamela Jenkins Thames Valley Archaeological Services Ltd Site Code SWO 05/67 August 2005 Summary Site name:

More information

Excavations at Shikarpur, Gujarat

Excavations at Shikarpur, Gujarat Excavations at Shikarpur, Gujarat 2008-2009 The Department of Archaeology and Ancient History, the M. S. University of Baroda continued excavations at Shikarpur in the second field season in 2008-09. In

More information

Greater London Region GREATER LONDON 3/567 (E.01.K099) TQ BERMONDSEY STREET AND GIFCO BUILDING AND CAR PARK

Greater London Region GREATER LONDON 3/567 (E.01.K099) TQ BERMONDSEY STREET AND GIFCO BUILDING AND CAR PARK GREATER LONDON 3/567 (E.01.K099) TQ 33307955 156-170 BERMONDSEY STREET AND GIFCO BUILDING AND CAR PARK Assessment of an Archaeological Excavation at 156-170 Bermondsey Street and GIFCO Building and Car

More information

Bronze Age 2, BC

Bronze Age 2, BC Bronze Age 2,000-600 BC There may be continuity with the Neolithic period in the Early Bronze Age, with the harbour being used for seasonal grazing, and perhaps butchering and hide preparation. In the

More information

Human remains from Estark, Iran, 2017

Human remains from Estark, Iran, 2017 Bioarchaeology of the Near East, 11:84 89 (2017) Short fieldwork report Human remains from Estark, Iran, 2017 Arkadiusz Sołtysiak *1, Javad Hosseinzadeh 2, Mohsen Javeri 2, Agata Bebel 1 1 Department of

More information

Forteviot, Palisaded Enclosure Excavations Data Structure Report. By Aoife Gould

Forteviot, Palisaded Enclosure Excavations Data Structure Report. By Aoife Gould Forteviot, Palisaded Enclosure Excavations 2010 Data Structure Report By Aoife Gould Table of Contents Summary 2 Introduction 2 Location 2 Archaeological Background 4 Aims 4 Methodology 5 Results 8 Subsoil

More information

Church of St Peter and St Paul, Great Missenden, Buckinghamshire

Church of St Peter and St Paul, Great Missenden, Buckinghamshire Church of St Peter and St Paul, Great Missenden, Buckinghamshire An Archaeological Watching Brief for the Parish of Great Missenden by Andrew Taylor Thames Valley Archaeological Services Ltd Site Code

More information

BALNUARAN. of C LAVA. a prehistoric cemetery. A Visitors Guide to

BALNUARAN. of C LAVA. a prehistoric cemetery. A Visitors Guide to A Visitors Guide to BALNUARAN of C LAVA a prehistoric cemetery Milton of Clava Chapel (?) Cairn River Nairn Balnuaran of Clava is the site of an exceptionally wellpreserved group of prehistoric burial

More information

Cetamura Results

Cetamura Results Cetamura 2000 2006 Results A major project during the years 2000-2006 was the excavation to bedrock of two large and deep units located on an escarpment between Zone I and Zone II (fig. 1 and fig. 2);

More information

1. Presumed Location of French Soundings Looking NW from the banks of the river.

1. Presumed Location of French Soundings Looking NW from the banks of the river. SG02? SGS SG01? SG4 1. Presumed Location of French Soundings Looking NW from the banks of the river. The presumed location of SG02 corresponds to a hump known locally as the Sheikh's tomb. Note also (1)

More information

THE EXCAVATION OF A BURNT MOUND AT HARBRIDGE, HAMPSHIRE

THE EXCAVATION OF A BURNT MOUND AT HARBRIDGE, HAMPSHIRE Proc Hampshire Field ClubArchaeolSoc5i, 1999,172-179 (Hampshire Studies 1999) THE EXCAVATION OF A BURNT MOUND AT HARBRIDGE, HAMPSHIRE by S J SHENNAN ABSTRACT A burnt mound of Late Brome Age date, as indicated

More information

FEATURE DESCRIPTIONS: PART 1. SAN AGUSTÍN MISSION LOCUS, THE CLEARWATER SITE, AZ BB:13:6 (ASM)

FEATURE DESCRIPTIONS: PART 1. SAN AGUSTÍN MISSION LOCUS, THE CLEARWATER SITE, AZ BB:13:6 (ASM) CHAPTER 4 FEATURE DESCRIPTIONS: PART 1. SAN AGUSTÍN MISSION LOCUS, THE CLEARWATER SITE, AZ BB:13:6 (ASM) Thomas Klimas, Caramia Williams, and J. Homer Thiel Desert Archaeology, Inc. Archaeological work

More information

Forteviot, Perthshire: Excavations at the Entrance Avenue of the Neolithic Palisaded Enclosure Interim Report and Data Structure Report

Forteviot, Perthshire: Excavations at the Entrance Avenue of the Neolithic Palisaded Enclosure Interim Report and Data Structure Report Forteviot, Perthshire: Excavations at the Entrance Avenue of the Neolithic Palisaded Enclosure 2007 Interim Report and Data Structure Report by Gordon Noble and Kenneth Brophy Table of Contents Summary

More information

LAND WEST OF ELM GROVE, EBRINGTON, GLOUCESTERSHIRE. NGR: SP (centred) ARCHAEOLOGICAL EVALUATION

LAND WEST OF ELM GROVE, EBRINGTON, GLOUCESTERSHIRE. NGR: SP (centred) ARCHAEOLOGICAL EVALUATION LAND WEST OF ELM GROVE, EBRINGTON, GLOUCESTERSHIRE. NGR: SP 1892 4012 (centred) ARCHAEOLOGICAL EVALUATION Report No. 640 May 2009 1q LAND WEST OF ELM GROVE, EBRINGTON, GLOUCESTERSHIRE. NGR: SP 1892 4012

More information

Monitoring Report No Sacred Heart Church Aghamore Boho Co. Fermanagh AE/10/116E. Brian Sloan L/2009/1262/F

Monitoring Report No Sacred Heart Church Aghamore Boho Co. Fermanagh AE/10/116E. Brian Sloan L/2009/1262/F Monitoring Report No. 202 Sacred Heart Church Aghamore Boho Co. Fermanagh AE/10/116E Brian Sloan L/2009/1262/F Site Specific Information Site Address: Sacred Heart Church, Aghamore, Boho, Co. Fermanagh

More information

MARSTON MICHAEL FARLEY

MARSTON MICHAEL FARLEY MARSTON MICHAEL FARLEY On 9 March agricultural contractors, laying field drains for Bucks County Council Land Agent's Department, cut through a limestone structure at SP 75852301 in an area otherwise consistently

More information

Evidence for the use of bronze mining tools in the Bronze Age copper mines on the Great Orme, Llandudno

Evidence for the use of bronze mining tools in the Bronze Age copper mines on the Great Orme, Llandudno Evidence for the use of bronze mining tools in the Bronze Age copper mines on the Great Orme, Llandudno Background The possible use of bronze mining tools has been widely debated since the discovery of

More information

Grange Farm, Widmer End, Hughenden, Buckinghamshire

Grange Farm, Widmer End, Hughenden, Buckinghamshire Grange Farm, Widmer End, Hughenden, Buckinghamshire An Archaeological Evaluation for British Flora by Andy Taylor Thames Valley Archaeological Services Ltd Site Code GFH 05/63 July 2005 Summary Site name:

More information

Excavation Report. Medieval Occupation at Challis Green Barrington Cambridgeshire. Excavation Report. Client: Hills Partnership Limited.

Excavation Report. Medieval Occupation at Challis Green Barrington Cambridgeshire. Excavation Report. Client: Hills Partnership Limited. Medieval Occupation at Challis Green Barrington Cambridgeshire Excavation Report Excavation Report April 2012 Client: Hills Partnership Limited OA East Report No: 1269 OASIS No: oxfordar3-102493 NGR: TL

More information

Forteviot, Perthshire 2008: Excavations of a henge monument and timber circle. Data Structure and Interim Report. by Gordon Noble and Kenneth Brophy

Forteviot, Perthshire 2008: Excavations of a henge monument and timber circle. Data Structure and Interim Report. by Gordon Noble and Kenneth Brophy Forteviot, Perthshire 2008: Excavations of a henge monument and timber circle Data Structure and Interim Report by Gordon Noble and Kenneth Brophy Summary This interim report will describe the provisional

More information

Moray Archaeology For All Project

Moray Archaeology For All Project School children learning how to identify finds. (Above) A flint tool found at Clarkly Hill. Copyright: Leanne Demay Moray Archaeology For All Project ational Museums Scotland have been excavating in Moray

More information

Archaeological trial-trenching evaluation at Chappel Farm, Little Totham, Essex. April 2013

Archaeological trial-trenching evaluation at Chappel Farm, Little Totham, Essex. April 2013 Archaeological trial-trenching evaluation at Chappel Farm, Little Totham, Essex April 2013 report prepared by Ben Holloway commissioned by Tim Harbord Associates on behalf of Mr Tom Howie Planning reference:

More information

Fort Arbeia and the Roman Empire in Britain 2012 FIELD REPORT

Fort Arbeia and the Roman Empire in Britain 2012 FIELD REPORT Fort Arbeia and the Roman Empire in Britain 2012 FIELD REPORT Background Information Lead PI: Paul Bidwell Report completed by: Paul Bidwell Period Covered by this report: 17 June to 25 August 2012 Date

More information

Intermediate School Gym RAF Lakenheath, Eriswell ERL 214

Intermediate School Gym RAF Lakenheath, Eriswell ERL 214 Intermediate School Gym RAF Lakenheath, Eriswell ERL 214 Archaeological Excavation Report SCCAS Report No. 2012/017 Client: Defence Infrastructure Organisation Author: Andrew Vaughan Beverton 06/2012 Intermediate

More information

An archaeological watching brief on one section of an Anglian Water main Spring Lane, Lexden, Colchester

An archaeological watching brief on one section of an Anglian Water main Spring Lane, Lexden, Colchester An archaeological watching brief on one section of an Anglian Water main Spring Lane, Lexden, Colchester April-September 2001 on behalf of Breheny Contractors CAT project ref.: 01/4D Colchester Museum

More information

AREA C. HENRY 0. THOMPSON American Center of Oriental Research Amman, Jordan

AREA C. HENRY 0. THOMPSON American Center of Oriental Research Amman, Jordan AREA C HENRY 0. THOMPSON American Center of Oriental Research Amman, Jordan Of the 1971 work previously reported,' Squares 4,5, and 6 were not excavated in 1973, but work in Squares 1, 2, and 3 was continued.

More information

DEMARCATION OF THE STONE AGES.

DEMARCATION OF THE STONE AGES. 20 HAMPSHIRE FLINTS. DEMARCATION OF THE STONE AGES. BY W, DALE, F.S.A., F.G.S. (Read before the Anthropological Section of -the British Association for the advancement of Science, at Birmingham, September

More information

Foreign Whaling in Iceland Archaeological Excavations at Strákatangi in Hveravík, Kaldrananeshreppi 2007 Data Structure Report

Foreign Whaling in Iceland Archaeological Excavations at Strákatangi in Hveravík, Kaldrananeshreppi 2007 Data Structure Report Foreign Whaling in Iceland Archaeological Excavations at Strákatangi in Hveravík, Kaldrananeshreppi 2007 Data Structure Report Caroline Paulsen, Magnús Rafnsson and Ragnar Edvardsson February 2008 NV nr.

More information

Land at Ullswater Road, Campsea Ashe, Suffolk CAA 032. Archaeological Post-excavation Assessment DRAFT. Client: Flagship Housing Group

Land at Ullswater Road, Campsea Ashe, Suffolk CAA 032. Archaeological Post-excavation Assessment DRAFT. Client: Flagship Housing Group DRAFT Land at Ullswater Road, Campsea Ashe, Suffolk CAA 032 Archaeological Post-excavation Assessment SCCAS Report No. 2013/131 Client: Flagship Housing Group Author: Mark Sommers March 2014 Suffolk County

More information

Archaeological. Monitoring & Recording Report. Fulbourn Primary School, Cambridgeshire. Archaeological Monitoring & Recording Report.

Archaeological. Monitoring & Recording Report. Fulbourn Primary School, Cambridgeshire. Archaeological Monitoring & Recording Report. Fulbourn Primary School, Cambridgeshire Archaeological Monitoring & Recording Report October 2014 Client: Cambridgeshire County Council OA East Report No: 1689 OASIS No: oxfordar3-192890 NGR: TL 5190 5613

More information

Excavations of Late Iron Age and Roman features and a Roman road north of Gosbecks Archaeological Park, Colchester, Essex

Excavations of Late Iron Age and Roman features and a Roman road north of Gosbecks Archaeological Park, Colchester, Essex Excavations of Late Iron Age and Roman features and a Roman road north of Gosbecks Archaeological Park, Colchester, Essex 1995-1996 report by Stephen Benfield with contributions from Justine Bailey, Peter

More information

Bangor University. The Meillionydd Project: Characterising the double ringwork enclosures in Gwynedd Preliminary Excavation Report

Bangor University. The Meillionydd Project: Characterising the double ringwork enclosures in Gwynedd Preliminary Excavation Report Bangor University The Meillionydd Project: Characterising the double ringwork enclosures in Gwynedd Preliminary Excavation Report Kate Waddington and Raimund Karl Bangor, August 2010 Contents Acknowledements

More information

A Fieldwalking Project At Sompting. West Sussex

A Fieldwalking Project At Sompting. West Sussex by John Funnell Introduction A Fieldwalking Project At Sompting. West Sussex During March -and April 1995 the Brighton and Hove Archaeological Society conducted fie1dwa1king in a field at Sompting West

More information

THE UNFOLDING ARCHAEOLOGY OF CHELTENHAM

THE UNFOLDING ARCHAEOLOGY OF CHELTENHAM THE UNFOLDING ARCHAEOLOGY OF CHELTENHAM The archaeology collection of Cheltenham Art Gallery & Museum contains a rich quantity of material relating to the prehistoric and Roman occupation of the North

More information

Azerbaijan National Academy of Sciences Institute of Archaeology and Ethnography. Safar Ashurov

Azerbaijan National Academy of Sciences Institute of Archaeology and Ethnography. Safar Ashurov Azerbaijan National Academy of Sciences Institute of Archaeology and Ethnography Safar Ashurov Zayamchay Report On Excavations of a Catacomb Burial At Kilometre Point 355 of Baku-Tbilisi-Ceyhan and South

More information

KNAP OF HOWAR HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE. Property in Care (PIC) ID: PIC301 Designations:

KNAP OF HOWAR HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE. Property in Care (PIC) ID: PIC301 Designations: Property in Care (PIC) ID: PIC301 Designations: Scheduled Monument (SM90195) Taken into State care: 1954 (Guardianship) Last reviewed: 2004 HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE KNAP

More information

Prehistoric and Romano-British Activity and Saxon Settlement at Hoo Road, Wainscott, Kent

Prehistoric and Romano-British Activity and Saxon Settlement at Hoo Road, Wainscott, Kent Prehistoric and Romano-British Activity and Saxon Settlement at Hoo Road, Wainscott, Kent This report has been downloaded from www.kentarchaeology.org.uk the website of the Kent Archaeological Society

More information

Rådhuspladsen, KBM 3827

Rådhuspladsen, KBM 3827 KØBENHAVNS MUSEUM / MUSEUM OF COPENHAGEN Rådhuspladsen, KBM 3827 Cultural Historical Report, Metro Cityring Excavation Ed Lyne and Hanna Dahlström Foreword As a result of the extensive archaeological

More information

Life and Death on a Romano-British estate: Turnershall Farm in Hertfordshire

Life and Death on a Romano-British estate: Turnershall Farm in Hertfordshire Introduction Life and Death on a Romano-British estate: Turnershall Farm in Hertfordshire In 2002 metal detectorists discovered two of the most significant burials to come from Roman Britain. The objects

More information

CHAPTER 14. Conclusions. Nicky Milner, Barry Taylor and Chantal Conneller

CHAPTER 14. Conclusions. Nicky Milner, Barry Taylor and Chantal Conneller PA RT 6 Conclusions In conclusion it is only fitting to emphasise that, useful though the investigations at Star Carr have been in helping to fill a gap in the prehistory of north-western Europe, much

More information

ARCHAEOLOGICAL MONITORING REPORT

ARCHAEOLOGICAL MONITORING REPORT ARCHAEOLOGICAL MONITORING REPORT SCCAS REPORT No. 2009/324 Thorington Hall, Stoke by Nayland SBN 087 HER Information Date of Fieldwork: November 2009 - January 2010 Grid Reference: TM 0131 3546 Funding

More information

TAI COCHION AND TREFARTHEN ROMAN SETTLEMENT

TAI COCHION AND TREFARTHEN ROMAN SETTLEMENT 1037 TAI COCHION AND TREFARTHEN ROMAN SETTLEMENT Interim Excavation Report Ymddiriedolaeth Archaeolegol Gwynedd Gwynedd Archaeological Trust TAI COCHION AND TREFARTHEN ROMAN SETTLEMENT Interim Excavation

More information

ARCHAEOLOGICAL EXCAVATION REPORT: THE PADDOCK, HIGH DIKE, NAVENBY, LINCOLNSHIRE

ARCHAEOLOGICAL EXCAVATION REPORT: THE PADDOCK, HIGH DIKE, NAVENBY, LINCOLNSHIRE ARCHAEOLOGICAL EXCAVATION REPORT: THE PADDOCK, HIGH DIKE, NAVENBY, LINCOLNSHIRE NGR: 499398, 357274 AAL Site Code: NAPA 13 OASIS Reference Number: allenarc1-205997 Report prepared for Navenby Archaeology

More information

ST PATRICK S CHAPEL, ST DAVIDS PEMBROKESHIRE 2015

ST PATRICK S CHAPEL, ST DAVIDS PEMBROKESHIRE 2015 ST PATRICK S CHAPEL, ST DAVIDS PEMBROKESHIRE 2015 REPORT FOR THE NINEVEH CHARITABLE TRUST THE UNIVERSITY OF SHEFFIELD AND DYFED ARCHAEOLOGICAL TRUST Introduction ST PATRICK S CHAPEL, ST DAVIDS, PEMBROKESHIRE,

More information

STONES OF STENNESS HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE

STONES OF STENNESS HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE Property in Care (PIC) ID: PIC321 Designations: Scheduled Monument (SM90285); Taken into State care: 1906 (Guardianship) Last reviewed: 2003 HISTORIC ENVIRONMENT SCOTLAND STATEMENT OF SIGNIFICANCE STONES

More information

By Lisa Brown. Trench 1. Residual pottery. 4.1 The later prehistoric pottery

By Lisa Brown. Trench 1. Residual pottery. 4.1 The later prehistoric pottery 4.1 The later prehistoric pottery By Lisa Brown 4.1.9 Introduction What follows are detailed trench by trench descriptions of the pottery, complementing the discussions in the Alfred s Castle monograph.

More information

Keflavík on Hegranes: Cemetery Excavation Interim Report 2016

Keflavík on Hegranes: Cemetery Excavation Interim Report 2016 SKAGAFJÖRÐUR CHURCH AND SETTLEMENT SURVEY Keflavík on Hegranes: Cemetery Excavation Interim Report 2016 Guðný Zoëga Douglas Bolender 2016/172 Picture on front page Drone photo of the Keflavík cemetery.

More information

Monitoring Report No. 99

Monitoring Report No. 99 Monitoring Report No. 99 Enniskillen Castle Co. Fermanagh AE/06/23 Cormac McSparron Site Specific Information Site Name: Townland: Enniskillen Castle Enniskillen SMR No: FER 211:039 Grid Ref: County: Excavation

More information

Part 10: Chapter 17 Pleated Buttoning

Part 10: Chapter 17 Pleated Buttoning Part 10: Chapter 17 Pleated Buttoning OUR last chapter covered the upholstering of one of the commonest forms of chair frames. The same chair may be upholstered with deeper buttoning, but instead of indenting

More information

CUMBRIA 2/635 (C ) NY

CUMBRIA 2/635 (C ) NY CUMBRIA Allerdale 2/635 (C.16.6017) NY 11803070 BRIDGE STREET, COCKERMOUTH Report on an Archaeological Evaluation at Bridge Street, Cockermouth, Cumbria (Report No. 5/00) Grahame, R Carlisle : Carlisle

More information

Rochester Road Soak-away

Rochester Road Soak-away Rochester Road Soak-away RAF Lakenheath, Suffolk Client: Defence Infrastructure Organisation Date: October 2015 ERL 236 Archaeological Excavation Report v0.3 SACIC Report No. 2015/005 Author: Rob Brooks

More information

The Living and the Dead

The Living and the Dead The Living and the Dead Round Barrows and cairns The transition from the late Neolithic to the early Bronze Age is traditionally associated with an influx of immigrants to the British Isles from continental

More information

An archaeological evaluation by trial-trenching at Playgolf, Bakers Lane, Westhouse Farm, Colchester, Essex

An archaeological evaluation by trial-trenching at Playgolf, Bakers Lane, Westhouse Farm, Colchester, Essex An archaeological evaluation by trial-trenching at Playgolf, Bakers Lane, Westhouse Farm, Colchester, Essex commissioned by Mr Stephen Belchem on behalf of ADP Ltd. report prepared by Chris Lister Planning

More information

An archaeological watching brief at Sheepen, Colchester, Essex November-December 2003

An archaeological watching brief at Sheepen, Colchester, Essex November-December 2003 An archaeological watching brief at Sheepen, Colchester, Essex November-December 2003 report prepared by Ben Holloway on behalf of Colchester Borough Council CAT project ref.: 03/11c Colchester Museums

More information