Richborough connection project: some evidence of early Bronze Age spelt wheat and late Iron Age/Roman field systems at Hoath

richborough connection project: some evidence of early bronze age spelt wheat and late iron age/roman field systems at hoath

james holman and caroline russell

Archaeological remains were encountered in fifteen positions along the 20km route. However, the remains were often isolated and commonly not dateable. As a consequence, only at Site PC 22 at Hoath was it considered that the identified archaeology might make a more meaningful contribution to the local and regional setting. The results of the excavation at this site form the main topic of this article.

Elsewhere, a small number of prehistoric features were identified on Sites PC 20 and 22 in the Stour Valley and PC 40 and 45 in the marshes of the former Wantsum Channel. Further information regarding these sites can be found in the post-excavation assessment (Macintosh et al. 2019). Characterised largely by scattered pits and post-holes or lengths of ditch these sites provided further evidence for the prehistoric agricultural landscape, though the amount of data retrieved from the sites was minimal.

The post-Roman phases related to further local development of the agricultural landscape, with the most significant discoveries pertaining to this period being medieval drove-ways identified at Sites PC 45 and 46 in the Monkton Marshes (see Macintosh et al. 2019). A substantial but undated oven was identified on Site PC 19, with this feature likely Roman to medieval in date.

the pc 22 excavation

Site PC 22 lay toward the eastern end of the 20km route, some 8km from the centre of Canterbury (Fig. 1). The site covered an area 20m wide by 40m long and was locally prominent, positioned at the base of a spur of land formed by the confluence of the Sarre Penn (also known as the Nethergong Penn) to the north and a now largely dry minor valley to the south (NGR 621036 163225; Fig. 2). Orientated downslope, west to east, the site lay at a height of 7.60-6.25m (aod), on gently sloping, north-east facing terrain.

The site is mapped by the British Geological Survey as lying on a bedrock geology of clay and silt of the London Clay Formation, overlain by a superficial Head deposit of clay and silt (BGS). The excavation confirmed this, with the Head geology encountered at a 0.52-0.64m below the existing ground surface. This deposit sloped in height from 6.95m aod to 5.74m aod, dropping from west to east.

Archaeological overview

Cut into the geology was a series of ditches and discrete features, concentrated in the south-western half of the excavation area (Fig. 3). The ditches varied in width and orientation, with some re-cut. The presence of field boundary systems was attested to by ditches running perpendicular to one another, with a parallel pair appearing to represent a north-south aligned trackway that ran across the west end of the site. Discrete features were largely represented by pits, with a single potential un-urned cremation burial also identified.

Horizontal truncation of the archaeology was noted by the shallowness of some features (the smallest pits had a depth of just 0.07-0.11m) and the tapering ends to several linear features. This truncation is likely to have resulted from ploughing activity.

Early Bronze Age (2200-1700 bc)

A single, probably early Bronze Age, feature was identified within the excavation area, consisting of a pit (S22009). Initially thought to be late Iron Age or Roman on the basis of a single pottery sherd recovered from the upper fill, subsequent radiocarbon dating of cereal remains from the middle fill of the pit provided a date of c.1700-1500 calibrated bc.

The cereal remains were represented by a substantial deposit of emmer-type grains (Triticum cf. dicoccum) and chaff bases (Triticum dicoccum) (Carruthers 2022, 2-3). Two potential spelt glume bases were also tentatively identified but were far too poorly preserved for the identification to be confirmed (ibid, 5). Also recovered from this assemblage was a small quantity of barley, possibly a relict crop, and seeds from various species of weed including black bindweed (Fallopia convolvulus) and persicary (Persicaria sp.), as well as more unusual taxa typical of early prehistoric cereal assemblages such as black nightshade (Solanum nigrum) and cf. bittersweet (Solanum cf. dulcamara).

In addition to the cereal remains, slender charred stems (1-3mm wide) were frequent in the middle fill of the pit and also present in the lower fill (Carruthers 2022, 3-4). Most of the stems were curved in section and some of them retained a ‘heel’ where they had been pulled from a thicker, more-woody stem. These may have represented rods of willow that had been split, with any thin bark or spongy pith burnt away.

Overall, it is suggested that this pit was cut in the early Bronze Age with the emmer wheat being deposited in this period in the upper part of the middle fill. The fills then settled, or were perhaps disturbed, with the upper fill that contained the later pottery collecting in a shallow depression, which survived later ploughing in the top of the feature.

Iron Age and late Iron Age (300 bc-ad 43)

A 23.60m long north-east to south-west aligned ditch (S22058) that ran for some 23.60m from the northern limit edge of excavation contained a single pottery sherd, with a broad date of 500 bc-0 ad (Lyne 2019, 31). This feature seemed to represent the earliest in several phases of a ditch system that developed across the site in the Iron Age and Roman periods.

As discussed above, it is suggested that the uppermost fill of Early Bronze Age pit S22009 may have developed at this time, perhaps being deposited in a depression caused by the settling and possible disturbance of the earlier fill deposits. Two sherds of late Iron Age pottery, dated c.25 bc-ad 70, were recovered from this deposit (Lyne 2019, 31).

Early/Middle Roman (ad 43-260)

More substantive activity was present from the Early to Middle Roman period, represented by a series of field boundaries, pits and a possible trackway. The earliest boundary in this phase was north-west to south-east aligned ditch (S22097), that ran for some 17.3m across the south-west corner of the site. This ditch yielded forty-five sherds (281g) of pottery, most of which (forty sherds) came from the primary silting and included twenty-three sherds from biconical beakers of Monaghan’s classes 2G1 (1987, c.70-110) and class 2G2 (c.43-100), as well as fourteen fragments from a hand-made jar in black fine-sanded fabric B8 (c.43-60): two abraded residual sherds in late Iron Age grog- and calcined-flint tempered fabric were also present (Lyne 2019, 31).

A potentially related ditch (S22019) lay perpendicular to ditch S22097 and produced pottery dated to c.43-200 (Lyne 2019, 25). The two features likely intersected some 3.4m to the south of the stripped area. An undated ditch (S22106) lay in the south-east corner of the site and ran parallel to S22019, perhaps suggesting that it too formed part of this system.

Located 4m north of ditch S22097 was a large, roughly oval pit (S22034), the earliest of three such features that can be ascribed to this phase of activity. This pit measured 1.51m long, 1.00m wide and 0.30m deep, producing a single pottery sherd, dated c.ad 43-80 (Lyne 2019, 25, 32), along with iron nail fragments and a fragment of daub.

Apparently slightly later in date were an oval pit (S22015) and a roughly circular one (S22017), with pottery suggesting they were Middle Roman, dated c.170-250 (Lyne 2019, 25, 32). Both pits also contained fragments of tile, daub and of indeterminate animal bone, with seven hobnails and two glass beads recovered from pit 22017. Small quantities of charred plant remains were recovered from pit S22015, with these containing moderate amounts of emmer/spelt grains, spelt chaff, and emmer/spelt chaff (including a possible emmer glume base), small amounts of barley (Hordeum sp.) and oat grains (Avena sp.) (Carruthers 2022, 5). Of particular interest was that the upper fill contained a wider range of fruit and nut remains. These included a fragment of sloe stone (Prunus spinosa), a hawthorn stone (Crataegus monogyna), three fragments of hazelnut shell and a fragment of stone pine seed coat or ‘nutshell’ (Pinus pinea).

Mid/Late Roman (260-410)

A north-west to south-east aligned ditch (S22088) was partially exposed for some 5.8m, running perpendicular to ditches S22019 and S22106. Its intersection with ditch S22106 lay directly outside the area stripped. Pottery from this feature ditch S22088 spanned the period c.150-300, within the Middle to Late Roman period (Lyne 2019, 25, 32).

The most substantial ditch on site (S22038) formed a re-cut of ditch S22088, running for some 39m on a north-west to south-east alignment. With a maximum depth of 0.85m, the ditch was V-shaped in profile. Pottery recovered from the upper fills included sherds from a jar of Lyne-type 7B2 in Richborough/Canterbury grog-tempered ware (Lyne 2015, c.350-420) and a convex-sided dish of type 7A.16 in the East Kent siltstone-grog-tempered variant (ibid, c.370-420) (Lyne 2019, 26, 32). The presence of two sherds from a Black Burnished (BB) 1 cooking-pot (c.280-370) and forty-six from Thames-side industry products suggests that the bulk of this pottery is of late third- to mid fourth-century date. A recut of this ditch S22026 survived for only a 9.7m length along its south-west side and was far less substantial.

Ditch (S22046), which was partly obscured by later ditches, ran north-east for 4.7m, from the south-west corner of the site, before turning and heading north to cut the Early Roman ditch (S22097). Ditch S22046 was recut as ditch S22044, which was ploughed out after some 8.6m. The pottery assemblage from this ditch was similar in make-up to that from S22038, though no Richborough/Canterbury grog-tempered ware was present (Lyne 2019, 26, 32).

A substantial pit (S22034), lay immediately adjacent to ditch S22024. The feature measured c.1.8m wide, 2.14m long and 1.25m deep, and was irregular in profile with an uneven base. The pit contained eleven fills which were intermixed with apparent dumps of domestic refuse. This feature yielded the largest assemblage of pottery from the site (193 sherds). The pottery was of largely undiagnostic late third- and early fourth-century date, but a sherd from a Harrold shell-tempered ware cooking-pot (c.360-400+) and three large fresh fragments from a type C14 flagon in Oxfordshire red colour-coat fabric (c.350-400 suggest it is a late fourth-century feature (Lyne 2019, 27). Fragments of tile and animal bone, one of Roman glass, an iron hobnail, three iron nails, and a poorly preserved assemblage of charred plant remains (mostly grain with the occasional fruit stone) were also recovered.

Ditch S22044 was cut by a pair of north-south aligned ditches S22022 and S22024, approximately 1.5m apart. Only the west ditch, S22024, ran for 19.6m, across the full width of the site, as S22022 stopped short of the north baulk by 6.3m. The ditches may have formed a track or drove-way that ran parallel with the lie of the land. Stratigraphically, the ditches are Late Roman in date at the earliest. The fills produced several unabraded Late Roman sherds, dating to 350-420, although most of the pottery recovered was of Late Iron Age or earlier Roman date and evidently residual.

Undated

A number of features on the site remain undated. These include three small, circular, or oval pits (S22056, S22011 and S22005), which formed a rough line close to the southern limit of excavation. The features were of similar size to each other, measuring 0.37-0.42m in diameter, and were shallow, with a depth of only 0.07-0.11m. Both pits S22056 and S22011 had moderately sloping sides and a flat base, while S22005 had gently sloping sides and a concave base. Pit S22005 produced 304.5g of calcinated bone fragments, perhaps forming an un-urned cremation burial. Unfortunately, no determinate fragments were identified, due to a high level of fragmentation (Teoaca 2019, 46). Microscopic analysis of the bone suggest that some is human, though burnt animal bone is also present with this not uncommon in prehistoric and Roman funerary practices (Mays 1998, 214; Booth et al. 2015, 292, 295).

The fourth and final undated feature is a possible pit (S22031), that extended out from the south baulk to cut ditch S22097. It had exposed measurements of 1.22m by 1.10m and a depth of 0.72m. A reasonably well-preserved assemblage of charred plant remains was also recovered. Cereal grains consisted primarily of hulled wheat grains and chaff, with smaller quantities of weed/wild seeds. The ratio of grain to chaff to weed was 3:2:1 indicating that either semi-processed spikelets of hulled wheat or mixed waste had been burned and deposited in the pit. Chaff of both emmer and spelt (Triticum spelta) were represented, but spelt chaff was six times more frequent. The function of this feature was not clear, though it would seem quite late in date due to its position at the top of the stratigraphic sequence.

discussion

Evidence observed during both the archaeological excavations and monitoring undertaken along the pylon route identified features relating to landscape changes from the prehistoric period onwards. This was most clearly demonstrated on Site PC 22, where some significant past activity was anticipated due to its location at the confluence of two watercourses.

Here, an isolated Bronze Age pit S22009 formed the earliest evidence for exploitation. This pit is mainly of interest due to the large and well-preserved deposit of emmer wheat recovered from its fill, with this deposit typical of the prehistoric period. However, the two glume bases that possessed characteristics of spelt wheat remain of interest due to their date. Only two sites from east Kent have produced similarly early evidence for the introduction of spelt to the British Isles in the Early to Middle Bronze Age: Princes Road, Dartford (two Middle Bronze Age radiocarbon dates from an organic deposit; Pelling 2003) and, producing the earliest date yet obtained for spelt from the British Isles, a mixed emmer and spelt assemblage from a ditch at Monkton Road, Minster in Thanet, 1900-1700 calibrated bc (Martin et al. 2012). It would seem that that spelt gradually spread along the Thames into other parts of England and Wales during the Bronze Age, having first arrived from the continent in the area around the Isle of Thanet (Carruthers et al. 2013). Emmer was gradually replaced by spelt through the Iron Age, though it lingered longer in Kent than in some parts of the country, perhaps because it was well-suited to the local soils. Overall, this reflects the situation elsewhere in Kent, for example at the East Kent Access Road, where emmer survived as a relict crop (Hunter 2015, 528).

A hiatus in activity seems to have lasted through the Middle Bronze Age into the latter part of the Iron Age. The single sherd, dated 500 bc to ad 0, recovered from ditch S22058 hints at Iron Age settlement activity in the vicinity, but the focus of this settlement is not clear and is beyond the area investigated. Potential discontinuity of settlement between the Early/Middle Iron Age and the Late Iron Age is not unexpected and has been noted elsewhere, for example on Plateau 8 of Thanet Earth (Holman and Henshaw 2019, 268) and, perhaps, at Island Road, Hersden (Barrett 2006, 20). The latter site is of particular interest, with intense activity during the Early Iron Age but a distinct paucity of Late Iron Age activity prior to a resumption of occupation during the Roman period. At Thanet Earth, a substantial Early to Middle Iron Age settlement had been abandoned by the Late Iron Age, with settlement shifting some 175m to the east, on the opposing side of a buried valley (Holman and Henshaw 2019, 268).

Despite the small size of the excavation area, evidence from the Late Iron Age running through to the Late Roman period was plentiful, with the landscape split up into a distinct landscape system. This is perhaps not unexpected: increasing regularisation of settlement and development of trackway systems has been long noted in both the Thames Valley and Northern France during the Late Iron Age, albeit less clearly evidenced in Kent (Booth et al. 2007; Taylor 2007, 57-65, 113).

At Site PC 22, a field system lay on a north-west to south-east alignment. Associated pottery suggests it was laid out soon after the Claudian Conquest and survived into the Late Roman period. That ditches S22097, S22019 and S22106 formed part of a small field within this system seems likely. If so, at approximately 16m in width, this would compare well with those at Coldswood Road, Thanet, where fields were only c.10-15m wide by 20-30m long (Dinwiddy and Schuster 2009, 94-95).

Associated pits indicate domestic activity in the vicinity of the site throughout the Roman period, as well as exploitation of plants that prefer damper soils. The suggestion of nearby settlement was supported by the plant remains, in particular the imported pine nuts, a luxury foodstuff (Carruthers 2022, 6). Other similarly dated finds of stone pine from Kent (either seed coat or cone scales) include seed coat from a large ditch at Shelford Farm Estate, Canterbury (Carruthers and Allison 2010), cone scales and seed coat fragments from Marlowe Theatre, Canterbury (Giorgi 2016) and charred nutshell fragments from two pits at Monkton, Isle of Thanet (Pelling 2008). Continuing the story of its decline, emmer was recovered in such small quantities that it was likely present only as a contaminant to the now more ubiquitous spelt.

In the Middle Roman period, substantial north-west to south-east ditch S22038 was cut on the same alignment as the earlier field system but located slightly to the north. The presence of the Richborough/Canterbury grog-tempered ware sherds in S22038 would seem to indicate at least limited settlement continued in the vicinity into at least the late fourth century.

Quite how the north-south aligned trackway relates to the Roman activity is not clear. Stratigraphically this is one of the latest features on the site (together with pit S22031), cutting through ditch S22038, and containing Late Roman pottery. However, the possibility remains that these sherds could be residual and that the trackway of a later, possibly medieval, date. It should be noted that the orientation of this feature runs parallel to that of a road and, latterly, a footpath extending from Island Road (on the line of the Roman road from Canterbury to Thanet) to Hoath. If of Roman or even medieval origin, the PC 22 trackway would seem to form part of a fossilised landscape system elements of which survive into the present day.

acknowledgements

This report includes contributions by Wendy Carruthers, Malcolm Lyne and Adelina Teoaca. The archaeological programme was initiated by National Grid under the direction of RSK Environmental. Simon Mason monitored the project on behalf of the Heritage Conservation Group of Kent County Council and Rosanne Cummings on behalf of Canterbury City Council.

The fieldwork was directed by Adrian Gollop, Andrew Macintosh and Julie Martin with the assistance of George Carstairs, Hazel Mosley, Åsa Pehrson and Juan-Paulo Vasquez. Digital survey was conducted by Paul-Samual Armour. Environmental archaeology was the responsibility of Enid Allison with bulk sample processing undertaken by Åsa Perhson. Finds processing and recording was carried out by Michele Johnson and Jacqui Matthews with the assistance of Jo Stephenson. The project was managed by Diarmuid O’ Seaneachain on behalf of RSK Environmental and James Holman on behalf of CAT. The figures in this report were prepared by James Holman and Matthew Charlwood.

references

Andrews, P., Dinwiddy, K., Ellis, C., Hutcheson, A., Philpotts, C., Powell, A. and Schuster, J., Kentish Sites and Sites of Kent. A miscellany of four archaeological excavations, 57-174.

Barrett, D 2006, ‘Island Road, Hersden’, in Canterbury’s Archaeology 29, CAT, 17-20.

Booth, P., Brady, P., Good, O., Leivers, M., Powell, J. and Thacker, G., 2015 ‘Conquest and change: the Later Iron Age and Roman period’, in Digging at the Gateway. Archaeological Landscapes of south Thanet. The Archaeology of East Kent Access (Phase II). Volume 1: The Sies, 229-383.

Booth, P., Dodd, A., Robinson, M., and Smith, A., 2007, The Thames through Time – The Archaeology of the Gravel Terraces of the Upper and Middle Thames: the early historical period, AD 1–1000, Thames Valley Landscapes Monograph 27, Oxford University School of Archaeology.

British Geological Survey (http://mapapps.bgs.ac.uk/geologyofbritain/home.html).

Carruthers, W., 2021, ‘Richborough Connection Project Kent (RCP EX 21): the charred plant remains’, unpubl. CAT report.

Carruthers, W. and Allison, E., 2010, ‘Shelford Farm Estate, Broadoak, Canterbury, Kent: charred plant remains and other biological material from the palaeoenvironmental samples’, unpubl. CAT rerport.

Carruthers, W., Campbell, G. and Pelling, R., 2013, ‘The Introduction, dispersal and survival of some key cereal crops into central and southern England’, poster for the Sixteenth International Work Group for Palaeoethnobotany Conference, Thessaloniki, Greece.

Dinwiddy, K. and Schuster, J., 2009, ‘Thanet’s Longest Excavation: archaeological investigations along the route of the Weatherlees–Margate–Broadstairs waterwater pipeline’, in P. Andrews et al.

Giorgi, J., 2016, ‘The plant remains from Marlowe Theatre (MTC.EX09) and St Peter’s Lane (SPLC), Canterbury’, unpubl. CAT report.

Holman, J. and Henshaw, R., 2019, ‘Iron Age’, in Rady and Holman.

Holman, J. and Weekes, J., 2019, ‘Late Iron Age and Romano-British’, in Rady and Holman. (https://www.canterburytrust.co.uk/_files/ugd/805c19_5292909a1f894de0 b45 ba2e7d86d7843.pdf?index=true).

Hunter, K., ‘Plant Macrofossils’, in P. Andrews et al.

Lyne, M., 2015, Late Roman handmade grog-tempered ware producing industries in South-East Britain, Archaeopress, Roman Archaeology 12.

Lyne, M., 2019, ‘The Pottery’, in A. Macintosh et al., 24-33.

Macintosh, A., Gollop, A. and Holman, J., 2019, ‘Richborough Connection Project: archaeological assessment report’, unpubl. CAT Client Report 2019/117.

Margetts, A., Dowsett, A., Krawiec, K. and Bates, M., 2020, ‘A Geoarchaeological Borehole Transect From the National Grid (Richborough Connection Project) Development Consent Order 2017 (as amended): A Post-excavation Assessment and Updated Project Design Report’, Archaeology South-East Report No: 2019110.

Martin, J., Schuster, J. and Barclay, A.J., 2012, ‘Evidence of an Early Bronze Age field system and spelt wheat growing, together with an Anglo-Saxon sunken featured building at Monkton Road, Minster in Thanet’, Archaeologia Cantiana, 132, 43-52.

Mays, S., 2011, The Archaeology of Human Bones, Routledge, London.

Monaghan, J., 1987, Upchurch and Thameside Roman Pottery: A Ceramic Typology; first to third centuries A.D, BAR, British Series 173, Archaeopress.

Pelling, R., 2003, ‘Charred plant remains’, in P. Hutchings, ‘Ritual and Riverside Settlement: a multi-period site at Princes Road, Dartford’, Archaeologia Cantiana, 123, 71-76.

Rady, J. and Holman, J., 2019, Beneath the Seamark: 6,000 years of an Island’s History, archaeological investigations at ‘Thanet Earth’, Kent 2007-2012, CAT Technical Paper 2.1 (https://www.canterburytrust.co.uk/_files/ugd/805c19_5292909a1 f894de0b45ba2e7d86d7843.pdf?index=true).

Taylor J., 2007, An Atlas of Roman Rural Settlement in England, CBA Research Report, 151.

Teoaca, A., 2019, ‘The Cremated Bone’, in Macintosh et al., 45-46.

Wilson, T., 2019, ‘Worked flint’, in A. Macintosh et al., 33-35.

Fig. 1 Location of site PC22.

Fig. 2 The topographic setting of site PC22. (Contains Ordnance Survey data © Crown copyright and database right 2019. Digital surface model based upon 2020 lidar data collected for the Environment Agency on a 1m or finer grid. Both dtat sets are public sector information licensed under the Open Government Licence v3.0

http://nationalarchives.gov.uk/doc//open-government-licence/version/3/).

Fig. 3 Site PC22 phased archaeology.