Kent Archaeological Review extract

The Evidence for Ancient Quern Production at Folkestone.

Since 1974, over 60 greensand querns of a type common in late Iron Age and early Romano-British contexts have been collected from the foreshore at East Wear Bay, Folkestone by five different finders. The majority of these has been recovered by local enthusiast, Mr M Dugdale of Saltwood, who reported his findings to the writer.

The querns have been recovered from the foot of a low, slumped gault cliff (centred at NGR TR 24263695) that projects onto the foreshore at the head of the shingle beach (Figure I) immediately below the site of the Roman villa excavated during 1924 and 1925 (Winbolt 1925). Many of the stones had been exposed from this section of eroding cliff following storms and were found lying at the foot in the resultant clay slurry at high water mark. In some cases the stones were found lying further down the foreshore having been washed out to sea by strong currents.

The geology here comprises a thick band of gault clay approximately 100 feet in thickness, which overlies Lower Greensand. At its junction with the Greensand, the gault is almost liquid. This has caused massive landslips over many years, the effects of which have drastically altered the landscape of this area. The cliff edge almost certainly stretched further seawards in Roman times, perhaps by as much as 400-500 metres. Investigations on this stretch of foreshore, carried out by the Kent Archaeological Rescue Unit following storms in the winter of 1974, confirmed that part of the cliff section at the head of the beach originally came from a higher level (for an interim note see Keller 1982; see also below).

Area location and finds spot.
Figure 1. Area location and finds spot. (Based on Ordnance Survey.)

The foreshore querns are cut from greensand which may be readily attributed to the Folkestone Beds, the uppermost formation of the Lower Greensand. The particular hard sandstone required for their manufacture is geologically confined to an area which lies between the small headland of Copt Point and Stanford, approximately 5 miles to the north-west (Smart et al, 1966 pages 88-97, Figure 55). The nearest and the most readily available source of stone within this area would be from the section conveniently exposed in the low cliff at Copt Point only some 400 metres to the south of the East Wear Bay foreshore. It is highly likely, therefore, that the Copt Point outcrop provided the source material for the quernstones found on the adjacent beach. The querns are of the semi-rotary, or 'oscillatory', domestic hand type and include both upper and lower stones. Many are water-worn and damaged, some badly enough that essential features are no longer recognisable. Although a few of the querns recovered from the beach appear to be finished products, such as could be found on rural settlement sites, significantly the majority appear to be unfinished. This is indicated by partially worked surfaces (Figure 2 Numbers 2 and 6); incomplete hoppers (Figure 2 Numbers 3 and 4); and partially bored spindle holes (Figure 2 Number 5). Several of the stones are almost certainly boulders that had been roughly shaped in preparation for final shaping and dressing, these may be termed 'rough-outs' (Figure 2 Number 1). In many cases the intended form is discernable from the `rough-out' and the unfinished quern undoubtedly, however, what may have started out as an upper stone may well have finished as a lower stone. In many cases the querns appear to have fractured during shaping, or most likely, the hole boring stage due to flaws in the rock and these would have been subsequently discarded. In some cases evidence of faulty workmanship can be seen such as misaligned central spindle holes (Figure 3 Number 10). It would seem here that holes had been bored from both sides of the stone but that they simply failed to meet up The finished stone, however, may still have been useable, functioning quite adequately as a lower stone. Only a small percentage of those recovered represent probable finished querns and include upper and lower stones (five are shown here, Figure 3 Numbers 7-11). In general appearance these have well-defined shapes, with evenly dressed surfaces showing clear traces of tooling. These can be arranged into basic groups according to their shape:

UPPER STONES: Three different forms have been recognised to date.
  1. (Figure 3 Number 7)

    Steep-sided, cylindrical form with wide, funnel-shaped hopper and narrow, circular spindle hole. The grinding surface is either flat or slightly concave. A socket hole for a handle is provided in the side (not shown here). This form is the most common so far found at Folkestone and as far as the writer is aware is neither a form well known at other sites nor in other materials. At present this would appear to be a Folkestone Type, typical of the East Wear Bay site. It is hoped that future work will confirm this view.

  2. (Figure 3 Number 8)

    A nearly hemispherical form with wide, shallow recessed hopper and narrow, circular spindle hole. The grinding surface is either flat or slightly concave. A socket for a handle is also provided in one side. This form superficially resembles Curwen's Number 10 (Curwen 1937 page 141).

  3. (Figure 3 Number 9)

    A rather squat, rounded, 'bun-shaped' form with narrow, funnel- shaped hopper and narrow, circular spindle hole. The grinding surface is either flat or slightly concave. None is sufficiently well-preserved to determine the presence of a socket hole for a handle, though one would presumably have been provided. This form superficially resembles Curwen's Number 7 (Curwen 1941 page 17).

LOWER STONES: Two forms have so far been recognised.
  1. (Figure 3 Number 10)

    These are essentially circular discs of varying thicknesses with flat or flattish upper and lower surfaces. These incorporate a roughly centrally placed spindle hole penetrating the full thick ness of the stone.

  2. (Figure 3 Number 11)

    This form (only one of which is currently represented in the assemblage) is broadly similar to the latter example but has a pronounced rounded, convex base. A circular spindle hole is incorporated into the upper flat grinding surface and penetrates up to a half of the full thickness of the stone. Similarly shaped lower stones made of Hertfordshire Puddingstone occur at Baldock (Foster in Stead and Rigby 1987, Figure 78 Numbers 793-4).

Clearly the querns recovered from the foreshore are at various stages of manufacture and include 'rough-outs' through to the finished objects. This seems to have important implications which have been considered further below.

During the excavations in 1924 and 1925 of the villa complex on the cliff above the East Wear Bay foreshore, some 20 greensand querns were recovered by Winbolt (Winbolt 1925 page 108). Sadly only one is provenanced (found in the North Corridor of Block B) and of the group only five survive. These are now housed in Folkestone Museum. One of these, an upper stone is unfinished and clearly resembles Type-1 of the foreshore assemblage. Others of this type may have gone unrecognised by the excavator. In fact Winbolt notes a form with 'broad cavities in the middle without holes' that he suggested were 'used as a kind of mortar for pounding grain with a pestle or stone rubber'. (Winbolt 1925 page 108). It now seems more likely that these are unfinished Folkestone-Types.

Unfinished quern stones from East Bay foreshore.
Figure 2 Unfinished stone querns from East Wear Bay foreshore, Folkestone. (1:6)

From hand-specimen and thin-section analysis it has recently been shown that the surviving querns from the villa site are manufactured from Folkestone Beds sandstone (Miss C Ingle pers. comm.). This stone was almost certainly obtained from the same local source as the foreshore querns. Indeed Winbolt himself was of the firm opinion that the villa querns were of sandstone obtained at Folkestone. The villa quern collection seems unusually high for a normal rural settlement and the discovery nearby of a further 60 stones seems far in excess of what might normally be expected from such a site. This is reflected in the numbers of querns reported from a selection of notable published rural sites both within and outside Kent listed in the accompanying table.

Number of querns from a selection of sites broadly similar to that of the Roman villa at Folkestone. A 7-entry table follows showing county, site, number of querns & reference.
County. Site Number of Querns. Reference.
Kent. Baston Manor, Hayes. 7 (min). Philp 1973, page 90.
Kent. Darenth. 8 (min). Payne 1897, page 74;Philp 1973, pages 143-4..
Kent. Faversham. 1. Philp 1968, page 85..
Kent. Wingham. 2 (min). Dowker 1883, pages 357.
Herts. Baldock. 30. Stead and Rigby 1987, Figure 78.
Herts. Gadebridge Park. 6. Neal 1974, pages 193-4, Figure 84..
Sussex. Fishbourne. 7. Cunliffe 1971,page 153 and Figure 71..

This list is by no means exhaustive and is merely intended to illustrate the smaller numbers of quernstones found from sites broadly similar to Folkestone. W S Penn wrote over 20 years ago that querns were 'the most sadly neglected artifacts in archaeology' (Penn 1966); a point that had already been made clear to Curwen some 29 years earlier (Curwen 1937, page 133). To some extent this situation has not changed and querns are still rarely published in any detail or remain completely unpublished.

The excavation by KARU during the winter of 1974 of a portion of the slumped cliff at the head of the foreshore revealed an oblique stratified section. This was examined and at least two layers were found to contain pottery and quernstones (of which two were recovered; another two were found lying on the beach). The pottery all appears to be native in character and suggests a date of between AD 10-80. No obvious Roman material was found. More significantly, one layer was composed entirely of greensand dust. This almost certainly represents stoneworking debris. It was difficult to interpret the section as the layers were set back steeply into the cliff. The layers may represent fills within either a slumped ditch or a pit or, they may possibly represent a series of horizontal deposits including the probable stoneworking level. It seems highly likely that these stratified deposits are remnants of an earlier native settlement that stretched further out to sea than did the later Roman villa estate.

From the material collected from the foreshore and from the evidence recovered from the excavations at the head of the beach and at the site of the Roman villa, it seems certain that large-scale quern production was being undertaken hereabouts, which made full use of the natural resources available at the Copt Point headland situated a little to the south. The material may have been quarried from the beds exposed in the cliff face or, alternatively, much of the material may have been readily available in the form of boulders lying on the ground at the foot of the eroding cliff. Convenient-sized boulders could easily have been collected and stored close to the working area. The dating of the pottery recovered from the foreshore cliff- section is significnat in that the earliest villa buildings at Folkestone do not appear to date from before AD100. The quern production, as evidenced by the greensand dust layer, must therefore relate to the pre-villa native settlement for which there is clear evidence. Thus quernstone production at Folkestone seems to have started in the first century AD and, not inconceivably, before the Roman conquest. The early dating of the foreshore querns is largely confirmed by the overall form of the objects themselves. The typology of querns proposed by Curwen (Curwen 1937) shows that pre-Roman querns are characterised by tall, nearly-hemispherical and 'Bee- hive' shaped upper stones. These characteristics are a broadly similar feature among the foreshore querns, although they remain fairly distinct in their style. Quernstone production at Folkestone probably continued into the succeeding centuries and there is certainly one quern of typical Roman form in Folkestone museum together with another found on the foreshore. Clearly a major problem still to be resolved is the intensity of production on the site throughout the Roman era. The writer is of the current opinion that quern production is essentially a feature of the period between the first- century BC to the first-century AD with the numbers produced declining in the later Roman period. It may be of significance that this decline in quern production coincides with the expansion of the villa complex.

Finished quern stones from East Bay foreshore.
Figure 3. Finished quernstones from East Wear Bay foreshore, Folkestone. (1:6)

What is certain is that a major quern-making industry flourished hereabouts on these lower slopes above the sheltered bay . In all probability this formed part of the extensive pre-Roman native settlement occupying the eastern headland that almost certainly developed into the later villa estate. This industry may have continued a tradition of quern making in the general area that stretches back to prehistoric times. Recent finds from Wingham and Deal may confirm this. A saddle-quern and rubber, both of greensand taken from the Folkestone Beds and thought to have been obtained at Folkestone, were found in a Neolithic pit at Wingham (Greenfield 1960, page page 66-7). A similar source is most likely for a greensand saddle-quern recovered from a late-Bronze Age/early-Iron Age pit from Dosset Court, Upper Deal. The pit also contained two pottery vessels dated between 1000-800/700 BC (McPherson-Grant in Farfitt 1985). Preliminary work has shown that broadly similar querns to those found at Folkestone occur on nearby Kent sites. These are also of sandstone from the Folkestone Beds and almost certainly derive from the same source. These finds indicate a local trade in and distribution of these Folkestone products. Broadly similar querns from the Folkestone Beds have also been found as far away as Hunsbury, Northants where at least two examples resembling Folkestone Upper Stone Type 1 (Figure 3 Number 7) have been found at the Iron Age hillfort (Miss C Ingle pers. comm.). These important finds indicate an even wider distribution and more significantly help confirm the early date of these locally made querns. 'Work on this project is still very much in progress with particular emphasis being placed on the aspects of trade and distribution. The writer would therefore like to hear from anybody who has or knows of any greensand querns (especially those thought to derive from the Folkestone Beds). It is to be emphasised that any stratified querns recovered from excavated sites from well dated contexts would be of immense value. A study of all the available regional (and non-regional) material- a task recently thought to be long overdue (Black 1987, page 63) - will hopefully enable the identification of many more querns from the Folkestone production site. This material and in particular material recovered from settlement sites, will hopefully illuminate the date, scale and nature of the production and distribution processes for the Folkestone products. The identification of a substantial quern production site at Folkestone must surely be of major importance to local studies and beyond. The writer feels that there are many important implications that have yet to be fully considered. The final results of this work will therefore be published in detail at a later date once more information has become available.

BIBLIOGRAPHY. A 16-entry table follows showing author, date, publication.
Author. Year of publication. Reference.
Black, E W. 1987. The Roman Villas of S E England. British Archaeological Reports: Brit. Ser. 171.
Cunliffe, B. 1971 . Excavations at Fishbourne 1961-69: Volume ii The Finds: Research Report of the Soc. of Antiquaries. 26.
Curwen, C E. 1937. Querns: Antiquity 11, pages 133-151.
Curwen, C E. 1941. More about Querns. Antiquity 15, pages 15-32.
Dowker, G. 1883. The Roman Villa at Wingham: Part 2. Arch. Cant. 15, pages 351-7.
Greenfield, E. 1960. A Neolithic Pit and other Finds from Wingham, East Kent. Arch. Cant. 74, pages 66-7 and Figure 3.
Keller, P T. 1982 . Rescue Excavations in Folkestone from 1973. KAR Number 69. pages 209-11 1.
Neal, D S. 1974 . The Excavation of the Roman Villa in Gadebridge Park, Hemel Hempstead 1963-68. Research Report of the Soc. of Antiquities. 31.
Parfitt, K. 1985 . Some Iron Age Sites in the Deal Area. KAR Number 79, page 213, Figure 5.
Payne, G. 1897. The Roman Villa at Darenth. Arch. Cant. 22, pages 49-84..
Penn, W S. 1966. A Quern Survey in Kent KAR Number 3, pages 46-8.
Philp, B. 1968 . Excavations at Faversham l965: The Royal Abbey, Roman Villa and Belgic Farmstead
Philp, B. 1973 . Excavations in West Kent I960-70.
Smart, J G O. 1966. Geology of the Country around Canterbury and Folkestone. Memoirs of et. al. the Geological Survey of Great Britain.
Stead, I M & Rigby, V. 1987. Baldock: The Excavation of a Roman and Pre-Roman Settlement. Britannia Monograph Number 7.
Winbolt, S E . 1925 . Roman Folkestone.