Nea Farm, Phase GP5, Somerley, Ringwood, Hampshire

Nea Farm, Phase GP5, Somerley, Ringwood, Hampshire An Archaeological Excavation for Tarmac Southern Ltd by Andy Taylor Thames Valley Archaeological Services Ltd Site Code SOM07/114

List of Figures Figure 1 Figure 2 Figure 3 Location of site in Southern England and Ringwood Location of site on the Somerley Estate and previous phases of work Plan of Excavation Areas showing all features Figure 4 Sections (1) Figure 5 Sections (2) Figure 6 Selected pottery (see text for details)

Nea Farm, Phase GP5, Somerley, Ringwood, Hampshire An Archaeological Excavation by Andy Taylor with contributions by Steve Ford, Malcolm Lyne, Rosalind McKenna and Frances Raymond Report 07/114 Summary Excavation prior to gravel extraction revealed a small number of archaeological deposits across the stripped areas mostly consisting of linear features representing field boundaries. These are tentatively dated to the Roman and/or medieval periods. Several pits and postholes were also observed, the majority of which were undated but some were medieval. Of particular significance due to their rarity was the discovery of three prehistoric pits. One of these was of Early Neolithic date with a radiocarbon determination of 3517 3393 cal BC (KIA39673), a second with Beaker pottery of Early Bronze Age date produced a radiocarbon determination of 2134 2078 cal BC (KIA39674),and a third was of Middle Bronze Age or Iron Age date. Introduction An archaeological excavation was carried out by Thames Valley Archaeological Services at Nea Farm, Somerley, Ringwood, Hampshire (SU 1290 0820) (Fig. 1). The work was commissioned by Mr Andy Cadell of Tarmac Southern Limited, Stancombe Quarry, Stancombe Lane, Flax Bourton, Bristol, BS48 3QD. Planning consent was obtained from Hampshire County Council to extract gravel from the area (a golf course prior to this work) as part of an ongoing extraction programme. The excavation was carried out in two phases to a specification approved by Mr Stephen Appleby, then Senior Archaeologist with Hampshire County Council, who also monitored the works along with Ms Hannah Fluck, the current incumbent, between 11th September and 5th November 2007 and between the 25th September and 20th October 2008. The site code is SOM07/114. This is one of a long series of archaeological investigations on this quarry site (see below). The earlier archaeological projects were numbered phases 1 6. More recent work is referred to by new quarry extraction references (GP3 and GP4 for the immediately previous phases, and GP5 for this phase). The archive is currently held by Thames Valley Archaeological Services, 47-49 De Beauvoir Road, Reading, RG1 5NR and will deposited in due course with Hampshire County Museum Service. 1

Topography and geology Nea Farm, Somerley lies to the north-west of Ringwood on the eastern margins of Ringwood Forest (Fig. 1). The underlying geology consists of plateau gravel capped with brickearth (BGS 1990), which was observed across the site. Approximately 1km to the east lies the River Avon. The site lies at a height of between 47m and 48m AOD. The steep slope of the main valley side lies 500m to the east and the small valley of a tributary stream lies 400m to the north (Fig. 2). The current phase of the extraction site comprised an irregular, c.10 hectare area of land, of which around 7ha were stripped in two phases (Fig. 3). Archaeological background The phased extraction of sand and gravel has been preceded by a number of archaeological investigations initiated as part of the planning process (Ford 1992; 2001a; 2001b; Cass 1998; Anthony 2002; Anthony and Ford 2003; Ford and Hall 1993; Oram 2004; Pine 2003; Smith 1996; Taylor 2008; Weaver 1995) (Fig. 2). The initial document set out the archaeological background to the environs of the site and established a framework for archaeological investigation of the quarry following the guidance in Archaeology and Planning (PPG16 1990). The preliminary study included minimally invasive fieldwork comprising fieldwalking, test pitting and geophysical survey. Of relevance to the present site, the preliminary fieldwork (phase 1) revealed a dense scatter of worked flint towards the western edge of the current area (Ford and Hall 1993) with the presence of a Roman building and other features revealed in evaluation trenches to the north (Weaver 1995). A short distance to the north-east of the site (phase 5) a complex of medieval boundary features was identified (Oram 2004) which continued into the adjacent area (phase 6). Of note from phase 6, was the discovery of an early Bronze Age Beaker with barbed and tanged arrowhead (Taylor 2008). A number of linear features from the previous investigation (SOM05/62, GP3) seemed likely to continue across the current area: these were not well dated but probably medieval. Nea Farm, though, is chiefly notable for its Upper Palaeolithic site (SOM 01/41: Barton et al. 2009). The Excavation Three disconnected areas were stripped of topsoil and overburden (Fig. 3). The stripping took place under constant archaeological supervision using a 360º type machine fitted with a toothless grading bucket to expose the uppermost surface of the archaeological features. This revealed a modest number of deposits consisting of one ditch, 25 gullies, 13 pits and 17 postholes. All discrete features were fully excavated with linear features 2

extended beyond the agreed 10% sample in order to retrieve further dating evidence, usually to no avail. Despite there having been a flint scatter identified during test pitting, it appears that since the golf course was instated some significant landscaping had taken place resulting in the loss of this material. A catalogue of features excavated is presented in Appendix 1. Ditches and gullies Just two of the numerous linear features produced dating evidence (pottery). Ditch 1041 produced 13 abraded sherds of late Bronze Age/ early Iron Age pottery, from a single excavated slot and its value as for dating this feature must be regarded with caution. Short gully 13 is more securely dated to the medieval period, but unfortunately is not obviously related to any other linear feature. Unless noted otherwise, none of the other linear features produced any finds. Ditch 1041 was aligned WNW ESE and terminated at its western end. Its relationship with gully 123 was uncertain, but it cut gullies 1040 and 1043. Eight slots (6, 19, 23, 120, 124, 203, 204 and 206) were dug across it measuring between 1.37m and 3.00m wide and between 0.27m and 1.20m deep. Slot 6 produced 13 abraded sherds of Late Bronze Age/Early Iron Age pottery and five pieces of burnt flint. Slots 23 and 120 produced four and three pieces of burnt flint respectively. Gully 1030 was aligned approximately north-south and terminated at its southern end. It continues as gullies 1031, 1032 and 1034. Its northern end was truncated by pit 128/129 but neither it nor the pit could be dated. Three slots (125, 126 and 127) were dug across gully 1030 measuring between 0.60m and 0.70m wide and between 0.20m and 0.24m deep. Gully 1031 terminated at both ends but is clearly an interrupted a continuation of gully 1030. Similarly, it continues to the south as 1032 and 1034. Two slots (26 and 27) were dug across 1031 measuring between 0.50m and 0.70m wide and between 0.09m and 0.19m deep. Three slots (28, 29 and 30) across gully 1032 showed it measured between 0.42m and 0.80m wide and between 0.05m and 0.10m deep. Gully 1034 terminated at its north eastern end but continued out of the excavation area to the south. It was a continuation of 1032 and was cut by 1033. Three slots (31, 32 and 35) were dug across it measuring 0.70m wide and 0.16m deep. Gully 1033 was aligned NW SE and terminated at both ends. It cut gully 1034. Six slots (33, 34, 36, 37, 38 and 121) were dug across it measuring between 0.36m and 0.60m wide and between 0.13m and 0.25m deep. Gully 1035 was aligned approximately north-south and terminated at its northern end. Its relationship with gully 1037 was unclear. Six slots (46, 47, 48, 49, 101 and 102) were dug across it measuring between 0.52m and 3

1.46m wide and between 0.13m and 0.36m deep. Slot 48 contained two pieces of burnt flint. No other finds were recovered. Gully 1036 was aligned east-west and terminated at its eastern end. Its western end was unclear. Two slots (117 and 118) were dug across it measuring between 0.33m and 0.40m wide and between 0.06m and 0.08m deep. Gully 1037 was a short length of gully aligned NE SW that terminated at its eastern end. It continued as gully 1038 and its relationship with 1035 was unclear. Two slots (100 and 104) were dug across it measuring 0.51m wide and 0.07m deep. Gully 1038 was aligned NE SW and terminated at both ends. It was a continuation of 1037 and continued as 1039. Three slots (105, 106 and 107) were dug across it measuring between 0.41m and 0.77m wide and between 0.07m and 0.19m deep. Gully 1039 terminated at both ends and was a continuation of 1038. Three slots (110, 111 and 112) were dug across it measuring between 0.35m and 0.73m wide and between 0.04m and 0.13m deep. Gully 1040 was a small stretch of curving gully aligned north-south but curved through a right angle to the west before terminating. It was cut by ditch 1041. Three slots (18, 20 and 114) were dug across 1040 measuring between 0.40m and 0.45m wide and between 0.09m and 0.14m deep. Gully 1042 was aligned north-south and terminated at its northern end. Two slots (11 and 12) were dug across it measuring respectively 0.70m and 0.80m wide and 0.13m and 0.17m deep. Slot 11 produced two pieces of burnt flint. No other finds were recovered. Gully 1043 was aligned east-west, terminated at both ends and a southward-extending spur of it was cut by ditch 1041. Two slots (21 and 22) were dug across it measuring 0.56m and 0.81m wide and 0.08m and 0.16m deep. Gully 1044 was aligned NW SE, broadly parallel to ditch 1041, and terminated at its western end. Two slots (4 and 5) were dug across it measuring 0.84m and 0.94m wide and 0.33m and 0.39m deep. Slot 4 contained a single flint flake. Gully 1045 was aligned north-south and terminated at its southern end. It is possible that this gully may be a continuation of gully 1042. Three slots (222, 223 and 224) were dug across it measuring between 0.70m and 0.80m wide and between 0.15m and 0.20m deep. Gully 1046 was at a right angle to 1045, possibly indicating the corner of an enclosure. It was aligned approximately east-west and terminated at its eastern end to form a short gap with the southern terminal of 1045. Two slots (220 and 221) across it measured 0.67m and 0.83m wide and 0.10m and 0.14m deep. 4

Gully 1047 was aligned NE SW and terminated at both ends. Three slots (213, 218 and 219) were dug across it measuring between 0.40m and 0.77m wide and between 0.15m and 0.20m deep. Short stretches of gully 216 and 217 may be a south-westerly extension of this feature. Gully 1048 was aligned NE SW, parallel to 1047 and at a right angle to gully 1049 and terminated at its south-western end. The north-eastern end was unclear. Three slots (207, 208 and 214) were dug across it measuring between 0.50m and 0.65m wide and between 0.15m and 0.26m deep. Gully 1049 was at a right angle to gully 1048 and was aligned NW SE and terminated at both ends. Three slots (209, 210 and 211) were dug across it measuring between 0.30m and 0.80m wide and between 0.04m and 0.16m deep. Gully 200 lay on the extreme eastern margins of the site with both ends terminating beyond the site boundaries. It was 0.66m wide and 0.21m deep, aligned NE SW. Gully 13 was just 4m long and the excavated slot measured 1.20m in length, 0.62m wide and 0.15m deep. Its mid grey brown clayey silt fill (64) contained seven sherds of 12th- to 13th-century pottery. Pits and postholes A number of discrete features were observed in this phase of works, with the majority again undatable. Pits 201, 202, 205 and 215 may be the remnants of burnt out trees, perhaps of post-medieval date. Similar features have been observed across many of the other previously examined areas. Posthole 1 measured 0.62m in diameter and 0.14m deep. Its mid grey brown silty clay fill (52) contained one piece of burnt flint. Prehistoric Pit 3 measured 1.31m in length, 1.29m wide and 0.43m deep. Its mid/dark grey gravelly silt fill (54) was found to contain 42 pieces of Beaker pottery representing four vessels, as well as six pieces of struck flint comprising three scrapers and three flakes of probable early Bronze Age date and one piece of burnt flint. A radiocarbon date on non-wood charcoal from this pit produced a date of 2134 2078 cal BC (KIA39674, see below). Wet sieving produced a small amount of other charred plant remains with single items of wheat and cereal chaff along with various weed seeds. Oak charcoal was dominant, followed by ash, hazel and some alder. Pit 24 measured 0.56m in diameter and 0.11m deep. Its dark brown grey silty sandy clay fill (77) contained three sherds of earlier Neolithic pottery and 31 pieces of struck lithic material. These included a serrated blade of black chert and 19 flint flakes. A radiocarbon date on hazelnut shell produced a date of 3517 3393 cal BC (KIA39673, below). Wet sieving produced a small amount of other charred plant remains with 5

single items of wheat, cereal and chaff present along with various weed seeds. Oak and alder charcoal were the main woody species present with hazel and ash also present. Pit 25 measured 0.40m in length, 0.70m wide and 0.22m deep. Its dark brown silty clay fill (78) contained seven sherds representing at least two vessels. The dating of these small sherds is difficult and they may represent vessels of Middle Bronze Age, or Iron Age date. Roman Pit 42 measured 0.42m in length, 0.95m wide and 0.36m deep. Its dark brown silty sand fill (95) contained a single sherd of Late Iron Age/Early Roman pottery and two pieces of burnt flint. Also of note was the presence of a Late Iron Age/Roman urn dated to between 50BC and AD100. This was not deposited within a cut feature but present within the subsoil in the north-west corner of the site and had somehow managed to survive ploughing and the impact of the golf course. This is most likely due to its presence close to a line of former large trees. Medieval Posthole 16 measured 0.70m in diameter and 0.22m deep. Its dark grey brown silty clay fill (67) contained three sherds of 12th- to 13th-century pottery. Posthole 17 measured 0.55m in diameter and 0.19m deep. Its dark grey brown silty clay fill (68) contained two sherds of 12th- to 13th-century pottery and two pieces of burnt flint. Although undated, postholes 14 and 15 in close proximity to these two medieval features are likely to have been related and formed an irregular structure next to ditch 1041, possibly suggesting the ditch should also be of this date, despite its earlier pottery. Finds Prehistoric Pottery by Frances Raymond A small assemblage of prehistoric pottery (70 sherds, weighing 464g.) came from three pits (3, 24 and 25). The earliest group from pit 24 is composed of a few tiny fragments of earlier Neolithic pottery associated with a radiocarbon date of 3517 3395 BC (KIA39673). Most of the pottery is from pit 3 (60 sherds, weighing 449g.), which produced a Beaker-related domestic assemblage deposited between 2134 2078 BC (KIA39674). The few sherds from pit 25 exhibit characteristics which suggest they are most likely to be of middle Bronze Age or middle Iron Age origin. 6

The pottery was analysed in accordance with the guidelines of the Prehistoric Ceramic Research Group (PCRG 1997). The variables recorded included fabric, form, vessel dimensions, decoration, surface treatment, colour, wall thickness, sherd size and condition. The pottery was quantified within each of these categories both by number and weight. The descriptive data relating to these recorded attributes are in the project archive. Pit 24 produced three lightly to moderately abraded sherds of pottery (weighing 4g.), including one rim fragment. This is simple, flattened and upright and provides no evidence for the vessel profile. The sherds have very dark grey to dark reddish grey exteriors (5YR3/1 and 4/2) and are made from the same soft glauconitic shelly ware (glv/1). This incorporates moderate quantities of glauconite altered to limonite (<0.06 to 1mm.) and similar amounts of shell, which has leached or burnt out during firing leaving a series of characteristic voids. The seven fragments of pottery from pit 25 (weighing 11g) are derived from two vessels. One of these has a simple, rounded and upright rim (not illustrated), a very dark grey burnished exterior and is made from a fine shelly ware tempered with flint (Fsh/1). This incorporates common amounts of crushed burnt flint (up to 2mm) and sparse shell (up to 2mm). Most of the sherds are in this fabric and are almost certainly part of a single vessel (six sherds, 10g). The only exception is a split wall fragment in a similar fine and soft ware tempered with common quantities of crushed burnt flint (up to 2mm; F/1). There is insufficient evidence to allow for the certain phasing of this assemblage, which exhibits characteristics pointing to two possible alternatives. The simple rim, burnished exterior and fabric of one of the vessels are typical of a middle Bronze Age globular urn, but are equally consistent with the ceramic repertoire of the middle Iron Age. The Beaker Related Assemblage from Pit 3 The assemblage from pit 3 (60 sherds, weighing 449g.) represents the remains of four vessels (Fig. 6: 1 5). Two have similar upper profiles with in-turned necks and both have reddish brown exteriors (Fig. 6: 1 and 2; 5YR4/4). It is uncertain whether these are from bipartite or long-necked beakers with relatively low and accentuated waists, although the profile would suggest that the latter is more probable. The cordoned vessel is embellished with zoned motifs composed of alternating bands of horizontal and vertical lines (Fig. 6: 1). The exterior is moderately abraded so that the decorative technique is no longer clear. The impressions display irregularities which rule out incision, while their slightly sinuous character points to the use of cord rather than comb. The vessel is represented by 18 sherds (106g), which include approximately 24% of the rim. It is made from a soft, fine ware (GS/1) containing rare glauconite (0.2 to 0.5mm) and sparse, rounded, fine to medium quartz sand (0.1 to 0.5mm), which has been tempered with common quantities of grog (up to 2mm). 7

The plain beaker is represented by one lightly abraded fragment comprising only 7% of its rim (Fig. 6: 2; weighing 63g). It is made from the same soft medium grade ware as the rest of the sherds from the pit (G/1; also used for Fig. 6: 3 5). This is similar to GS/1 in that it is tempered with common amounts of grog, but is slightly coarser (up to 3mm.). The shape of the rim from the larger cordoned vessel is most consistent with a domestic style beaker (Fig. 6: 3). The technique used to produce the short crescentic impressions above the cordon is unclear. The beaker has a dark brown exterior (7.5YR4/3) and is represented by two sherds (weighing 40g.) including 10% of the rim. A small wall fragment (6g) decorated with short straight impressions (either comb or cord) may also be from this vessel (not illustrated). The assemblage additionally includes approximately 26% of the base and lower walls of a rusticated vessel, consistent with a domestic style beaker (Fig. 6: 4; four sherds, 80g.). This has a brown (7.5YR5/3) lightly to moderately abraded exterior and is embellished with apparently random fingertip impressions, some of which appear to be paired. Two light brown to reddish yellow (7.5YR6/4 and 6/6) wall sherds with crescentic impressions may be from this same vessel or from Fig. 6: 3 (Fig. 6: 5). As with much of the assemblage the technique used to produce these impressions is unclear. Discussion The few sherds of pottery from pit 24 are consistent in character with the radiocarbon date, which indicates that they were deposited towards the end of the earlier Neolithic. Shelly fabrics are common on sites of the period in the Wessex region (cf. Cleal 1995), being represented to the south of Somerley in the Bournemouth area at Haddon Hill, the Holdenhurst long barrow and at Southbourne (Wilson and Freestone 1980). The simple character of the rim is consistent with the south-western sub-style, but the evidence is far too slight to allow for an attribution of this type. The beaker assemblage is noteworthy for providing a partial insight into the range of vessels in contemporary use. Furthermore the bias towards rim and base/lower wall fragments (Fig. 6: 1 to 4 comprise half the assemblage by weight) suggests that the sherds were selected for deposition. The upper profiles of the decorated and plain beaker (Fig. 6: 1 and 2) indicate that they are long-necked Style 3 vessels (after Case 1993). The in-turned rims, the banded decoration and the association with domestic and rusticated forms (Fig. 6: 3 and 4) would place them within Case s (1993) Southern Group B. This exhibits northern influences and is thought to have emerged in southern England in the later third millennium BC. The Somerley ceramics recall a 8

comparable assemblage from a pit at Easton Lane near Winchester, which produced sherds from a Style 3 beaker and fingernail impressed domestic vessels, including one with a horizontal ridge and internal rim bevel comparable to Fig. 6: 3 (Ellison 1989, fig. 86, P28-P30). Case (1995) suggested a date range in the fourth quarter of the third millennium BC for Group B assemblages in the south, a chronology which is supported by the radiocarbon date from Somerley. Late Bronze Age, Early Iron Age, Roman and Medieval Pottery by Malcolm Lyne The pottery fabrics are the same as those identified for previous phases of work on the site (Appendix 2). Late Bronze Age/Early Iron Age Thirteen abraded sherds (22g) were recovered from ditch 1041 (slot 6, fill 57). Late Iron Age/Roman 183 sherds were of this date but 182 of those sherds came from subsoil contexts. One abraded sherd was recovered from pit 42. 175 of the subsoil sherds were from a single vessel. Medieval Twelve sherds of medieval pottery of 12th- to 13th-century date wee recovered from a short gully (13) and two posthole-sized features (16,17). Struck Flint by Steve Ford A small collection comprising 39 struck lithics was recovered from the site, 37 of which came from two stratified deposits of Early Neolithic and Early Bronze Age date (Appendix 3). Pit 3 (fill 54) of Early Bronze Age produced just six struck flints: three flakes and three scrapers. The scrapers are all well made on flakes, with one particularly so, with extensive lateral invasive retouch, almost sufficient to classify it as a knife (draw???). The flint is in a fresh condition and is derived from a local gravel source. Pit 24 (fill 77) of Early Neolithic date produced 31 struck lithic pieces. One serrated blade was made of a fine grained black chert, probably originating from the Isle of Portland. The remaining items were made of local gravel flint. The assemblage comprised 19 flakes, five certain or possible narrow flakes, four bashed lumps (tested nodules), two core fragments and two retouched flakes. The core fragments and bashed lumps show many frost fractures. Two of the flakes had been burnt. The relatively high proportion of narrow flakes in the assemblage is consistent with the date of this pit. 9

Gully 1044 (slot 4, fill 55) contained a single flake in a fresh condition. One patinated and abraded flake was recovered from the subsoil. Burnt Flint by Andy Taylor Twenty-three pieces of unworked burnt flint were recovered during the excavation weighing a total of 1049g, with no marked concentrations. Carbonized Plant Remains by Rosalind McKenna Four samples were submitted for an analysis of their palaeoenvironmental remains. Three samples are from Early Neolithic, Early Bronze Age and Middle Bronze Age/Iron Age pits and the other from an undated gully (Appendix 4). The samples were floated with the residues collected using a 0.25mm mesh. The heavy residues (the material which does not float) were not examined, and therefore the results presented here are based entirely on the material from the flot, with the exceptions of some larger pieces of charcoal which had been sorted from the heavy residues. The flot was examined under a low-power binocular microscope at magnifications between x12 and x40. Full data are available in the archive. The flots were sieved into convenient fractions (4mm, 2mm, 1mm and 0.3mm) for sorting and identification of charcoal fragments. Identifiable material was only present within the 4mm and 2mm fractions. For each sample, a random selection of 50 fragments of charcoal of varying sizes was made, and these were identified using the wood identification guides of Schweingruber (1978) and Hather (2000). Taxa identified only to genus cannot be identified more closely due to a lack of defining characteristics in charcoal material. Charcoal fragments were present in all of the samples, however, only samples from pits 3 (54) and 24 (77) produced enough identifiable charcoal fragments to be of interpretable value. Both samples consisted primarily of wood charcoal. Fragments of up to 50mm were present in sample 1, and up to 58mm in sample 2. Randomly selected charcoal fragments were taken from the 4mm and 2mm fractions fragments smaller than this lack the defining characteristics required for a confident identification. The preservation of the charcoal fragments was variable even within the samples: the majority of the charcoal in sample 2 (24,77) were poorly preserved. Both samples contained oak as the dominating species, with sample 3 having a significant component of ash and sample 14 a significant component of hazel. There were also occasional fragments of other taxa: Table 2 (Appendix 4). The fragments of charcoal were too small to give any indication of the size of the timber which 10

they came from. Bark was also present on some of the charcoal fragments, and this indicates that the material is more likely to have been firewood, or the result of a natural fire. All of the taxa identified are native British trees. The types represented are deciduous types, and are all relatively common woody taxa of open woodland, hedgerow and scrub. The presence of alder may indicate local availability and exploitation of wetter areas such as fen or floodplain. Most of these species, with the exception of Alnus, have good burning properties and would have made a fire suitable for most purposes (Edlin 1949). Similar assemblages have been recovered from Neolithic features at other sites in southern Britain (e.g. Gale 2004). Charred plant macrofossils were present in the two flots from samples 1 (pit 3, 54) and 2 (pit 24,77), but only in small volumes and poorly preserved. Root / rootlet fragments were also present within the samples. This indicates disturbance of the archaeological features, and this may be due to the nature of some features being relatively close to the surface, as well as deep root action from vegetation that covered the site. This disturbance is further confirmed by the presence of waterlogged plant macrofossils present in limited volumes in a lot of the samples. The preservation of these was excellent and it is probable that they are modern contaminants. Those present (Sambucus nigra, and Rubus sp.) are often found in archaeological samples as a modern contaminant. Although prior work has been carried out at this site, and sites within the local proximity, there has been a lack of environmental data present. These findings therefore begin to establish an initial dataset for the area, which can be added to with further excavations that may take place. Radiocarbon dating Two radiocarbon determinations were obtained by the University of Kiel on sample of hazelnut and miscellaneous non-wood charcoal taken from pits 3 (54) and 24 (77). The dates were calibrated using the INTCAL04 curve (Reimer et al.2004). Details of methodology are in the archive: in summary, the results are considered reliable. KIA39673 Sample 2 pit 24 fill 77 Hazelnut shell Radiocarbon Age: BP 4663 ± 24 One Sigma Range: cal BC 3507 487 (Probability 16.2 %) (Probability 68.3%) cal BC 3473 3427 (Probability 44.0 %) cal BC 3381 3372 (Probability 8.1 %) Two Sigma Range: cal BC 3517 3395 (Probability 84.0 %) (Probability 95.4 %) cal BC 3386 3368 (Probability 11.4 %) 11

KIA39674 Sample 1 cut 3 fill 54 miscellaneous non-wood charcoal Radiocarbon Age: BP 3653 ± 28 One Sigma Range: cal BC 2120 2095 (Probability 16.4 %) (Probability 68,3 %) cal BC 2041 1973 (Probability 51.9 %) Two Sigma Range: cal BC 2134 2078 (Probability 94.4 %) (Probability 95,4 %) cal BC 2063 1943 (Probability 1.0 %) Conclusion This phase of the ongoing fieldwork has again revealed a moderate volume of archaeological deposits that can now be seen to be typical of the use of much of the area of Nea Farm as a whole. The earliest deposits are few in number and are of prehistoric date. The most significant finding, and the first event of this period to be revealed at Nea Farm, is an Early Neolithic pit. The pit contained pottery and struck flint and a charred plant remains repertoire of economic species that includes both gathered and cultivated crops. No other cut features, not even undated ones were present in the vicinity of the pit, but nevertheless it is considered to reflect the presence of an occupation site, with other evidence lost to the plough. Prior to this phase of fieldwork a test-pitting evaluation had recovered significant quantities of struck flint from the northern part of this extraction area (Ford and Hall 1993). Some of the durable flintwork recovered, especially the distinctive narrow flake component, may relate to Early Neolithic use of the site as well as the previously recorded Mesolithic use, though few flint finds were coincident with the location of the Neolithic pit. During monitoring of the topsoil stripping, surprisingly, no struck flint finds were observed from the topsoil and only a single piece from the subsoil and it appears possible that a significant amount of de-stoning of the topsoil had taken place during construction of the golf course. Early Neolithic occupation deposits are generally rare, especially so for the Middle Avon Valley (eg Light et al. 1995) in contrast to the better known chalkland areas to the north west. Much of the above discussion is equally applicable to the single, isolated Early Bronze Age pit on the site which produced Beaker pottery, struck flint and a little charred plant evidence for arable farming. It is again considered to reflect the presence of an occupation site for this period, when below ground traces of occupation deposits are even rarer than for the Early Neolithic, although flint scatter evidence is more frequent (Light et al. 1995). Deposits of this period though, have been recorded at Nea Farm previously, with the presence of a pit or treehole with a beaker and barbed and tanged arrowhead observed in an earlier phase of excavation to the northwest (Taylor 2008). That deposit is considered to have been a burial though no bones were preserved, probably due to the acidic ground conditions. In contrast to the Neolithic, Early Bronze Age activity in the valley is very well attested, especially from the large number of round barrows present on both the valley floor and adjacent 12

terraces, and a curiosity is the absence (so far) from Nea Farm of any further round barrows or their levelled counterparts (ring ditches). The final prehistoric feature, is difficult to assess due to uncertainty over its actual date with the pottery recovered being too fragmented to allow a distinction between Middle Bronze Age or Iron Age to be determined. In the light of the above discussion, it may also be a vestigial trace of a former occupation site. The most distinctive features within the excavated area are gullies and a ditch representing boundary features which indicate the presence of a managed landscape. The layout is intermittent, perhaps by design, but more likely to be due to subsequent removal of the shallowest elements by later ploughing. All of the previous phases of excavation work at Nea Farm have revealed similar ephemeral traces of such landscape organization. However, there is a serious difficulty in placing these features in context due to the almost complete lack of dating evidence, an aspect frequently shared with the previous areas. The previous work (e.g. Cass 2008; Taylor 2008) has hinted at a medieval date for some of the features but with the possibility that some are of Roman date. There are stray finds and cut features of prehistoric, possibly Roman and medieval date, and the linear features here could belong to all or some of these periods. The largest pottery assemblage from a linear feature actually comprised several sherds of LBA/EIA pottery from a single excavated segment but residuality of this group cannot be ruled out. Several of the linear features excavated here form rectilinear arrangements with axes shared with elements recorded in earlier phases. Two orientations are clearest: one aligned north-south and the other NW SE. At the least, this indicates some degree of remodelling of the land divisions, but whether this can be taken to indicate the presence of, say, Roman and medieval systems is unclear. The evidence of extensive land divisions at Lower Farm, Pennington, Hampshire indicates numerous reorganizations all within the medieval period (Moore et al. 2008). The presence of the boundary features here dating from the medieval period (perhaps), as at Pennington also raises the topic of whether classic medieval open field ridge and furrow was ever present at Nea Farm and if so where it fitted within the sequence of development of the landscape at this site. Acknowledgements The excavation was funded by Tarmac Southern Limited. The author wishes to thank all who participated on site and behind the scenes. The excavation team consisted of the author, Aidan Colyer, James Earley and David Platt. Illustrations are by the author and Andy Mundin. References 13

Anthony, S, 2002, Nea Farm Quarry Phase 3, Somerley, near Ringwood, Hampshire, An archaeological evaluation, Thames Valley Archaeological Services, rep 02/03, Reading Anthony, S and Ford, S, 2003, Nea Farm Somerley, near Ringwood, Hampshire phase 2, A post-excavation assessment, Thames Valley Archaeological Services, rep 01/41c, Reading Barton, R N E, Ford, S, Collcutt, S N, Crowther, J, Macphail, R I, Rhodes, E and Van Gijn, A, 2009, A Final Upper Palaeolithic site at Nea Farm, Somerley, Hampshire and some reflections on the occupation of Britain in the Late Glacial Interstadial, Quartar, 56, 7 35 BGS, 1990, British Geological Survey, 1:50000, Sheet 314, Solid and Drift Edition, Keyworth Case, H, 1993, Beakers: deconstruction and after, Proc Prehist Soc 59, 241 68 Case, H, 1995, Some Wiltshire beakers and their contexts, Wiltshire Archaeol Natur Hist Mag 88, 1 17 Cass, S, 2008, Nea Farm, Phase GP 4, Somerley, Ringwood, Hampshire, Thames Valley Archaeological Services rep 07/53, Reading Coles, S, 2004, The excavation of three ring ditches at Mockbeggar Lane, Ibsley, Ringwood, Hampshire Stud 59, 31 64 Cleal, R M J, 1995, Pottery fabrics in Wessex in the fourth to second millennia BC, in I Kinnes and G Varndell (eds), Unbaked Urnes of Rudely Shape Essays on British and Irish Pottery for Ian Longworth, Oxbow Monogr 55, Oxford, 185 94 Edlin, H L, 1949, Woodland crafts in Britain: an account of the traditional uses of trees and timbers in the British countryside, London Ellison, A B, 1989, The Neolithic and Bronze Age pottery, in P J Fasham, D E Farwell and R J B Whinney, The Archaeological Site at Easton Lane, Winchester, Hampshire Fld Club Monogr 6, Winchester, 83 91 Ford, S, 1992, Nea Farm Quarry, Somerley, near Ringwood, Hampshire, an assessment of the proposed sand and gravel extraction (input to Environmental Statement), Thames Valley Archaeological Services rep 92/21, Reading Ford, S, 2001a, Nea Farm Quarry, Somerley, near Ringwood, Hampshire Phase 2, an archaeological evaluation, Thames Valley Archaeological Services rep 01/41, Reading Ford, S, 2001b, Nea Farm Quarry, Somerley, near Ringwood, Hampshire Phase 2, An archaeological evaluation - stage 2 (test-pitting), Thames Valley Archaeological Services rep 01/41b, Reading Ford, S and Hall, M, 1993, Somerley Estate, Ringwood, Hampshire, Archaeological Evaluation, Thames Valley Archaeological Services rep 92/21b, Reading Fulford, M G, 1975, New Forest Roman Pottery, BAR Brit Ser 17, Oxford Gale, R, 2004, Charcoal from later Neolithic/early Bronze Age, Iron Age and early Roman contexts, in G Lambrick and T Allen, Gravelly Guy, Stanton Harcourt, Oxfordshire: the development of a prehistoric and Romano-British community, Oxford Archaeol Thames Valley Landscapes Monogr 21, Oxford, 445 6 Hather, J, G, 2000, The identification of Northern European woods; a guide for archaeologists and conservators, London Light, A, Schofield, J and Shennan, S J, 1995, The Middle Avon Valley, Survey: a study in settlement history, Proc Hampshire Fld Club Archaeol Soc 50, 43 101 Moore, H, Pine, J and Taylor, A, 2008, Prehistoric and Saxon features and Medieval land allotment at Lower Farm, Pennington, Hampshire, Hampshire Stud 63, 88 100 Oram, R, 2004, Medieval field systems at Nea Farm, Somerley, Ringwood, Hampshire: an archaeological excavation (Draft publication report), Thames Valley Archaeological Services rep 04/53, Reading PCRG, 1997, The Study of Later Prehistoric Pottery: General Policies and Guidelines for Analysis and Publication, Prehistoric Ceramics Research Group Occas Pap 1 and 2 (revised) Pine, J, 2003, Nea Farm Quarry, Somerley, Ringwood, Hampshire Phase 4, An archaeological evaluation, Thames Valley Archaeological Services rep 03/68, Reading PPG16, 1990, Archaeology and Planning, Dept Environment Planning Policy Guidance 16, HMSO Reimer, P J, Baillie, M G L, Bard, E, Bayliss, A, Beck, J W, Bertrand, C J H, Blackwell P G, Buck, C E, Burr, G S, Cutler, K B, Damon, P E, Edwards, R L, Fairbanks, R G, Friedrich, M, Guilderson, T P, Hogg, A G, Hughen, K A, Kromer, B, McCormac, G, Manning, S, Bronk Ramsey, C, Reimer, R W, Remmele, S, Southon, J R, Stuiver, M, Talamo, S, Taylor, F W, van der Plicht, J and Weyhenmeyer, C E, 2004, IntCal04 terrestrial radiocarbon age calibration, 0 26cal kyr BP, Radiocarbon, 46(3), 1029 58 Schweingruber, F H, 1978, Microscopic wood anatomy, Birmensdorf, Swiss Federal Inst Forestry Research Smith, A, 1996, Somerley Estate, Ringwood, Hampshire, Archaeological Excavation, Thames Valley Archaeological Services rep 95/64b, Reading Stace, C, 1997, New Flora of the British Isles, Cambridge Taylor, A, 2008, Nea Farm, Phase GP3, Somerley, Ringwood, Hampshire, 2005, An Archaeological Excavation, (Draft publication report), Thames Valley Archaeological Services rep 05/62, Reading. Weaver, S, 1995, Somerley Estate, Ringwood, Hampshire, Archaeological Evaluation, Thames Valley Archaeological Services rep 95/64, Reading 14

Wilson, G, and Freestone, I C, 1980, A Neolithic pot from Hadden Hill, Bournemouth, Proc Dorset Natur Hist Archaeol Soc 102, 95 6 15

APPENDIX 1: CATALOGUE OF EXCAVATED FEATURES Group Cut Deposit Type 50 Topsoil 51 Subsoil 1 52 Posthole 2 53 Posthole 3 54 Pit 1044 4 55 Gully Terminus 1044 5 56 Gully 1041 6 57 Ditch 7 58 Pit 8 59 Pit 9 60 Pit 10 61 Pit 1042 11 62 Gully Terminus 1042 12 63 Gully 13 64 Gully Terminus 14 65 Posthole 15 66 Posthole 16 67 Posthole 17 68 Posthole 1040 18 69 Gully 1041 19 70 Ditch 1040 20 71 Gully 1043 21 72 Gully Terminus 1043 22 73 Gully Terminus 1041 23 74 Ditch 1041 23 75 Ditch 1041 23 76 Ditch 24 77 Pit 25 78 Pit 1031 26 79 Gully Terminus 1031 27 80 Gully Terminus 1032 28 81 Gully Terminus 1032 29 82 Gully 1032 30 83 Gully Terminus 1034 31 84 Gully Terminus 1033 33 85 Gully 1034 32 86 Gully 1033 34 87 Gully 1034 35 88 Gully 1033 36 89 Gully 1033 37 90 Gully 1033 38 91 Gully Terminus 39 92 Posthole 40 93 Posthole 41 94 Posthole 42 95 Pit 43 96 Pit 44 97 Posthole 45 98 Posthole 1035 46 99 Gully Terminus 1035 47 150 Gully 1035 48 151 Gully 1035 49 152 Gully 1037 100 153 Gully 1035 101 154 Gully 1035 102 155 Gully 103 156 Gully Terminus 1037 104 157 Gully Terminus 1038 105 158 Gully Terminus 1038 106 159 Gully 1038 107 160 Gully Terminus 108 161 Gully Terminus 109 162 Gully Terminus 1039 110 163 Gully Terminus 1039 111 164 Gully Group Cut Deposit Type 1039 112 165 Gully Terminus 113 166 Posthole 1040 114 167 Gully Terminus 115 168 Posthole 116 169 Posthole 1036 117 170 Gully Terminus 1036 118 171 Gully 119 172 Posthole 1041 120 173 Ditch Terminus 1041 120 174 Ditch Terminus 1041 120 175 Ditch Terminus 1041 120 176 Ditch Terminus 1033 121 177 Gully Terminus 122 178 Gully Terminus 123 179 Gully 1041 124 180 Ditch 1030 125 181 Gully Terminus 1030 126 182 Gully 1030 127 183 Gully 128 184 Pit 129 185 Pit 200 250 Gully 201 251 Pit 202 252 Posthole 1041 203 253 Ditch 1041 204 254 Ditch 205 255 Pit 1041 206 256 Ditch 1048 207 257 Gully Terminus 1048 208 258 Gully 1049 209 259 Gully Terminus 1049 210 260 Gully 1049 211 261 Gully Terminus 212 262 Gully Terminus 1047 213 263 Gully Terminus 1048 214 264 Gully 215 265 Posthole 216 266 Gully Terminus 217 267 Gully Terminus 1047 218 268 Gully 1047 219 269 Gully Terminus 1046 220 270 Gully Terminus 1046 221 271 Gully 1045 222 272 Gully Terminus 1045 223 273 Gully 1045 224 274 Gully

APPENDIX 2: CATALOGUE OF POTTERY Cut Deposit Fabric Form Date-range No. sherds Wt (g) Comments Subsoil 3 BC3.11 bowl Bead-rim jarsx6 c.ad40 120 c.50bc AD100 Fresh Fresh Handled vessel 182 852 Fresh 6 57 1 Jars Late Bronze/Early Iron Age 13 22 Abraded 13 64 5 Scratch-marked cooking pot AD1100 1300 7 330 Fresh. 16 67 5 Cooking-pots AD1100 1300 3 8 Fresh 17 68 5 Cooking-pot AD1100 1300 2 16 Fresh 42 95 2 Bead-rim jar Late Iron Age to AD60 1 20 Abraded 1. Handmade irregular fabric with sparse up-to 5.00mm calcined flint fired patchy black/brown. Prehistoric. 2. Handmade rough black fabric with profuse up-to 0.30mm multi-coloured quartz filler. 3. Dorset Black-Burnished ware (BB1). 5. Very-fine-sanded rough buff fabric with profuse up-to 0.50mm multi-coloured quartz filler, fired black internally. Possibly Medieval.

APPENDIX 3: CATALOGUE OF STRUCK FLINT Feature Cut Deposit Subsoil 51 flake Gully 1044 4 55 flake Pit 3 54 2 flakes, 3 scrapers Pit 24 77 19 flakes; 3 narrow flakes; 2 spalls; serrated blade (chert); 2 retouched flakes; 4 bashed lumps; 2 core fragments

APPENDIX 3: PLANT REMAINS Table 1: components recorded from each of the samples. Sample Cut Context Charcoal Plant macrofossils 1 3 54 xxx x 2 24 77 xxx x 3 25 78 xxx - 4 120 175 xxx - x present xx some xxx abundant Table 2. Taxonomy and nomenclature follow Schweingruber (1978) for charcoal and Stace (1997) for the rest. Numbers are on a relative scale for plant remains (from 1 one or a few remains (less than an estimated six per kg of raw sediment) to 4 abundant remains (many per kg or a major component of the matrix)). and % of identified charcoal fragments for each sample. w/l) waterlogged. Sample 1 2 Cut 3 24 Context 54 77 Plant remains Corylus avellana L. Hazel - 1 Chenopodium spp./ Atriplex spp. Goosefoot/ orache 1 - Polygonum aviculare L. Knotgrass 3 - Rumex acetosella L. agg. Sheep s sorrel 2 - Rubus spp. (w/l) Bramble - 1 Sambucus nigra L. (w/l) Elder 1 - POACEAE unidentified small caryopsis Grass Family - 1 POACEAE unidentified medium caryopsis Grass Family - 1 cf. Triticum spp. Wheat 1 1 Indet. Cereal fragments - 1 Indet. Cereal chaff fragments. 1 1 UNIDENTIFIED 2 2 Charcoal Corylus avellana Hazel 12% 3% Fraxinus excelsior Ash 22% 2% Quercus Oak 58% 13% Alnus glutinosa Alder 4% 12% Indet. 4% 70%