A regionally important Early Iron Age Pottery Group: The Manor Farm Pub Site, High Street, Rainham, Kent

A regionally important Early Iron Age pottery group: the Manor Farm Pub site, High Street, Rainham

mike seager thomas

with Sarah Barrowman, Barry Bishop, K. Le Hégerat and Kevin Reilly

Excavations by Pre-Construct Archaeology to the rear of the former Manor Farm pub, High Street, Rainham, have yielded a major assemblage of Early Iron Age pottery of a type previously unknown in the region but with close analogues from east Kent. By aiding its recognition locally and demonstrating the presence or absence of relationships between the region and elsewhere, this will contribute greatly to our understanding of the period in the county.

The site consists of a probable post-built roundhouse, containing and surrounded by pits, a four-post structure, and a number of other features, all bounded to one side by a ditch – a typical Iron Age enclosure. The excavated area was a focus of activity for an extended period of time and provides evidence for a range of everyday domestic activities including pottery use, food preparation, the long-term care of animals, and localized middening.

None of us really knows how many Iron Age sites have been excavated in Kent but it must now be well in excess of a hundred. Reliably documented sites range from groups of postholes and pits and stray cremations, through house excavations, to cremation cemeteries, hillforts and other huge multi-feature complexes. Pottery from these sites can be ascribed to a variety of well known and well sequenced traditions – post Deverel-Rimbury, the final ‘decorated’ phase of which is associated with the end of the Late Bronze Age and the beginning of the Iron Age (Needham 1996); early La Tène or ‘Marnian’, in Britain also usually dated to the beginning of the Iron Age (Cunliffe 2005, 98); saucepan and later La Tène pottery, conventionally dated to the Middle Iron Age (Orton and Cunliffe 1984); Wealden pottery, likewise probably Middle Iron Age (Seager Thomas 2010, 16); and Belgic, Atrebatic and other immediately pre-Roman forms, dated to the Late Iron Age (Green 1980; Thompson 1982). Confusingly, the nomenclature applied to these traditions varies in the literature, as does their precise Three Age System dating. There is also some overlap between chronologically adjacent traditions. Nonetheless the overall sequence is clear and should provide a secure foundation upon which to build our knowledge of Iron Age Kent.

There are problems to be taken into account however. As recently as 2007, Professor Tim Champion stated that pottery of the Early Iron Age, the period to which this Rainham site and pottery assemblage belong, was unknown in central and western parts of the county, inferring ‘a markedly regional distribution’ for contemporary traditions recognized in east Kent (Champion 2007, 297). In fact at the time the assertion was made it was already erroneous. Probable Early Iron Age pottery, which had not been and still has not been published, had been excavated on the Isle of Grain in 1999, while British analogues for the east Kent material were known from as far away as West Sussex (where they also remained unpublished) (Seager Thomas 2008, 41). Such failure to disseminate information is endemic in archaeology and has the unhappy result of specialist knowledge remaining unshared between professionals.

Equally problematic is the failure of specialists to agree what things are called or to what date they belong – even while acknowledging them to belong to the same, or related traditions! Kent has form here. As late as the 1990s Nigel Macpherson-Grant was calling Deverel-Rimbury pottery Late Bronze Age and ‘developed’ (as opposed to ‘decorated’) post Deverel-Rimbury pottery Late Bronze Age/ Early Iron Age (e.g. Macpherson-Grant 1992; 1994), years after the archaeological community outside the county had revised their Three Age System dating for these traditions backwards (to the Middle and Late Bronze Ages, respectively). Likewise he eschewed the Early Iron Age date for early La Tène/ ‘Marnian’ pottery usual in Britain, instead following the Dutch in calling it Early to Middle Iron Age (Broeke 1987, fig. 5; Macpherson-Grant 1989). When a site is fully published this is not a problem; we can see for ourselves to what tradition it belongs and where it falls in the foregoing sequence. But when it is not, and data reaches us in a summary form, it plays havoc with our interpretations. Typical is the attribution of pottery such as that from the present site to the Middle Iron Age, which, owing to the scarcity of the saucepan pots, is not widely recognized locally. It might just date to the middle of the Iron Age, but it was not Middle Iron Age, and it does not fill the gap left by the absence or non-recognition of saucepan pottery locally.

By contrast there is nothing ambiguous about the present assemblage. Its key features are the tradition to which it belongs: essentially the same as east Kent’s Early Iron Age pottery, which remains poorly understood regionally and – till the excavation of the present assemblage – unrepresented locally. The presence in the material of large, diverse, mostly well-preserved and apparently closed context assemblages therefore contributes significantly to our understanding. This is exactly what is required for reliable interpretation. Their reconstruction provides a useful point of comparison with other assemblages of the same and different dates from elsewhere, and helps place both the pottery assemblage and the site as a whole in a wide regional context, while their composition and associations as revealed through excavation contributes to archaeology’s first next step: understanding the site itself. Should the analysis published here prove wrong, it is at least published, so that colleagues can make up their own minds about it. Either way, it has the potential to add greatly to the clarification of a range of outstanding issues related to our understanding of the Iron Age in the county.

interpreting the site

Behind the former Manor Farm pub, at the junction of High Street, Rainham (the Roman Watling Street), and Maidstone Road (TQ 81320 66040), the site lay on gently rising ground close to the 45m contour, from where it overlooked the Medway estuary, 2km to the north (Figs 1 and 2). Excavation showed it to be thick with cut features (Fig. 3). Nearly all – where dateable – belonged to the Early Iron Age. These comprised pits and probable postholes, burnt features, and two successive ditches. There were also many undated features – postholes and pits of course, and hundreds of stake holes. Owing to their large numbers and a lack of clear patterning in their distribution, interpreting these is difficult, and any inferences about the site’s structure must be tempered by a recognition of the area’s extended use, which was demonstrated by the recovery from the site of a handful of early (Mesolithic to Early Bronze Age) struck flints (Bishop 2012a, 123-4) and small assemblages of Late Bronze Age, Middle Iron Age and Roman pottery (Seager Thomas 2012a). Only the Early Iron Age occupants of the site, however, demonstrably dug pits and postholes and it is our view that most of the excavated features pertain to that date. The stake holes by contrast include several that cut recent services and make up deposits – Barrowman 2012, 57. From their density and the intercutting of a handful of them, we thus infer that the area was a focus of activity during the Early Iron Age, and that this was of some time depth. We also suggest that the ditches bounded this activity or these activities. Most likely, therefore, we are looking at an Early Iron Age enclosure.

Amongst the excavated features it is possible to distinguish three or four probable structural groups. Out of a cluster of features in the south-west corner of the main excavation can be reconstructed a fairly small (5m diameter) post-built roundhouse, containing and surrounded by pits, and demarcated to the north and east (the west was not excavated) by an area with few features (Fig. 3 and Fig. 4; Tables 1 and 2). Most of the pottery came from the vicinity of this cluster of features, as did all of the larger context assemblages. A pit inside of the post-ring filled with burnt flint had characteristics of a Polynesian-type oven or a storage heater (pit 279) (cf. Seager Thomas 2005, 95-6). Also noteworthy are four postholes, two related by cross-joining sherds, which form a perfect rectangle close to the centre of the main excavation (a third feature, in line with one of the rectangle’s short sides, yielded another sherd from the same pot) (Fig. 3 and Fig. 4; Table 1), a wnw/ese linear alignment – or fence line – across the north of the trench (Fig. 3), and three widely spaced clay-lined pits (Figs 3 and 4; Table 2).

As for how the site was used we have two further clues. We will return to the distribution of the pottery below, but it should be emphasised here that as a group the finds comprise a typical domestic assemblage and do not appear to reflect any particular activity, ritual, industrial or otherwise, of the type occasionally associated with contemporary sites (but see Conclusion). The assemblage includes a wide range of pottery fabrics and types, part of a triangular loom weight, a few struck flints, charred cereals, the bones of cattle, sheep/goat, horse and pig, and large quantities of burnt flint. Absent was evidence for on-site metalworking (indeed the site yielded hardly any metal) and pottery making, and querns or quern fragments, which would have been used in processing the foregoing cereals. Of these, only the stones filling the ‘Polynesian oven’ gave any indication of being in functional situ. Comprising 50 per cent of the fill, these were apparently unmixed with unburned material (except for a single small sherd) and must have been at least partly clast-supported. (In a Polynesian oven, stones are heated in a pit, the food placed upon them, and the pit backfilled. Unless they are deliberately cleared out afterwards, the stones underlying the food remain in the pit – where there are sufficient stones, in a clast-supported layer. Characteristically these are soot-soaked. Where stones are to be used for saunas or boiling water, they are not normally heated in a pit). Otherwise different categories of find tended to be mixed, very often in distinguishable layers or dumps within features (Table 3). Three successive layers in one pit [199] can be related by the presence in them of cross-joining sherds; but otherwise the impression is of repeated episodes of rubbish disposal into opportunistically used pit features.

The Pottery

The Prehistoric pottery assemblage comprises 2,000-odd sherds weighing approx-imately 26kg. The bulk of it, characterized by a suite of flint- and decalcified shell-tempered fabrics recurrently associated with two chronologically diagnostic pottery finishes – rustication, which involves deeply fingering the body of a pot or the application of a rough surface finish, and painting – and several chronologically diagnostic forms, including the ‘onion-shaped jar’, with a rounded shoulder and a flared neck, the pedestal base and the open mouthed convex-sided jar, belongs to a poorly understood and currently unnamed pottery tradition that falls between post Deverel-Rimbury, dated at its latest to the very beginning of the Iron Age (c.700 bc), and the saucepan pot continuum, dated at its earliest to the beginning of the Middle Iron Age (c.400 bc), i.e. to the Early Iron Age. (Although extra-Kent associations show the forms comprising it to fall between post Deverel-Rimbury proper and saucepan pottery, owing to the coincidence of this period in Britain and on the near continent with the radiocarbon calibration plateau, close calendar dating of it is impossible). In addition, a small number of sherds, mostly from Early Iron Age-dated contexts, are attributable to the post Deverel-Rimbury (Late Bronze Age or Late Bronze Age/ Early Iron Age) tradition (less than 100 sherds), and another eight to the Middle (two only) and Late Iron Ages.

This report focuses on the Early Iron Age material. Discussed first is the structure of the assemblage. There are indications of localized middening on site, which has produced a series of pottery groups that are closed and yet incomplete, which has major implications in terms of what we can and cannot usefully say about the material. Typology can be profitably discussed, use context cannot. The composition of the assemblage in terms of the forms and fabrics comprising it is considered. This part of Kent lacks published Early Iron Age pottery groups and the 45-odd pots from the Manor Farm pub site go some way towards filling this gap. Thirdly, and perhaps most importantly, there are issues of chronology to be considered. As noted in the Introduction, owing to the absence from large parts of the county of identifiable Middle Iron Age pottery there has been a tendency locally to push Early Iron Age pottery traditions forward in time, to treat them as a ‘missing link’ as it were. But this assemblage is demonstrably not a missing link. Of interest finally are the regional relationships of the assemblage, which though including Sussex, Essex and the near Continent, are shown to be narrower than those of the preceding, post Deverel-Rimbury pottery tradition, but significantly wider than those of later, saucepan pot traditions. The range of possible uses to which the assemblage might have been put is not discussed but can be inferred by the reader from details of pot type, size and relationship present throughout the text and illustrations.

Pottery deposition/ assemblage integrity

The accuracy of any pottery study depends in large part on the integrity of the assemblage, whether it is closed or incorporates pottery of many different periods, whether it is representative of the context or site from which it comes as a whole or comprises an unrepresentative sample only. In these respects the Manor Farm pub group is certainly promising. Where paralleled, the bulk of the assemblage belongs to a single – if poorly understood – pottery tradition, and there is no reason to believe it has been much disturbed since the Early Iron Age, and while there are a number of possibly earlier sherds, no single context can be reliably dated to this period. Moreover, for a site of the Manor Farm pub’s modest size, it incorporates both significant numbers of sherds and a wide range of forms and fabrics. But accurate assessment also requires that we understand something of the way the assemblage was deposited in the first place and in this respect the record is more ambiguous.

Table 3, which quantifies the fabrics comprising several of the larger context assemblages, shows quite different suites of fabrics to have come from different features. Pit 199 for example yielded a more restricted range of fabrics than pit 192 and posthole 428, while pit 192 and posthole 428 yielded a similar range of fabrics but in very different proportions. This is what one would expect of a functionally and/or chronologically determined distribution – the implication being that individual fabrics, which on site were utilized for different vessel types (Table 4), had either different roles or are of different dates. On the other hand, the assemblage as a whole displays characteristics that elsewhere have been taken as signs of redeposition (cf. Davey and Macpherson Grant 1996, 67; Seager Thomas 2008, 46; 2010, 22), the frequent burning of pots (including sherds from a minimum of 31 of the 45 reconstructable pots), their mixing with finds of other categories (Table 3) and the small numbers of sherds by which individual pot are represented – despite the 100 percent sampling of surviving feature fills – as well as the presence of sherds from the same pot in different features (pit 294 and postholes 347 and 428; Fig. 3). It is also notable that the ratio of fine to coarse fabrics is lower than that of distinguishable fine to coarse ware pots, a likely consequence of the disturbance of fragile fine wares.

So, what are we to make of the assemblage? While there is good evidence for the deposition of disused pottery prior to its burial in the features from which it was recovered, it remains our view that this was not centralized, that there was no single homogenizing deposit, and that therefore individual context assemblages might well reflect functionally and/or chronologically discrete episodes of activity. The implications of this for our understanding of the assemblage are two-fold. On the one hand, we can probably assume a close functional and/or chronological relationship between the sherds comprising any single context assemblage. Even, for example, where large sherds belonging to apparently different pottery traditions are found together (as in pits 294 and 322), we have to assume they were in use more or less concurrently. On the other hand, we cannot know what escaped redeposition, and it is probably safest therefore to assume that these same context assemblages are only incompletely representative of pottery and pottery-using activity on site.

Pottery Typology (Figs 5-10 showing Pots 1-45)

Form: owing to the variable quality of the site’s coarse wares (roughly finished, coarsely tempered, mostly thick-bodied pots) and the poor preservation and incompleteness of some context assemblages, it is not possible to reconstruct every pot form represented in the assemblage with complete confidence. For the most part, however, enough of each pot survives both to distinguish it as an individual pot and to place it within a broad typological grouping, which can be paralleled elsewhere (Table 4), and thereby to situate the assemblage as a whole. In terms of distinguishable pots, the ratio of coarse to fine wares (burnished, finely tempered, mostly thin bodied pots) is about 4:1. The assemblage as a whole, however, is dominated by the upright jar form and incorporates very few bowl types.

Most representative of the assemblage is a coarse ware form shaped some-thing like a wonky flowerpot, individual examples of which have, or appear to have, straight (pots 4 and 27), convex (pots 1, 19 and 21), or straight and convex sides (e.g. pots 3 and 5), and are either open or closed, the latter profile occasionally displaying a pronounced hooked rim (pots 1 and 3). This form has mostly very roughly executed rounded (pots 4 and 21), bead (pot 44), plain-squared (pots 9 and 25), cabled (pots 3 and 5) and fingertip-impressed rims (pot 6). Rim diameters range in size from about 15cm (pot 19) to 32cm (pot 21).

Possibly also belonging to a ‘flower pot’ is a tiny sherd that best reconstructs as a bord festonné or festooned rim (from pit 224 – not illustrated), a rare form in Britain, in which the lip of the rim, which is wavy or cog-like, hangs down in front of the pot.

Otherwise coarse wares on site are represented by three essentially different but nonetheless overlapping shouldered jar variants (once again exact morphological characterization is impossible). These range from more or less tripartite, with an upright (pots 24 and 28) or out-turned neck (pots 15 and 35), through strongly bipartite, with a long (pot 30) or a very short shoulder (pots 16, 20 and 32). Lying between this and the previous form is a chronologically important variant with a short upright/vestigial neck (pots 10 and 13), to what are really variants of the convex-sided jar with a weakly-moulded shoulder/neck, which are either open (pots 1, 29 and 40) or closed (pot 34). The tripartite variants have plain squared (pots 15 and 24), cabled (pot 28) and cabled, internally and externally expanded ‘hammerhead’ rims (pot 35) (pot 28 is also notable for its markedly angular shoulder), the long shouldered bipartite, ‘hammerhead’ rims (pots 30 and, probably, 8), the short necked bipartite, slightly expanded, rounded (pot 32) and/ or internally bevelled rims (pots 16 and 20), and the weakly shouldered, internally bevelled (pots 1 and 29), slightly externally expanded (pot 34 and 43) and simple rounded rims (pot 40). Another very ‘wonky’ pot, which we reckon bipartite (note the slight out-turn at the bottom of the reconstruction drawing), but which one colleague prefers to see reconstructed as upright (pot 22) (S. Hamilton pers. comm.), has an internally bevelled rim.

With the exception of pot 34, a weakly shouldered jar, the smallest of these, perhaps predictably, are the complicated tripartite forms, with rim diameters ranging from 14cm (pot 28) to 24 cm (pot 15). The largest variants, the bipartite and weakly shouldered, have rim diameters ranging from 22 (pot 20) to over 30 and most probably in excess of 40cm (pots 38 and 43). This size distribution contrasts with that of the simplest form on site, the ‘flower pot’.

Amongst the reconstructable fine ware forms, finally, there are five jars, three bases – one certainly from a jar – and three bowls. Of the former, two recall coarse wares from the site – pot 37, a tripartite jar with an out-turned neck and rounded rim, and pot 11, a long shouldered bipartite jar with a ‘hammerhead’ rim. Both have rim diameters of about 22cm. Two are smaller (18cm diameter) onion-shaped jars. These comprise a bulbous body sherd with painted decoration (pot 26), which would very likely have had pedestal bases like that represented in pot 41, and a rounded shoulder with a pronounced flared neck (pot 45). The pots to which the other fine ware bases, pot 36 and an un-illustrated foot ring found with pots 28-30, belong cannot be reconstructed but the base forms themselves are to be expected of a fine ware assemblage of this sort. The last jar, by contrast, cannot easily be placed (pot 39). Consisting of a massively expanded rim, an upright neck, and a rounded body (which does not join the neck), it is currently without parallel in Britain or on the near continent and, accordingly, it is impossible to be confident when suggesting a reconstruction.

The bowls – all of quite large size (20-22cm diameter) – are round shouldered with an upright or slightly flared rim (pot 14), bipartite with an upright shoulder/ neck and, unusually for a bowl, an internally expanded rim (pot 17), and, probably, tripartite with a sharply angular shoulder (the pot is represented by a single tiny sherd with neither rim nor neck but the upper part of the shoulder bends out slightly – pot 31).

Finish: three or four coarse ware finishes are distinguishable in the Manor Farm pub assemblage – rough burnish (RB on the reconstruction drawings) (pots 16 and 30), simple fingering (F in the figures) (pots 4, 15 and 43), and deliberate roughening (known locally as rustication – R), using either the fingers (pots 2 and 27), some kind of coarse wipe or a comb (C) (pot 42), and/ or applied clay slurry (AR) (pot 38). (On the reconstruction drawings, where incidental fingering cannot be distinguished with certainty from deliberate fingering, it is marked RF – e.g. pot 21). Typically on ‘flower pots’ roughening/ rustication extends to the rim (pot 19), while on shouldered pots it extends to the shoulder angle, the shoulder itself being roughly burnished (pots 20 and, possibly, 16).

The fine wares here are by definition burnished, sometimes highly. Pot 14, for example, has a very high burnish both inside and outside, while both pots 11 and 17 retain a high burnish on the exterior. In addition pot 26 was painted with three wide horizontal bands of red, probably hematite paint, and in the gaps between these, which remained unoxidized, marginal lines, zigzags and multiple chevrons, in what was probably white but is now orange paint. Two other unillustrated sherds in a rare sandy fabric (RFFQ), one associated with pots 9, 10, 11 and 13, also have maroon hematite coats, while pots 11 and 39, which incorporate visible siderite nodules may have been intended to oxidize to a similar vivid red.

Pottery Fabrics: initially full fabric analysis of the assemblage seemed a good idea, the apparent integrity of the assemblage holding out the rare promise of an unambiguous fabric series for the period against which pottery from the more usual, chronologically mixed assemblages could be usefully compared. Indeed most of the bigger context groups are divisible into eight and ten, mostly clearly divisible fabrics, with an overall ratio of coarse to fine wares of about 11:2, noticeably different from that of distinguishable coarse to fine ware pots (Table 3 and Table 5). By the time the analysis was finished, however, most of these had resolved themselves into a continuum of fine to coarse, mostly flint-tempered fabrics typical of the earlier first millennium bc locally, with its usual interpretatively troubling overlaps with – from the perspective of the Early Iron Age koine – both earlier, Late Bronze Age post Deverel-Rimbury and much later, Iron Age traditions. Accordingly, for the assemblage as a whole, details of this analysis have been reserved for the archive (Seager Thomas 2012b). It’s worth mentioning a handful of features of it, however, which are apparently diagnostic of Early Iron Age traditions locally, and, in some cases, further afield.

Overall the assemblage is dominated by sherds the surfaces of which are oxidized red, often vividly so. In large part this is attributable to secondary firing, in many cases the red colouration continuing across the broken edges of sherds, which have dark grey unoxidized cores. This vividness of colour, which is most striking, would only have been possible had the clay comprising them been iron-rich, a view confirmed for the fabrics of some pots by the visible presence in them of small siderite concretions (pots 11, 35 and 39). In addition, in fabrics for which natural clay matrices are distinguishable from the inclusions deliberately added to them, these are sandy, a feature, which though a function of clay source(s), is nonetheless recurrent in late post Deverel-Rimbury and Early Iron Age pottery from South-East England as a whole (e.g. O’Connor 1986, 61-2; Seager Thomas 2001, 36; 2008, 41). Sherds give a false impression of friability, while original surfaces that are unburnished and free of deliberately added inclusions have a texture similar to that of fine grade sandpaper. Finally, two of the better defined fabric types stand out, FCF, tempered with fine and coarse burnt flint (as opposed to fine to coarse burnt flint), and DSF, a coarse decalcified shelly fabric with widely varying quantities of fine and/or coarse burnt flint, used only in ‘flowerpots’ (pots 4 and 5). When found together, these features, which reoccur through the assemblage, can probably be taken as characteristic of Early Iron Age pottery locally.

Pottery dating

Unusually in South-East England, the dating of much Kent earlier first-millennium bc pottery remains open, due in large part to the absence from the county of significant assemblages belonging to two key ‘marker’ traditions, very late post Deverel-Rimbury, characterized in particular by angular, often highly decorated pots, associated with the very beginning of the Iron Age, c.700 bc (Needham 1996, 134-7), and the saucepan pot continuum, characterized in particular by the saucepan pot, which dates at its earliest from the beginning of the MIA, c.400 cal bc (Orton and Cunliffe 1984, fig. 5). Instead, what we have – at least in the east of the county – is France and the Low Countries’ ‘Marnian’ or early La Tène pottery (e.g. Hawkes 1940; Macpherson-Grant 1989), which on the continent emerges out of their equivalent angular, decorated horizon, but whose end point cannot yet be closely correlated with any clearly established British Iron Age pottery tradition. In the absence of precise radiocarbon dating (which is not possible for the period owing to its coincidence with the earlier first millennium bc radiocarbon plateau – Pearson and Stuiver 1986, fig. 1a), the placing and dating of middle and west Kent assemblages, such as this one, which are neither strictly Marnian/ La Tène or any other currently clearly and completely defined tradition, rests on analogy with a range of very different, and sometimes themselves imprecisely dated assemblages.

In Kent the assemblage as a whole is best paralleled by groups from Barham Downs and Highstead, and in ‘Marnian’/early La Tène assemblages from the east of the county (Table 4), and it can be assumed therefore to belong to a related tradition. Of the forms – and fabrics – comprising it, however, a handful are of some longevity, complicating the chronological attribution of the assemblage. In Sussex, for example, rustication, one of the present tradition’s principal diagnostic traits, is associated with its latest post Deverel-Rimbury (Seager Thomas 2008, 41), while at Holland’s Oss Ussen, it first appeared, albeit in small quantities, in its earliest (Late Bronze Age/ Early Iron Age) phases growing in significance through its Early and Early to Middle Iron Age (our Early Iron Age) (Van den Broeke 1987, table 5). Likewise it is occasionally present in later French post Deverel-Rimbury-like assemblages, although it is also widely associated with immediately succeeding traditions (e.g. at Coquelles and Frethun outside Calais – Blancquaert 1989, figs 5 and 12). Similarly the hooked rim convex-sided jar is associated in particular with early post Deverel-Rimbury traditions (e.g. Bradley and Ellison 1975) and ‘flowerpots’ generally and forms similar to some of the site’s weakly tripartite shouldered jar variants, with both post Deverel-Rimbury (Seager Thomas 2008, fig. 9) and later, saucepan pottery (e.g. at Norton in East Sussex, and Little Waltham in Essex – Drury 1978, figs 44.59 and 48.208; Seager Thomas 2005, figs 15 and 16). Indeed, examples of the latter (pots 28 and 37), tend – at least by some specialists – exclusively to be associated with later post Deverel-Rimbury traditions.

To exactly what period then does this assemblage belong? While we cannot categorically exclude the possibility that it incorporates pottery belonging to the Late Bronze Age, the Early Iron Age and the Middle Iron Age, there is strong evidence that it does not, but rather that it falls between the two extremes.

Firstly, one of the post Deverel-Rimbury tradition’s principal defining characteristics is its thin-bodied coarse wares, something largely absent from the Manor Farm pub material; though it incorporates some post Deverel-Rimbury traits, it is not a post Deverel-Rimbury assemblage. Secondly, while there are overlaps with other pottery traditions, the tradition represented by the assemblage as a whole has occasionally been found isolated from other ceramics – notably at sites like Barham Downs and Highstead (Table 4), at the latter of which it was stratified above an earlier post Deverel-Rimbury assemblage (Couldrey 2007, figs 56-62). If there are, as there appear to be, late post Deverel-Rimbury sherds in it, and our inferences above regarding pottery deposition on site are correct, this might suggest continuity between the two traditions and a date within the Early Iron Age soon after the demise of post Deverel-Rimbury proper. Thirdly, it incorporates a number of features – such as the bipartite shouldered bowl (pot 17), deliberated roughening/ rustication (pot 28), the angular bowl (pot 31) and the ‘hammerhead’ rim (pot 33), which, in surrounding regions where Middle Iron Age pottery is distinguishable, are present in earlier Iron Age but not Middle Iron Age assemblages (sites where both Early and Middle Iron Age assemblages occur side by side include, for example, Hawk’s Hill in Surrey, and Carne’s Seat, Park Brow and Slonk Hill in Sussex – Cunliffe 1965; Hamilton 1986; Hartridge 1978; Wolseley and Smith 1924; Wolseley et al. 1927). Finally, Kent forms and fabrics, which are associated with the Middle Iron Age such as the Wealden S-shaped jar and glauconitic wares (Champion 2007, 297; Couldrey 1984, 38-40 and fig. 15; Seager Thomas 2010, 6 and 15) are conspicuous for their scarcity. Two sherds only fall into this group, both from the ditch.

Other Finds from the Site

In order fully to contextualize the pottery assemblage, we must say a few words about the finds with which it was associated – if only to show their relative proportions. In terms of numbers, the most common type of find was pottery. Next most common was burnt flint, of which 903 pieces, weighing 53kg were found, most from just ten features, four in or close to the roundhouse and the rest spread widely across the site (Bishop 2012b, 226). As noted, in only one case was it found in functional situ, but it is clear that several flint burnings were represented.

Compared to pottery and burnt flint, all other finds categories were poorly represented. There were just 105 pieces of struck flint pieces. Most of these comprised crudely struck flakes and minimally reduced cores typical of late prehistoric flint working traditions in Britain (Young and Humphrey 1999). Their principal interest here is their Iron Age associations, which suggest that flint working and use, albeit of an ad hoc nature, continued into that period. Struck flint too came from all over the site, although one of the largest individual assemblages was from pit 199, just outside the house (Bishop 2012a, 123-4).

Charred remains included unidentified charcoal, barley (Hordeum sp.) and wheat (Triticum sp.) grains, wheat chaff (Triticum spelta and Triticum sp.), weed seeds, a single hazelnut shell and some indeterminate fruit stones. These were most abundant in pit 199, which contained several layers that were thick with charred material; which otherwise was poorly represented. A less than 2g flot from the ‘Polynesian oven’ yielded between 50 and 250 pieces of charcoal of less than 2mm in size (Le Hégerat 2012).

Animal remains comprised 50 bones and teeth recovered by hand, with a few more added from sample residues. The animals identified were mostly adults and included cattle, sheep and/or goat, horse and pig. Amongst these were the teeth of a horse of more than 19 years old, indicating a notable level of care, a group of semi-articulated cattle-sized vertebrae (from pit 199 again) and a sheep or goat humerus with defleshing cuts. The lack of small and younger animals, the small number of bones overall, and a disproportionately high proportion of head parts and teeth, is attributed to poor preservation conditions on site and is not considered interpretatively important (Reilly 2012).

The Site’s Place in the World

Perusing Table 4 with its references to typological parallels from sites in West Sussex, the other side of the Thames estuary and in France and the Low Countries, the reader would be forgiven for thinking the pottery from the Manor Farm pub site belonged to far-ranging cultural continuum. Up to a point of course, this is true. There are good individual parallels; but there are very few group parallels (none at all beyond Kent), while key fine ware types present at Manor Farm are more or less unknown off-site and visa versa. Pot 39 is a good case in point; currently our best guess is that it is aping something Mediterranean, but this is so far fetched that independent development seems much more likely. The same is true of the site’s fabrics. The present writer is unable to comment on the wider distribution of fabric FCF, but the other Manor Farm pub fabric to stand out, DSF, though paralleled locally (in an unpublished assemblage from the Isle of Grain) and on the Essex coast (Wymer and Brown 1995, 83), is untypical of Kent sites further east, while grog-tempering, present to the east of the county (in unpublished assemblages from Castle Hill, Folkestone, Hawkinge and Highstead¹) and common across the channel (e.g. Bailleul and Ham – Barbet and Buchez 2005, 34; Hurtrelle 1989), and glauconitic fabrics, common in East Sussex through both the Early and Middle Iron Ages (Seager Thomas 2005, table 2; 2008, 41), are represented in the Manor Farm pub assemblage by a handful of sherds only, most of them arguably belonging to later traditions. Although not a ceramic island, therefore, the assemblage and the tradition to which it belongs stand out regionally. This should be contrasted, on the one hand, with preceding post Deverel-Rimbury traditions, which were present more completely over a wide area including much of southern Britain and the near continent (e.g. Burgess 1987, fig. 4), and, on the other, with the Middle Iron Age saucepan pot continuum, which although interconnected regionally (Morris 1994, figs 3 and 4; Seager Thomas 2010, 21), is largely absent from Kent (Champion 2007, 297) and completely absent from the rest of Europe.

conclusions

To summarize: we have an Iron Age enclosure with a roundhouse, albeit rather a small one, a rectangular structure, a fence and some storage pits, about which were carried out a range of everyday activities. Culturally the site had connections across South-East England and beyond, which were adapted locally. Except perhaps for the absence of quern, a typical Early Iron Age site, showing the Early Iron Age in this part of Kent to have been similar to the same period elsewhere in the South-East. As such the pottery assemblage from it is peculiarly useful, since we can take it as representative, and so use it as a yardstick against which to compare assemblages from other sites and periods.

There are always alternative interpretations in archaeology and this is no less true of the Manor Farm pub site than anywhere else. Excavations upslope of the enclosure ditches were more limited than downslope of it, and may not accurately reflect the distribution of features there (Barrowman 2012, 65). Indeed a glauconitic sherd from the earlier of the two might just be from a ‘Wealden’ jar, thus placing it and the postulated enclosure in the Middle Iron Age, tens if not hundreds of years later than most of the dated pits and postholes. The house may not be a house at all. It certainly was not recognized as such in the field. On analogy with Wessex structured deposits comprising ‘individual layers dominated by a single category of find’ (Hill 1994, 4), the group of articulated vertebrae and some pottery groups might likewise be interpreted as structured deposits (Barrowman 2012, 65) and therefore not reliably representative of everyday life during the period. The same could apply if the occupants of the site had a peculiarly conservative taste in pottery, and/or carefully curated pottery belonging to earlier traditions. On reflection we have discarded all these possibilities, but as is always the case in archaeology, we will certainly have got some things wrong, and we might have to reconsider.

In the final analysis however none of this is important. As noted in the introduction, so long as we present our data, and present it honestly, the reader can make up his or her own mind about it. However it is interpreted, the Manor Farm pub pottery is important, and with other seminal pottery groups from the county, such as those from the Bridge bypass excavations, Bigberry, Monkton Court Farm and Worth, and will be referred to again and again in the literature by specialists inside and outside the county. Let us hope that the many unpublished assemblages referred to above (and most of them are still unpublished) join it soon.

acknowledgements

Pre-Construct Archaeology would like to thank Walsingham Planning and the Whitbread Group plc group for commissioning these important excavations, and all those involved in the excavations and subsequent post-excavation work for their invaluable contributions.

bibliography

Barbet, P. and Buchez, N., 2005, ‘Les habitats protohistoriques de Ham “Le Bois à Cailloux”’, Revue Archéologique de Picardie, 1/2, 25–50.

Barrowman, S., 2012, An Assessment of an Archaeological Evaluation, Excavation and Watching Brief at the Manor Farm Public House, High Street, Rainham, Gillingham, Kent, ME8 7JE, unpubl. report: Pre-Construct Archaeology Ltd.

Bishop, B., 2012a, ‘The lithic assessment’, in S. Barrowman.

Bishop, B., 2012b, ‘The burnt stone assessment’, in S. Barrowman.

Blancquaert, G., 1998, ‘L’âge du fer à Coquelles et Fréthun (Pas-de-Calais)’, Revue du Nord, 328, 109-37.

Bourgeois, I., Leman-Delerive, G. and Révillion, S., 2003, ‘Houplin-Ancoisne: un aménage-ment d’accès à la rivière pendant l’époque gauloise?’, Revue du Nord, 85, 51-88.

Bradley, R. and Ellison, A., 1975, Rams Hill: a Bronze Age Defended Enclosure and its Landscape, 99-118, Oxford: BAR.

Broeke, P. van den, 1987a, ‘De dateringsmiddelen voor de ijzertijd van Zuid-Nederland,’ in W. van der Sanden and P van den Broeke (eds), Getekendzand. Tien Jaar Archaeologisch Onderzoek in Oss-Ussen, Waalre, 22-43, Bijdragen tot de Studie van het Brabantse heem, 31.

Burgess, C., 1987, ‘Les rapports entre la France et La Grande-Bretagne pendant l’Âge du Bronze: problèmes de poterie et d’habitats’, in J-C. Blanchet (ed.), Les Relations entre le Continent et Les Iles Britanniques a l’Âge du Bronze, 307-18, Actes du Colloque de Bronze de Lille. Amiens: Revue Archéologique de Picardie.

Champion, T., 2007, ‘Settlement in Kent from 1500 to 300 BC’, in C. Haselgrove and R. Pope (eds), The Earlier Iron Age in Britain and the Near Continent, 239-305, Oxford: Oxbow.

Couldrey, P., 1984, ‘The Iron Age pottery’, in B. Philp, Excavations in the Darent Valley, Kent, 30-70, Dover: KARU.

Couldrey, P., 2007, ‘The Late Bronze Age/Early Iron Age pottery’, in P. Bennett, P. Couldrey and N. Macpherson-Grant, Highstead, near Chislet, Kent: Excavations 1975-1977, 101-175, Canterbury: CAT.

Cunliffe, B., 1965, ‘The Pottery’, in F. Hastings, ‘Excavation of an Iron Age farmstead at Hawk’s Hill, Leatherhead’, Surrey Archaeological Collections, 62, 13-39.

Cunliffe, B., 2005, Iron Age Communities in Britain. An Account of England, Scotland and Wales from the Seventh Century BC until the Roman Conquest, 4th edition, Routledge: Abingdon.

Cunliffe, B. and Orton, C., 1984, ‘Radiocarbon age assessment’, in B. Cunliffe, Danebury: an Iron Age hillfort in Hampshire. Volume 1. The excavations, 1969-1978: the excavations, 180-90, CBA Research Report, 52, London.

Davey, M. and Macpherson-Grant, N., 1996, ‘The ceramics from the Whitfield-Eastry Bypass, Site 2’, Canterbury’s Archaeology 1995-96, 67-9.

Doorselaer, A. Van, 1989, ‘Un site fortifié de l’âge du Fer avec enclos cultuel à Kooigem, commune de Courtrai (Flandre Occidentale)’, in M. Otte and M. Ulrix-Closset (eds), La Civilisation de Hallstatt, 357-66, Rencontre Internationale Liege.

Green, C., 1980, ‘Handmade pottery and society in Late Iron Age and Roman East Sussex’, Sussex Archaeological Collections, 118, 69-86.

Hamilton, S., 1977, ‘The Iron Age pottery’, in M. Bell, ‘Excavations at Bishopstone’, Sussex Archaeological Collections, 115, 83-117.

Hamilton, S., 1986, ‘Late Bronze Age and Iron Age pottery’, in R Holgate, ‘Excavations at the late prehistoric and Romano-British enclosure complex at Carne’s Seat, Goodwood, West Sussex, 1984’, Sussex Archaeological Collections, 124, 43-4.

Hamilton, S., 2004, ‘Early first millennium pottery of the West Sussex Coastal Plain’, in C. Place, Excavations at Ford Airfield, Yapton, West Sussex, 1999, 18-38, Kings Lynn: Heritage.

Hamilton, S. and Seager Thomas, M., 2005, ‘The nature and importance of the Iwade earlier prehistoric pottery’, in B. Bishop and M. Bagwell, Iwade: the Occupation of a North Kent Village from the Mesolithic to the Medieval Period, 20-38, Pre-Construct Archaeology Monograph, 3, London.

Hawkes, C., 1940, ‘The Marnian pottery and La Tène 1 brooch from Worth, Kent’, Antiquaries Journal, 20, 117-21.

Hill, J., 1994, ‘Why we should not take the data from Iron Age settlements for granted: recent studies of intra-settlement patterning’, in A. Fitzpatrick and E. Morris (eds), The Iron Age in Wessex: Recent Work, 4-9, Salisbury: Trust for Wessex Archaeology.

Hodson, F., 1962, ‘Some pottery from Eastbourne, the “Marnians” and the pre-Roman Iron Age in southern England’, PPS, 7, 140-55.

Hartridge, R., 1978, ‘Excavations at the prehistoric and Romano-British site on Slonk Hill, Shoreham, Sussex’, Sussex Archaeological Collections, 116, 69-141.

Hurtrelle, J., Monchy, E., Roger, S., Rossignol, P. and Villes, A., 1989, Les débuts du second âge du fer dans le Nord de la France, Les Dossiers de Gauheria, 1.

Kinnes, I., Cameron, F., Trow, S. and Thomson, D., 1998, Excavations at Cliffe, Kent, British Museum Occasional Paper, 69. London.

Lardy, J-M., 1983, ‘Les Gauloise du Val d’Oise’, Les Dossiers Histoire et Archaeologie, 76, 34-45.

Le Hégerat, K., 2012, ‘The charred plant macrofossil and wood charcoal assessment report’, in S. Barrowman.

Macpherson-Grant, N., 1980, ‘Archaeological work along the A2: 1966-1974’, Archaeologia Cantiana, 96, 133-83.

Macpherson-Grant, N., 1989, ‘The pottery from the 1987-1989 Channel Tunnel excavations’, Canterbury’s Archaeology 1988-89, 60-3.

Macpherson-Grant, N., 1991, ‘A re-appraisal of prehistoric pottery from Canterbury’, Canterbury’s Archaeology 1990-91, 38-48.

Macpherson-Grant, N., 1992, ‘A review of Late Bronze Age pottery from east Kent’, Canterbury’s Archaeology 1991-92, 55-63, CAT: Canterbury.

Macpherson-Grant, N., 1994, ‘The Pottery’, in D. Perkins, N. Macpherson-Grant and E. Healey, ‘Monkton Court Farm evaluation, 1992’, Archaeologia Cantiana, 114, 248-88.

Morris, E., 1994, ‘Production and distribution of pottery and salt in Iron Age Britain: a review’, PPS, 60, 371-93.

Needham, S., 1996, ‘Chronology and periodisation in the British Bronze Age’, Acta Archaeologica, 67, 121-40.

O’Connell, M., 1986, Petters Sports Field, Egham. Excavations of a Late Bronze Age/ Early Iron Age Site, Research Volume of the Surrey Archaeological Society, 10, Guildford.

Parfitt, K., 1985, ‘Some Iron Age sites in the Deal area’, Kent Archaeological Review, 79, 206-19.

Pearson, G. and Stuiver, M., 1986, ‘High-precision calibration of the radiocarbon time scale, 500-2500 BC’, Radiocarbon, 28 (2B), 839-62.

Reilly, K., 2012, ‘Assessment of animal bone’, in S. Barrowman.

Seager Thomas, M., 2001, ‘Two early first millennium BC wells at Selsey, West Sussex and their wider significance’, Antiquaries Journal, 81, 15-51.

Seager Thomas, M., 2005, ‘Understanding Iron Age Norton’, Sussex Archaeological Collections, 143, 83-117.

Seager Thomas, M., 2006, ‘The Iron Age pottery’, in T. Carew, B. Bishop, F. Meddens and V. Ridgeway, Unlocking the Landscape: Archaeological Excavations at Ashford Prison, Middlesex, 56-68, Pre-Construct Archaeology monograph, 5, London.

Seager Thomas, M., 2008, ‘From potsherds, to people. Sussex prehistoric pottery: Collared Urns to post Deverel-Rimbury’, Sussex Archaeological Collections, 146, 19-51.

Seager Thomas, M., 2010, ‘A re-contextualization of the prehistoric pottery from the Surrey hillforts of Hascombe, Holmbury and Anstiebury’, Surrey Archaeological Collections, 95, 1-33.

Seager Thomas, M., 2012a, ‘Assessment of the Roman pottery’, in S. Barrowman.

Seager Thomas, M., 2012b, ‘Pottery data table’, in S. Barrowman.

Thompson, I., 1982, Grog-tempered ‘Belgic’ Pottery of South-Eastern England, BAR, 108, Oxford.

Wolesley G. and Smith R., 1924, ‘Discoveries near Cissbury’, Antiquaries Journal, 4, 347-59.

Wolseley, G., Smith, R. and Hawley, W., 1927, ‘Prehistoric and Roman remains on Park Brow’, Archaeologia, 76, 1-40.

Wymer, J. and Brown, N., 1995, Excavations at North Shoebury: Settlement and Economy in South-East Essex, 1500BC-AD1500, East Anglian Archaeology, 75, Colchester: Essex County Council.

Young, R. and Humphrey, J., 1999, ‘Flint Use in England after the Bronze Age: time for a re-evaluation?’, PPS, 65, 231-42.

endnotes

¹ The published material from Highstead did not incorporate grog – Couldrey 2007.

Fig. 1 The location of the site.

Fig. 2 The excavated area.

Fig. 3 The possible roundhouse and rectangular structure. The shallow depth of the postholes to the north of the post ring (Table 1), and the gap in it to the north-west, cast some doubt upon its identification as a roundhouse. Probably, however, this – upslope – part of the site was more deeply truncated than the rest.

Table 1. Postholes and other features directly related to the possible roundhouse and the rectangular structure

Cut	Fill	Sides	Base	Size (cm)	Depth (cm)*
Roundhouse
219	216-8	Vertical; straight	Flat	30 x 28	33
241	240	Vertical	Flat; round break	48 x 42	35
249	248, 252, 257	Steep	Flat; gradual break	53 x 46	28
281	280, 267, 273	Steep	Flat; curved break	52 x 43	23
292	288, 320	Steep	Stepped; gradual break	60 x 60	40
333	332	Vertical	Flat; sharp break	26 x 26	8
361	350-1, 358, 360	Steep to vertical	Flat; sharp break	80 x 67	32
379	378	Steep, concave	Flat; gradual break	38 x 20	3
399	398	Steep; concave	Concave; gradual break	30 x 30	9
418	417	Near vertical; straight	Flat; sharp break	40 x 40	33
Rectangular setting in roundhouse
270	263, 266	Steep	Flat	40 x 32	16
355	354	Very steep	Concave; gradual break	27 x 27	9
416	415	Near vertical; straight	Flat; sharp break	30 x 30	16
Rectangular structure
294 (pit)		Steep; slightly irregular	Concave	45 x 44	45
347	345-6	Steep	Flat; sharp break	61 x 55	61
368	367	Vertical	Near flat; sharp break	60 x 49	70
395	394	Steep	Flat; sharp break	55 x 24	52
404	401-3	Near vertical	Very slightly concave	52 x 50	57
428	426-7	Near vertical	Flat	58 x 54	67

*Add 18-57cm+ of developed/agricultural soil machine-stripped prior to excavation.

Fig. 4 Detail of roundhouse and rectangular structure.

Table 2. Selected pits

Cut	Fill(s) from the top down	Sides	Base	Size (cm)	Depth (cm)*
192	190-1, 210	Steep	Flat; concave break	95 x 121	52
199	178, 183-7, 198	Steeply sloping	Flat; sharp break	154 x 190	92
209	207, 208 (clay lining)	Near vertical	Flat; moderate break	72 x 94	8
256	255	Concave	Rounded; no break	56 x 57	14
277	275, 276 (clay lining)	Near vertical	Flat; sharp to gradual	80 x 95	35
279	278	Concave	Concave	40 x 45	10
287	286	Near vertical; slightly concave	Flat; sharp break	70 x 90	56
294	293, 298	Steep; irregular	Concave; sloping	44 x 45	45
301	297	n/a	Concave	42 x 52	9
322	321, 340, 364	Vertical	Flat; sharp break	150 x 290	124
336	334, 335	Vertical	Flat; sharp break	44 x 48	18
374	371, 372, 373 (clay lining)	Steep	Slightly concave	76 x 82	16

*Add 18-57cm+ of developed/agricultural soil machine-stripped prior to excavation.

Table 3. The fabric quantification and artefactual associations of the Early Iron Age pottery

Cut	Fill	Fabrics (weight in grams)										Associations
		FF	FMF	FMFS	MF	SMCF	FCF	MCF	CF	DSF	Other
Pit 192	190		149		254	271	577	841	86	799	81	burnt flint, cereal grain, charcoal, daub, struck flint
	191	15	14		169		72	27		341		cereal grains, charcoal
Pit 199	183	75			19			51				burnt clay, burnt flint, cereal grains, charcoal
	184	17			37			123	48			burnt clay, burnt flint, cereal grains, charcoal, struck flint
	185	30			32			112	91		42	burnt clay, burnt flint, cereal grains, charcoal, struck flint
	186	6						48				burnt clay, charcoal
	187	71	36					275	768		3	bone, burnt flint, cereal grains, charcoal, daub
	197	72	7					110				charcoal, cereal grains and chaff, hazel nut shell
Pit 256	255							635	51			burnt flint, struck flint
Pit 277	275	555	274			3,684	2,185				214	burnt clay, burnt flint, charcoal, struck flint
Pit 294	293	3	282		25		375	151				bone, charcoal
Pit 301	297	2	52		10		111	320				charcoal, struck flint
Pit 322	321		14		354			129			11	burnt flint, charcoal, daub, metal, struck flint
	340							296				burnt flint, charcoal, daub, struck flint
	364	30	44		7		185	136			9	burnt flint, charcoal, daub, struck flint
Posthole 428	426	1	736	145		54	383	29		38		none
	427			29								none

Table 4. The regional context of the Early Iron Age pottery (approximate typological parallels)

Pottery form	Kent	S.E. England	France/Low Countries
‘Flower pots’ – open mouthed 4, 19, 21, 44	Barham Downs 5 Cby Road, Hawkinge 119 Castle Hill 37 Hawkinge Aerodrome 19, 78, 155, 166 Highstead 335	Eastbourne 8, 10 Bishopstone 8, 28	Bailleul 1 Neuville-sur-Escaut 2
Flower pots – closed mouthed (hook rim) 1, 3, 6, 9, 25	Cby Road, Hawkinge 120 Hawkinge Aerodrome 93 Highstead 365, 406 etc Kingsnorth 22, 26	Eastbourne 12, 13 Bishopstone 1, 6	Houplin-Ancoisne 20.6 Kooigem 9 Neuville-sur-Escaut 17
Festooned rim	Cby Road, Hawkinge 74		Bailleul 4 Ham 381.1 Kooigem 20
Tri-partite shouldered jar – upright or flared neck 15, 24, 28, 37	Hawkinge A’drome 104 Highstead 300, 495 Kingsnorth 25	Slonk Hill (?)179	Kooigem 16
Tri-partite shouldered jar – cabled (or fingertip impressed) ‘hammerhead’ rim 35	Castle Hill Hawkinge A’drome 101 Iwade 21 Kingsnorth 14, 23	Hawk’s Hill 8.31
Bi-partite shouldered jar – short/ vestigial neck 10, 13	Cby Road, Hawkinge 143 Hawkinge A’drome 164, 177 Worth 3-5		Bailleul 10 Frethun 52bis.10, (?)38.1 Kooigem (?)10
Bipartite (short) shouldered jar – plain or bevelled rim 16, 20, 32	Castle Hill 38, 59 Deal 37, 41 Hawkinge A’drome 32, 47 Highstead 474	Bishopstone 31 North Shoebury 97, 123 Slonk Hill 57	Bailleul 9 Frethun 32bis.4
Bipartite (short or long) shouldered jar – ‘hammerhead’ rim 8, 11, 18, 30, 33	Barham Downs 10 Castle Hill Deal 39 Hawkinge A’drome 1 Highstead 373, 429, (?)451 Iwade 11	Bishopstone 11 North Shoebury 121, 124	Frethun (?)38.1, 52bis.19 Houplin-Ancoisne 18.2 Kooigem 18
Weakly shouldered jar – open 1, 29, 40, 43	Cby Road, Hawkinge 2, 20 Highstead 372, 502 Hawkinge A’drome 2 Kingsnorth 13	Park Brow 12	Houplin-Ancoisne (?)21.3 Neuville-sur-Escaut 6, 9
Pottery form	Kent	S.E. England	France/Low Countries
Weakly shouldered jar – closed 34	Cby Road, Hawkinge 132 Highstead 400	North Shoebury 94	Kooigem 4
Rustication – clay spatter 38	Cby Road, Hawkinge 126 Castle Hill 59 Deal 37, 41 Dolland’s Moor Hawkinge A’drome 32, 51 Highstead 388, 454, 456 Kingsnorth 19	Angmering	Frethun Houplin-Ancoisne Oss Ussen
Rustication – finger gooved 20, 27	Cby Road, Hawkinge 2 Castle Hill 37 Highstead 365
Rustication – combed 42	Dollands Moor Hawkinge A’drome 29, 47 Worth 5	Patcham-Fawcett	Bailleul 11 Frethun 10.17 Houplin-Ancoisne 22.8–10 Neuville-sur-Escaut 2 Oss Ussen
‘Onion shaped’ jar 26, 45	Barham Downs 8 Cby Road, Hawkinge 153 Hawkinge A’drome 176	Eastbourne 1, 5 Ford 47, 52 Slonk Hill 2 Fawley, Hants, context 6042
Bi-partite bowl – upright neck/ flat, internally expanded rim 17	Cliffe 91		Ham 381.11
Round shouldered bowl – upright neck/ simple rim 14	Highstead 461	North Shoebury 87	Genainville
(?) Angular bi- or tri-partite bowl 31	Dolland’s Moor	Hawk’s Hill 12. 50, 51 North Shoebury 82, 99, 104	Frethun 10.1 Houplin-Ancoisne 13.1
Pedestal/ footring base 41	Barham Downs 8, 13 Hawkinge A’drome 176 Highstead 380, 446 Worth 6	Bishopstone 17, 22 etc. Ford 58 North Shoebury 81, 92, 98 etc. Park Brow 8
Painted decoration 26	Barham Downs 8 Castle Hill Dolland’s Moor Highstead 368	Eastbourne 1

Note: the numbers alongside the sites are a shorthand reference to the particular vessel type noted.

References

Angmering (Seager Thomas 2008, 41),

Bailleul (Hurtrelle et al. 1989),

Barham Downs (Macpherson-Grant 1980),

Bishopstone (Hamilton 1977),

Canterbury Road, Hawkinge (Hamilton & Seager Thomas unpubl.),

Castle Hill, Folkstone (Couldrey unpubl.),

Cliffe (Kinnes et al. 1998),

Deal (Parfitt 1985),

Dolland’s Moor (Macpherson-Grant 1989),

Green Lane, Eastbourne (Hodson 1962),

Fawley (Seager Thomas unpubl.),

Ford (Hamilton 2004),

Frethun (Blancquaert 1998),

Genainville (Lardy 1983),

Ham (Barbet and Buchez 2005),

Hawkinge Aerodrome (Seager Thomas & Hamilton unpubl.),

Hawk’s Hill (Cunliffe 1965),

Highstead (Couldrey 2007),

Houplin-Ancoisne (Bourgeoise et al. 2003),

Iwade (Hamilton and, Seager Thomas 2005),

Kingsnorth, Isle of Grain (Seager Thomas unpubl.),

Kooigem (Van Doorselaer 1989),

Neuville-sur-Escaut (Hurtrelle et al. 1989),

North Shoebury (Wymer and Brown 1995),

Oss Ussen (Van den Broeke 1987),

Park Brow (Wolseley and Smith 1924),

Patcham-Fawcett (Seager Thomas 2008, 41),

Worth (Hawkes 1940).

Fig. 5 Pots 1-7.

Fig. 6 Pots 8-11, 13-15.

Fig. 7 Pot 16.

Fig. 8 Pots 17-25.

Fig. 9 Pots 26, 28-38.

Fig. 10 Pots 27, 39-45.

Table 5. The Early Iron Age pottery fabrics

Fabric code	Description	Comments	Illustrated pots
FF	A typical earlier first millennium bc flint-tempered fine ware. 5-7% burnt flint of <1mm with a very few larger fragments. Unquantifiable c.1mm+ siderite nodules.		11, 14, 17, 26, 31, 36, 41, 45
RFFQ	A densely sandy fabric with less than 1% very fine (usually <0.5mm) burnt flint.	Hematite coated. Individual sherds from 2 contexts only – [185] & [207].
FMF	Another typical earlier first millennium bc flint-tempered fine ware. 5-7% burnt flint of between <1 and 1.5 or even 2mm with a few larger fragments.	Similar to FF, occasionally grading up into FCF (e.g. in pot 27). Burnished and roughly finished.	18, (?)27 & 37
FMFS	A densely sandy fabric (up to medium-sized quartz sand) with patchy, 5-10% burnt flint of between <1 and 2mm. Some much larger water-rolled stone (?chert).	Combed	42
MF	A typical earlier first millennium bc flint-tempered medium ware. c. 5% (occasionally as low as 2 and as high as 10%) burnt flint of between <0.5 and 2.5-3mm. Some sherds also incorporate probably rare but unquantifiable c. 1mm+ siderite nodules	Two sherds in this size grade – from [222] & [240] – incorporate abundant glauconite. Grades into MCF.	1, 8, 13, 28, (?)34 & 35
SMCF	A typical earlier first millennium bc flint-tempered coarse ware. c.3% burnt flint of between <0.5% and 4 (and occasionally more) mm.		16, 22, 23, 25, 44
FCF	An unusual mix of 7-10% burnt flint of <0.5-1.5 and >3mm and frequently much larger size (slivers and flakes up to 10mm).	Usually very roughly finished. Occasionally grading down into FMF.	2, 3, 19, 20, 21, (?)27, 29, 32, 43
MCF	Another typical earlier first millennium bc flint-tempered coarse ware. c.5-7% burnt flint of between <0.5% and c.4mm.		6, 9, 10, 15, 24, 30, (?)34 & 35, 38, 40
CF	As MCF but with burnt flint up to 5mm.
DSF	Probably several related fabrics. c.3-10% platy voids (decalcified shell) with a very variable burnt flint fraction ranging from c.3% at <0.5-2mm to <1% at >5mm.	Restricted to ‘flowerpots’.	4, 5, 7