Researches and Discoveries

RESEARCHES AND DISCOVERIES IN KENT A NEOLITHIC STONE AXE: OFFHAM, NEAR WEST MALLING A fine Neolithic greenstone axe was discovered in July 2000 on the surface of a field near West Malling (Fig. 1). The find-spot lay about 1km north-east of Offham village (NGR TQ 6676 5780). The land here lies on the dip-slope of the Lower Greensand ridge at an elevation of about 50m above OD. The axe is complete and largely undamaged, with a maximum length of 110mm. The polished, oblique cutting edge has a width of 57mm and remains reasonably sharp. The sides of the axe taper to a narrow butt, which shows some evidence of battering through use. In profile, the specimen is somewhat asymmetrical, with a maximum thickness of 36mm. Neolithic axes, of both flint and stone, with oblique cutting edges as in the present example, are not uncommon finds. The form seems to be deliberate and could be a reflection of the method of hafting and use. Particularly noticeable on the implement is a thin dark vein that runs diagonally through the matrix of the natural rock. This was created during metamorphosis, when a crack in the rock developed and became in-filled with am phi bole, a mixture of magnesium, iron and silicates. Nonetheless, this 'fault' in the axe did not seem to compromise its value as a functional tool and the feature might well have held some special significance for the Neolithic owner. The rock has been identified as a uralitized gabbro, belonging to Group I of the Implement Petrology Group's classification, with a suspected source in the Mount's Bay area of Cornwall (McK Clough and Cummins 1979, 127). Although clearly of imported rock, it is not known whether this axe-head was actually manufactured in Kent. It seems most likely that it arrived in the County in its finished state. The high distribution of Group I implements (50%) within I 00 miles (160km) of London suggests that a Neolithic precursor of London may have been a secondary distribution centre for such axes (Cummins 1979, 12). Group I implements are the most prolific group in the 'south-eastern area'; hitherto eight examples have been recorded 397 RESEARCHES AND DISCOVERIES Fig. I Cornish greenstone axe from Offham. Scale, I :2. Drawn by Jo Bacon. from Kent. 'To what degree this is due to the volume of production of the south-western manufacturers, or the relish with which these implements were acquired by the south-eastern population is not clear' (Woodcock, Kelly and Woolley 1988, 23-4, 132). Always axe-heads, Group I implements have so far only been found in central and west Kent, with examples known from Boughton Monchelsea (TQ 779495), Charing (TQ 9549), Greenwich (TQ 384780), Hawley (TQ 548719), Keston (TQ 4163, TQ 4164), Lenham Heath (TQ 913503) and Murston (TQ 9164). All but one of the Group I axes found in the region have a rounded butt and an oval section (ibid., 24). A Group XIX axe, also from Cornwall, has been discovered at Mill Hill, Deal, in the east of the County (Parfitt l 990, 10). The area where the present discovery was made is of some interest and clearly implies Neolithic occupation, or at least activity, closeby. That this general region of Kent was something of a focus for activity during the Neolithic period is clearly demonstrated by the presence of the Medway megalithic barrow group (Holgate 1981; Ashbee 1999; Ashbee 2000) and the recently discovered long-house on Blue Bell Hill (Glass 1999, 192). The two long barrows at Addington lie just 2km to the north-west of the present find-spot. Acknowledgements: Thanks are due to Nigel Betts for allowing his find to be recorded and published here. John Francis (Department of 398 RESEARCHES AND DISCOVERIES Mineralogy, Natural History Museum) and Alan Tyler (Bromley Museum) analysed the rock composition and gave valuable opinion and comment. David Williams (Dept. Archaeology, University of Southampton) formally studied and identified the rock petrology. Gillian Varndell (Curator (Neolithic) Department of Prehistory and Early Europe, British Museum) gave a formal identification and assessment of the axe-head, and drew on similar comparisons. Claire Mason (Maidstone Museum) gave access to the museum collections in her care. Jo Bacon kindly provided the illustration. MICHAEL LEWIS KEITH PARFITT Ashbee, P., I 999, 'The Medway Megaliths in a European Context', Archaeologia Cantiana, CXIX, 269-284. Ashbee, P., 2000, 'The Medway's Megalithic Long Barrows', Archaeologia Cantiana, CXX, 319-345. Cummins, W. A., 1979, 'Neolithic Stone Axes: Distribution and Trade in England and Wales', in (eds) McK Clough, T. H. and Cummins, W. A., Stone Axe Studies, CBA Research Report, No. 23, 5-12. Glass, H., 1999, 'Archaeology of the Channel Tunnel Rail Link', Archaeologia Cantiana, CXIX, 189-220. Holgate, R., 1981, 'The Medway Megaliths and Neolithic Kent', Archaeologia Cantiana, xcv11, 221-234. Parfitt, K., 1990, 'Excavations at Mill Hill, Deal I 982-1989: An Interim Report', Kent Arch. Rev., IOI, 9-18. Woodcock, A. G., Kelly, D. B. and Woolley, A. R., 1988, 'The Petrological Identification of Stone Implements from South East England', in McK Clough, T. H. and Cumming, W. A., (eds.) Stone Axe Studies, Vol. 2, CBA Research Report, No. 67, London, 21-33. 399 RESEARCHES AND DISCOVERIES AN INSULAR LA TENE TOGGLE: WROTHAM In February 2001 a looped toggle was found by Keith Smallwood (S.E. London Metal Detecting Club) in plough soil whilst metal detecting on farmland near Wrotham. Although only finds of medieval or postmedieval date had been previously recorded from this site, a number of late Iron Age coins have been recovered elsewhere in Wrotham parish suggesting that this area may have formed a focus for late Iron Age depositional activity. The toggle was recorded under the Kent pilot of the Portable Antiquities Recording Scheme; full details of its precise find spot are held by the Kent SMR. The toggle is described and a short note on its typology is appended. The cast copper alloy looped toggle is of double bulbous shape with flattened knob-like terminals. Its integrally cast attachment loop, which 399 RESEARCHES AND DISCOVERIES 􀀁 \..J Fig. I Insular La Tene toggle from Wrotham. Scale I: I (kindly drawn by Jo Bacon). is set above a near circular void, is only slightly raised above the edge of the toggle. Length of toggle 27 .5mm, diameter of flattened knob mouldings 9.5-!0mm, weight 9.5g. [KENT3569- PAK52/5] (Fig. I). Looped toggles formed a similar type of dress or harness fastening to the button-and-loop fasteners discussed by Wild (1970). They probably held together two pieces of either fabric or leather strap; the loop would have either been sewn to a fabric base or attached to it by a thong and the toggle was inserted through a slit or loop in an adjoining piece of fabric or leather. Jackson has noted that, in size and form, looped toggles are similar to the late Iron Age and early Roman Class I button-and-loop fasteners with solid double-boss head and heavy ring-loop, but that they may also be prototypes of the late Roman Class IX fasteners which have a transverse bar for a shank (Jackson 1990, 40). Given the dates of examples of the type recovered during archaeological excavations, it is reasonable to suggest that the looped toggle is most probably a distinct sub-type of Wild's Class I fasteners (Macdonaldforthcoming). Three main types looped toggle are recognisable (Macdonald forthcoming). The first (Group A) consists of those toggles whose loops are attached directly to one side of the toggle. The second (Group B) consists of those toggles which also have loops attached directly to one side of the toggle but whose loops are only slightly raised above the edge of the toggle and therefore are set above a circular void or depression. The third (Group C) consists of those toggles which are connected by a neck, set midway along the toggle, to a loop which is invariably turned at an angle to the neck. These three 400 RESEARCHES AND DISCOVERlES groups can be further sub-divided according to the shape of the toggle. Three forms of toggle are recognisable: those which are barrelshaped with flanged ends (sub-type i), those which take the form of a double bulbous moulding with flanged ends (sub-type ii) and those which have flanged ends that are separated from a protruding, central rib by thinner and often waisted sections (sub-type iii). The classificatory system provided by the combination of these two criteria is reltively inclusive; only the looped toggle from Bishop's Stortford, Hertfordshire (Hattatt 1989, 416-no.30C, fig. 7), which has a toggle made up of a flanged, double bulbous moulding separated by a central rib, represents a 'hybrid' (Group Cii/iii) form. Looped toggles have been identified as an insular La Tene III type whose distribution is largely restricted to southern Britain (Jackson 1990, 40). Independently dated examples suggest that the type's vogue extended from the late Iron Age through to the early Roman period (Macdonaldforthcoming). The Wrotham looped toggle is the only example, known to the authors, of the Group Bii type. It is most closely paralleled by the Group Bi looped toggle from Hengistbury Head, Hampshire (Bushe-Fox 1915, 61, no.9, pl. XXIX; Cunliffe 1987, 153, no.46, ill.111 ). The Hengistbury Head looped toggle was found on site 11 during Bushe-Fox's excavations of 1911-12, but the date and associations of its precise context are unknown. Occupation at Hengistbury Head is continuous throughout the late pre-Roman Iron Age into at least the mid second century AD (Cunliffe 1987, 336-346) and the looped toggle presumably dates to some point during this period. To suggest a similar date range for the Wrotham looped toggle would not be unreasonable. MICHAEL LEWIS PHILIP MACDONALD KEITH SMALLWOOD Bushe-Fox, J. P., 1915, Excavations at Hengistbury Head, Hampshire in 1911-12 (Rep. Res. Comm. Soc. Antiq. London No. 3) Horace Hart (for the Society of Antiquaries), Oxford. Cunliffe, B., 1987, Hengistbury Head, Dorset. Volume I: the prehistoric and Roman settlement, 3500BC-AD500, Oxford Univ. Comm. Archaeol. Mono. No. I 3, Oxford. Hattatt, R., 1989, Ancient brooches and other artefacts. A fourth selection of brooches together with some other antiquities from the author's collections, Oxbow Books, Oxford. Jackson, R., 1990, Camerton. The late Iron Age and early Roman metalwork, British Museum Publications, London. Macdonald, P., forthcoming, 'A looped toggle from near Little Mill, near Caerleon, Monmouthshire' (and a note on the type). Wild, J. P., 1970, Button-and-loop fasteners in the Roman Provinces, Britannia, I, 137-155. 401 RESEARCHES AND DISCOVERIES IRON AGE MATERIAL: CAPEL-LE-FERNE During the summer of 1996 the writer and Rob Masefield (RPS Consultants, formerly RPS Clouston) monitored the construction of a new water main west of Dover. The pipeline, for Folkestone and Dover Water Services, ran from Lower Standen pumping station in the west to Dover Priory Station reservoir in the east, a distance of approximately 7.5km. North-east of Hockley Sole House (at TR 2464 4040), on land belonging to Mr Hector Walker, a pit was found after topsoil stripping. This was at 142m OD where the south-facing slope of the field begins to drop sharply towards the road. The rectangular flat-based pit, with slightly sloping sides and rounded corners, measured 1.12m by 0.8m. Although shallow (on average 0.2m deep) it is presumed that it was originally more substantial. The feature was cut into firm yellowbrown clay and the pit fill, remarkably similar in colour and texture to the natural Clay-with-Flints, was identified by the large numbers of pottery sherds within it. On excavation, the pit was found to be well-defined with a single fill (Context 3/2) containing a varied assemblage of pottery totalling 155 sherds, described below. Five flint flakes, probably residual, were found in the fill and also a length of strap-like material decorated by copper alloy studs or rivets. Ian Riddler of the Canterbury Archaeological Trust arranged for this to be examined and the strap was identified as mineralised leather, 20mm wide and 65mm in length but broken into two sections during deposition. Two studs are visible on the longer length of strap and one on the smaller section. However, the break has come at a point where the perforation for another stud is visible, as is also the case on the far end of the small section. The Leather Strap by Penelope Walton Rogers A fragmentary leather strap, pierced by three copper alloy rivets, is almost certainly of Iron Age date. The leather survives in a mineralised form and includes accreted plant remains, amongst which fern can be identified. Three copper alloy rivets survive, which have hollow, domed heads. They are closelyspaced, at intervals of approximately 15mm between their centres. The metal has blackened in some areas, suggesting that it could be a high-tin bronze. The belt is relatively narrow, with a width of approximately 20mm and, to judge from the surviving remains, it included a considerable number of decorative rivets. With this in mind, it is perhaps more likely that it formed part of a harness rather than a dress accessory. The Pottery by Nigel Macpherson-Grant An overall total of 155 prehistoric and later sherds (weight: 815gm) was re- 402 RESEARCHES AND DISCOVERIES covered from Context 2. All the pottery has been statistically analysed according to fabric, manufacturing, form and decorative characteristics. The fabric codes used belong to the Kent-based series now employed by the CAT. This data is held in the site's and regional ceramic type series archives. Overall character: The assemblage contains three main ceramic traditions: two indigenous Late Iron Age, represented by both flint-tempered and sandy wares (109 sherds); and an intrusive 'Belgic'-style tradition with grog-tempered wares (46). Of the former, the sandy wares visually dominate, with smallmedium sized sherds with surfaces pitted from leached-out calcareous inclusions. Analysis of wear-patterns indicates that the assemblage includes indigenous flint-tempered sherds some (but not all) of which are residual and derived from an occupation of some length, some part-profiles and sherds with unifacial wear suggesting deposition into context some time before final burial, and some unworn material, including [9) (Fig. 1.9). Vessel fabrics: Fabric types were identified using a x I O hands-lens and coded according to regional fabric types. These are summarised in the following table with obviously intentional filler types italicised. An assessment of fabric matrices indicates that up to eight different sources are represented in this assemblage. Those used for the indigenous flinttempered fabrics llA 1-2, 7-8, 30, and possibly IB J (and free from macroscopic calcareous inclusions) probably stem from three to four sources fairly local to the settlement. Another source altogether is represented by the dominant group of sandy fabrics, LIA 3-5, 21, and may not be from the immediate environment. These are epitomised by the presence of vari-sized voids from leached calcareous inclusions. Though they indicate one general clay source, two subsources were used, one with both fine black and quartz sand, and one predominantly with fine black sand. The leached-out void appearance-range and sandy matrices of Fabrics 3-6, 21, are virtually identical to sandy Early Roman wares from the Folkestone district which use clays derived from the local Greensand Beds. These wares are characterised by the presence of sparse to frequently profuse inclusions of a marly aggregate containing grains of black sand. In both the present and Fo!kestone material there is a high degree of variability in inclusion grain size, from fine dust, 1-2mm average, or frequently coarser up to 4mm. Though detailed comparisons have still to be made, the prevalence of this trend (even in the later more refined Conquest-period Folkestone material) suggests that these inclusions are natural to the clays used and to some extent may have been deliberately retained as a naturally occurring filler type. A further source is represented by the fine marl inclusions present in the 'Belgic'-style fabric B 28. Amongst the deliberately tempered LIA fabrics flint fillers are generally rather profuse and, fairly characteristically for the period, are frequently up to a coarse 4mm or even 7mm size. For grog-tempered 'Belgic'-style elements, 2-3mm grains occur in coarseware vessels, with fairly consistent fine I mm grog and sparse fine flint in the Fabric B 28 beaker [9J. Vessel surface finishes: Burnishing on the LIA fineware bowls [3-4) is su- 403 RESEARCHES AND DISCOVERIES TABLE l. DOVER SPINE MAIN, SITE 1: FABRIC TYPES AND SHERD FREQUENCIES Fabrics Description Qu'y Illus. No. Jndig. LIA, tempered LIA! Flint-tempered in sparse-mod. quartz 9 1-2 sand LIA 2 Flint-temDered in profuse quartz sand 4 - LIA 7 Sp. flint and sp-m organic-temp., fine 7 8 s'dy LIA 8 Flint-temp. with sp organic and I combed quartz incl. LIA 30 Flint-temf)ered in fine sandy matrix I - Jndig. LIA. untempered LIA 3 Fine sandy, sp-m blk sand, sp crs qtz, I - sp-m calc. inc LIA 4 As LIA 3 but with fairly profuse FeO I - LIA 5 Sandy, prof qtz, blk sand, mod-prof 48 Bowls calc. incls 3-4, 6 LIA 6 Fine sandy, sp-mod quartz and black 10 - sand LIA 21 Fine s'dy, prof blk sand, sp-mod qtz, 26 5, 7 fine marl incls 'Belgic' -style B 22 Grog-tempered with sp-mod quartz 3 - incls B 28 Grog-temp. with sp flint, sp-m qtz, 43 Conical fine marl incls 9 Indigenous • Belgic' -style IB 1 Grog-tempered with sp-moderate flint 2 - perficial and the irregular horizontal finish on [3] is more a question of broad, shallow, almost fluted tooling rather than a true burnish. The finish on the 'Belgic' -style beaker [9] is of good quality, fairly even and regular, but again a question of the surface being 'pressed down' smooth rather than truly burnished. For the coarsewares, the tooled inner rim of (I] has a distinctive broad, almost facetted, treatment and its body exterior has a rough diagonal gritdragged finish. Finishes on sandy coarsewares such as [7] consist of light superficial horizontal smoothing or wiping. 404 RESEARCHES AND DISCOVERIES 􀀁 F􀀄-- 1 2 9 1􀀂􀀃 . . . . 7 ( •· 5 6 Fig. I Capel-le-Ferne: Indigenous handmade wares ( 1-8); 'Belgic' -style wheel made ware (9). (Scale I :4) Firing trends: The assemblage contains predominantly reduced material (! 17 sherds) with only a small quantity of partially (25) or wholly oxidised sherds (3). There is a greater degree of firing variation amongst the indigenous flinttempered fabrics and this may indicate bonfire firing, rather than the likelihood of clamp-kiln firing which is suggested by the generally consistent trend for reduced products amongst the indigenous sandy and 'Belgic'-style wares. Vessel and decoration types: There are no fabric-related vessel class preferences, with both finewares and coarsewares occurring in the three main fabric 405 RESEARCHES AND DISCOVERIES classes. There is one fine ware vessel, the conical 'Belgic' -style beaker [9]. Despite their rather crude finish, four vessels can be classed as sub-fineware types, the sandy ware S-profiled bowls [3-4] (and one not illustrated) and the flint-tempered jar rim [8], which is probably from a medium-diameter Sprofiled jar. A total of up to 28 coarseware vessels are represented including the probable medium-diameter storage jar [ 1] and the smaller cooking or storage-vessel categories represented by [2, 5-7). Forms and sherds in the similar sandy fabrics LIA 5 and 21, are likely to come from the same source and the wall-thickness frequency peak for these, and diameter ranges for the sub-fineware bowls, are a reflection of the sort of general productional consistency that one should expect for products from the same potter/workshop. Only one small coarseware sherd in the indigenous flint-tempered fabric LIA 2 is decorated, with fine close-set combing in the 'Belgic' manner. Though technically a finishing technique, the deliberate sub-division of the burnished finish on the 'Belgic' -style beaker [9] is worth noting in terms of its likely freshly-made appearance. Dating The combination of heavy thickened everted rim and facetted internal finish on the large jar [I] is typical of a distinctive type of Late Iron Age coarse ware that occurs in a number of later Iron Age assemblages from eastern Kent, including Bigbury, Canterbury, Highstead, Barham Downs, Ebbsfleet, Worth, and most recently in an assemblage from a Late Iron Age enclosure near Whitfield. Some relevant parallels are summarised in Table 2. With the exception of Whitfield, most examples come from fairly intensively occupied multi-period sites lacking clear-cut stratigraphic sequences with an inevitably high degree of residuality and intrusion problems. For these reasons the main difficulty has been in assessing the temporal relationship between the indigenous flint-tempered and 'Belgic' -style traditions. So far, dating has been based on a complex inter-weave of likely, but rather tenuous, regional ceramic and stratigraphic linkages. Only one small assemblage from Barham Downs was free of multi-period 'clouding' and had no immediately associated grogged ware. This example and the implied stratigraphic evidence from sites such as Ebbsfleet in Thanet has allowed for the tentative application of a c. 150-75/50 BC date bracket for material similar to Nos. 1-2. In addition, though not in direct association, the consistent occurrence from Ebbsfleet and other sites of indigenous LIA-type S-profiled finewares, has suggested that this date could be partly under-pinned by the c. I 00-70 BC archaeomagnetic dating of the 'waterhole' assemblage from Bigbury (Clark 1983, 276). This was a mixed-tradition assemblage producing similar native flint-tempered and sandy finewares, together with 'Belgic' -style grogged material, and virtually free of earlier and later prehistoric and Roman material. Illustrated sherd sizes in this group imply that at least some of each tradition were contemporary. However, it also appears to contain a fair qua􀃃tity of small-size material and may be an abandonment clearance-deposit which could well contain both primary and already-residual secondary rub- 406 RESEARCHES AND DISCOVERIES TABLE 2: DOVER SPINE MAIN 1996, SITE 1: INTER-ASSEMBLAGE PARALLELS Zone/Site/Source Finewares Coarsewares MEDWAY TO STOUR Bigbury (Nr. Canterbury) Bowls 3-4, cf. Jar 2, close to Bigbury 68, (Thompson 1983) Bigbury 75 (flint), 76 cf. Bigbury 88 (gro2) Highstead/Chislet 1975-6 Bowls 3-4, related to Jars 1-2, related to High'd Macpherson-Grant Highstead LIA 9 LIA 9d, 10, 21, 40, 62 (forthcoming) Jars 6-7, cf. High'd LIA 5 R. STOUR TO COAST Barham Downs 1971 None Jar I, general type (Macpherson-Grant 1980) parallels Mill Hill, Deal 1904 Conical jar 9, close None (Rigby 1995) parallel Mill Hill X4. I Stour Street, Canterbury None Jars 1-2, general type Cakebread-Robey 1976 parallels Whitfield/Eastry Bypass Bowls 3-4, general Jars 1-2, general type 1994, Site 2 parallels (in sandy parallels (Macpherson-Grant fabrics) forthcoming) Worth, Sandwich None Jars 1-2, general type (Unpub. assemblage) parallels ISLE OF THANET Birchington (Worsfold Bowls 3-4, general Coll) 1938-40 type (Thompson 1982) parallel (grog) Ebbsfleet 1990 None Jar 2, cf. Ebbsfleet 48-9 (Macpherson-Grant 1992) Jar 7, cf. larger Ebbsfleet 47 bish: but until it has received a thorough wear analysis, certainty regarding degrees of tradition contemporaneity must wait. The apparent relatively pure single-period content of the Bigbury 'waterhole' underlines the regional need for uncontaminated assemblages, not only for the broad period c. 150 BC to 50 AD, but particularly for the crucial c. I 00-50 BC initial phases of interaction between native and 'Belgic' traditions. Towards this end the present assemblage and the fairly large body of material recovered from Whitfield Site 2 are useful contributions, partly because there are bone and C-14 dates from the latter site. These results extend the dating that can be applied to indigenous pre- 'Belgic' LIA wares but easy comprehension requires graphic presentation (Parfitt in prep. a). These new dates do not appear to affect the current start-dates for 'Belgic' -style grogged pottery 407 RESEARCHES AND DISCOVERIES and since this tradition occurs at both Whitfield and Dover Spine Main assemblages, dating centres around this aspect. For Whitfield the percentage ratio between native and 'Belgic'-grogged material is approximately 80%-20% respectively and, though grogged wares do occur in primary contexts, the quantities are extremely low and most is likely to represent late-phase discard or post-abandonment rubbish. Initial date estimates place the enclosure between c. 75-50/25 BC with an arguably conservative emphasis towards mid-century. For the current site relative tradition percentages and associated ceramic types are similar but obviously this single-context assemblage cannot be used to closely date the associated settlement, though it epitomises some of the problems associated with dating LIA ceramics. Indigenous flint-tempered coarsewares: The period characteriser [I] is not readily paralleled in the Bigbury waterhole assemblage but it is well-represented at Whitfield; (2) is present at both sites. It is worth remembering that the bulk of the flint-tempered material in the present assemblage is generally smaller and more worn than the other ware types. At this stage the general dating given above for c. 150-75/50 BC can be applied. Indigenous S-profiled sandy ware bowls: [3-4] are a formal type well-represented at both Whitfield and Bigbury. At Bigbury they occur in both flinttempered and grogged fabrics. The only difference is that none of the Bigbury examples are in sandy fabrics; however, they are essentially a fineware type and there is at least one small S-profiled fineware jar represented by Bigbury 81 in a sandy Greensand fabric. Of relevance is that [3-4] are simple forms and likely to be long-lived as a vessel class. One general example that may be considerably later is a grogged Thompson Dl-4 bowl from Birchington (Thompson 1982, 315, 625-6). For the time being, and on the basis of the Bigbury archaeomagnetic dating, bowls [3-4) can be dated c. I 00-70 BC. 'Belgic '-style grog-tempered wares: There are no parallels from either Bigbury or Whitfield for the conical jar [9]. It was almost certainly a pedestalled beaker or jar and broadly similar to a more flaring-walled example from Rose Lane, Canterbury - a Thompson Type A 10; she gives late earlier first-century AD dates for these (op. cit., 83-4, No. I), but implies the likelihood of occurrences in the first century BC. A somewhat coarser, but still good, parallel comes from Mill Hill, Deal, with identical finishing treatment as (9) (Rigby I 995, 47-8, Fig.63 X4. l , and Plate XX); this has a general dating of late first century BC-early first century AD, with the likelihood of some occurrences predating c. 25 BC. Conservatively, (9) can be placed between c. 25 BC-25/40 AD. Dichotomy: There is an immediate apparent difference between the dating of the indigenous wares and the 'Belgic' -style [9). Even if (9) does pre-date c. 25 BC there is still a potential 25-50 year gap between the likely dating applied to the flint-tempered and sandy wares - assuming that this is correct. Source: One key factor here is represented by the sandy ware bowls [3-4) and 408 RESEARCHES AND DISCOVERIES Fabrics LIA 3-6 and 21, all with fine black sand and some containing inclusions of calcareous aggregates. All are likely to be related in some way (i.e. from the same general source) and some may be related to broadly contemporary material from the Whitfield LIA enclosure. As indicated, the closest parallels for this general group of fabrics in terms of both production-locale and ware types, are Conquest-period sandy wares from Folkestone, using Greensand-derived clays. One site in particular, the native farmstead near Newington (CT site F25A 1998) produced a large essentially 'Belgic' -style assemblage containing both traditional grogged wares and both handmade and wheelmade sandy wares. The latter probably represents a pre-Conquest phase of Romanisation, possibly due to the increasing influence of the Gaulish imports and reflected throughout eastern Kent by the appearance of sandy wares or fine non-grogged wares in immediately pre- or post-Conquest contexts. The implications of this material has still to be adequately studied, so estimates for the start-date of the wheelmade material are tentative, but somewhere around c. 25-30 AD is likely. The important aspect regarding the Folkestone material is the presence (unfortunately from a heavily-occupied multi-phase settlement) of handmade sandy 'Belgic' -style vessels; whilst some of these may be contemporary off-the-cuff non-workshop productions, their presence could also imply earlier roots for this tradition. ls there a direct relationship between the apparently earlier Dover Spine Main sandy wares and the Conquest-period Folkestone material? If there is, it means that there were native workshops producing pre-'Belgic' wares (at least in the sense of prior to the introduction of purely grogged fabrics) as early as c. 75-50 BC or, alternatively, that whilst still workshop-produced, (3-4) are somewhat later and dateable to between c. 50/25 BC-25 AD (to take into account the suggested date range for the conical beaker No. 9). Canterbury: It is felt that a date quite as late as c. 25 BC is unlikely for the flint-tempered coarsewares ( 1-2). No pottery of this type has been noted from Phase I of the Canterbury Marlowe Car Park 'Belgic' enclosure (Blackley et al. 1995, Fig.4), though unpublished material from this area makes it clear that some specifically indigenous-style flint-tempered wares were still being produced (alongside grogged wares and local flint-tempered copies of the latter) up to c. 20-15 BC or later. Unpublished facetted-rim coarsewares have been found from Canterbury but from only one site close to the river. This potentially earlier location is some distance from the Marlowe enclosure and could imply settlement-shift (though other scenarios are possible) at some point after c. 50 or 25 BC. But the potential support that this could represent for the region's chronological framework has to remain in abeyance until a comprehensive review of both Canterbury's prehistoric pottery and the material from Bigbury and other sites is undertaken. The available regional data indicates that [9) ought technically to be the latest element present. However, on a purely intrinsic basis there is no reason to suspect that the larger sherds in all three main fabric traditions present are not broadly contemporary. Which means that either beaker (9) is significantly earlier than c. 25 BC or the flint-tempered jars [ 1-2) are later. Since dating of the 409 RESEARCHES AND DISCOVERIES latter is based on comparative (albeit slight) stratigraphic data which is indirectly under-pinned by Bigbury's scientific date, it is felt that if (1-2] are late, it is towards the end of their existing estimated date range rather than radically later, i.e. around c. 50 BC. As a result a date of c. 75-50/25 BC is proposed for the time being. Implications for pottery production Regarding the sandy fabrics Fabrics LIA 3-6 and 21, there are no equivalent clay types in the immediate vicinity of the site; the equation of these with wheelmade Conquest-period and slightly earlier handmade fabrics recorded from the Folkestone area is therefore slightly unexpected. It is an equation that is highly likely but really needs consolidation via petrological analysis; samples from this site have been submitted for examination and the results will be included in the forthcoming discussion of the similar wares from Whitfield Site 2. To this assessor's knowledge no similar material dated as early as c.75-50 BC has been recorded from the immediate Folkestone area. Since these sandy wares are present at least three to four miles further east than the likely clay sources, they are likely to represent marketable or at least exchangeable products, and in this assemblage are in sufficient quantities to imply workshop production beyond the immediate needs of the home settlement. This whole point is further strengthened by the occurrence of an LIA 2 I-type fabric (and a range of probably related sandy wares) from the Whitfield Site 2 enclosure, a full ten miles from likely source, and again in a Downland zone with apparently no equivalent clay types. This point is important since it is the first significantly pre-Conquest AD evidence suggesting indigenous (non- 'Belgic') organised pottery production and its dispersal in eastern Kent. It is likely that the same clay sources may have been employed in the production of both the Conquest-period AD material and the sandy wares in the Dover Spine Main and Whitfield assemblages; if this data reflects genuine continuity of production then it implies a degree of workshop marketorientated production from at least c. 75-50 BC, if not earlier. In turn this likelihood needs to be considered alongside that from other parts of Kent for this general period. In the upper Medway valley, sandy Greensandderived clays were being used for pottery production during the Mid-Late Iron Age (and earlier, cf. the Late Bronze/Early Iron Age sandy wares from Aylesford in an unpublished British Museum collection). This area (Thompson's Zone 4, op. cit., 11 and Map 2) is at the western end of the Holmesdale (lying between the chalk North Downs and the sandstone beds of the Weald to the south); Folkestone is at its eastern end and the presence of sandy wares need be no more than a reflection of the same availability of suitable potting clays (macroscopically the major difference between the two is that those from Folkestone mostly, but clearly not always, contain calcareous inclusions). However some Medway zone S-profiled fineware bowls and jars were being exported from c. 250 BC onwards across the Thames Estuary to Little Waltham and Mucking, Essex (Drury 1978, Fabric A, 56, 58-9, 126-7 and Figs.38 (Form 13), Fig.71), and within Kent to Holbury and Oldbury and down the coast to Birchington, Thanet. The LIA sandy wares from Bigbury are similar and may 410 RESEARCHES AND DISCOVERIES also represent traded wares, either from the Medway region or another source. In addition, the LIA material from Ebbsfleet, Thanet, contained S-profiled finewares made in lightly flint-tempered profusely sandy (non-Greensand) fabrics suggesting that despite the local traditional use of flint fillers, sandy clays were being deliberately sought for the production of at least some finewares. Whilst some regional occurrences of LIA sandy wares may represent no more than isolated traded items or localised sporadic influence-adoption, the combined implication of the Medway source(s), and now possibly Folkestone, suggests that there may have been a much greater degree of organisation to the production and dispersal of later second/first-century BC Kentish ceramics than previously suspected. The tone of uncertainty underlines yet again the need for a comprehensive review of regional LIA assemblages. This new Dover Spine Main, Whitfield and Folkestone data-bundle signposts the potential of a further indigenous Kentish centre, near Folkestone, producing marketable vessel quantities during the later Iron Age. Even though, compared with the Medway source(s), establishment may be later and around c. 75-50 BC, the potential it represents is well in line with the later Iron Age trend for increasing commercialisation represented by the introduction of coinage. This potential now needs to be thoroughly confirmed, both in the field and petrologically, and any implications compared with other regional data, including both indigenous and 'Belgic' -style ceramics, metalwork finds and the increasingly numerous quantities of Late Iron Age coinage and the model suggested by Cunliffe for Kentish socio-economic territories, confirmed or modified (Cunliffe 1982, fig.22). The primary significance of the isolated, shallow pit on the route of the Dover Spine Main has proved to be that, because it contained an uncontaminated assemblage of Late Iron Age pottery including indigenous and 'Belgic' -type wares, it has assisted in the dating of other less secure assemblages and pinpointed the temporal relationship between the different traditions. It has also given extra support to theories of possible early workshop production of pottery in eastern Kent and suggestions that there may have been a degree of organisation in its marketing and exchange. Secondly it has identified a Late Iron Age site in a part of the Kent where finds from this period are sparse. A few sherds of similar date were found approximately 750m to the south during work on the A20 road (Rady 1991) and a number of Late Iron Age cremation urns were excavated at the extensive Great Hougham Court Farm site, 2.6km to the south-east (Parfitt in prep.b ).Pottery of a comparable date was also recovered at Hawkinge Airfield, 3km to the west (Stevens in prep.), but other parallels discussed in the pottery report above are at some distance from the site. The feature is likely to be a clearance pit at the edge of a settlement, dug and quickly backfilled with clean natural clay. Judging from the condition of the pottery, which is a mixture of fresh and partly worn 411 RESEARCHES AND DISCOVERIES sherds, the settlement was occupied for some length of time. The varied nature of the assemblage, which includes cooking pots and fine wares made from clays from eight different sources, only some of which are local, and the inclusion of the possible harness fragment, suggest that the settlement was well established. MAUREEN BENNELL Bennell, M., 1997, Dover Spine Main: Archaeological Monitoring Report (RPS Clouston archived report). Blockley, K., et al., 1996, Excavations in the Marlowe Car Park and Surrounding Areas, The Archaeology of Canterbury V (1996), Parts I-II. Clark, A. J ., 1983, 'Archaeomagnetic Dating at Bigbury', in Thompson, F. J ., 'Excavations at Bigbury, near Canterbury 1978-80', The Antiquaries Journal, !xiii (1983), 237-275-6. Cunliffe, B., 1982, 'Social and economic development in Kent in the preRoman Iron Age', in Leach, P. E. (ed.), Archaeology in Kent to AD 1500, CBA Res. Rep. 48, 40-50. Drury, P. J., 1978, Excavations at Little Waltham 1970-71, CBA Res. Rep. 26. Macpherson-Grant, N., 1980, 'Archaeological Work along the A2: 1966- 1974', Archaeologia Cantiana, XCVI ( 1980), 133- I 83. Macpherson-Grant, N., 1992, 'Appendix II : The Pottery', in Perkins, D. R. J ., 'Archaeological Evaluations at Ebbsflet in the Isle of Thanet', Archaeologia Cantiana, ex ( 1992), 269-311. Parfitt, K., in prep. a, Excavation of the late Iron Age enclosure at WhitfieldEastry Bypass Site 2, CAT Occasional Paper. Parfitt, K., in prep. b, Settlement Continuity and Change on the Chalklands of East Kent: Evidence from the Folkestone-Dover Transfer Pipeline, 1998. Rady, J., 1991, Extension of the A20: Folkestone to Dover, Canterbury's Archaeology 1990-1991 (CAT 15th Annual Report). Rigby, V., 1995, 'The Pottery', in Parfitt, K. et al., Iron Age Burials from Mill Hill, Deal, British Museum. Stevens, S., in prep., Excavations at Hawkinge Airfield, Archaeology SouthEast report. Thompson, F. J ., 1983, 'Excavations at Bigbury, near Canterbury 1978-80', The Antiquaries Journal, !xiii, 237-278. Thompson, I., 1982, Grog-tempered 'Belgic' pottery of South-eastern England, British Archaeological Reports I 08, Oxford. 412 RESEARCHES AND DISCOVERIES SECTIONS OF ROMAN ROAD: BENENDEN Fieldwork at Bishopsden Farm has involved tracing the course of the Benenden-Canterbury Roman road and selective excavation along it. It is the most heavily metalled iron-slag road investigated so far in the area. The first section (TR 8445 3485) showed the road surface to be 2.30m wide with a depth of slag of 0.26m. Wheel ruts were present in the surface and there were indications of a single road ditch on the north side (Fig. 1 a). At a second site, 280m west, towards Bexhill Farm (TR 8423 3487), only the upper profile of the road was recorded in order to leave this well-preserved road intact (Fig. 1 b).The total width of the DARK SOIL \ IROAD EDGE 0 --c=c.:.c:::_-_--1...1 "' - -- '- IRON SLAG ROAD ..- ROAD/ EDGE 0 .: ■ I■ =-•==::tl ==r:::::::::r􀀆"' Fig. I. (a) (above) Roman Road at Bishopsden Farm: Section I, east face. (b) (below) Section 2, facing west. 413 RESEARCHES AND DISCOVERIES road was 6m, significantly wider than section 1, possibly due to its position at the base of a steep incline. There appeared to be two adjacent road surfaces composed of slag and ironstone. The first, on the north side was 4m wide, and the second at a slightly lower level was 2m wide. Further work continues where access can be arranged. Acknowledgements: Thanks are due to Mr & Mrs Honnywill for their interest and hospitality; also to W. Coomber and E. Pollard for their help with the fieldwork and excavations. NEIL ALDRIDGE 414 RESEARCHES AND DISCOVERIES MEDIEVAL DENEHOLE: WILMINGTON, DARTFORD On 20 August 2000 a survey was made by the Kent Underground Research Group of a denehole which had suddenly re-opened the previous month in the rear garden of number 50, Birchwood Road, Wilmington. (NGR TQ 5036 7179) A small hole had appeared in a lawn area which revealed a large irregular underground cavity. The roof of a chalk chamber had failed and, over many years, the fall had migrated upwards to the surface. At the side of the hole the edge of a substantial concrete raft was visible, evidence that the original entrance shaft had been capped at some time in the past. An examination of the intact underground chambers accessible from the collapse revealed that the denehole was a typical six-chambered type, with two sets of three chambers on each side of the narrow entrance shaft giving a double clover leaf or trefoil ground plan. The shaft had been sunk through 1.6m of very loamy Thanet Sand and a further 3.9m in chalk before the miners started to dig the horizontal galleries. With such a shallow cover above the chalk the miners prudently extended the shaft to ensure greater stability. The collapse of the eastern chambers occurred at the junction of the primary and lateral chambers. These chambers are of smaller dimensions than those on the other side of the shaft, which suggests that tbe mediaeval excavators recognised some form of structural threat and limited their activities in this area. The junction of adits or galleries was always a potentially dangerous area and the miners would take great care on the profile of the tunnels to ensure structural integrity. Any geological anomaly at this point could weaken the structure and cause roof falls and ultimately total collapse. The subsidence also destroyed half of the original l .Om diameter shaft. The miner's footholds were still clearly visible in the intact portion, as were grooves 414 RESEARCHES AND DISCOVERIES worn in the chalk by the hauling ropes taking the excavated chalk to the surface. The chambers to the west of the shaft were found to be in very good condition, 3.6m high with no signs of roof falls or flaking from the walls. Tool marks from the usual short-headed iron pick were abundant on the walls and roof. To make up for the reduced chalk output from the eastern chambers, those on the west were enlarged. The digging of small headings at 90 degrees from the ends of the lateral chambers allowed the miners to compensate and achieve the required output from the denehole (approximately 350 to 380 tons). The excavated chalk was spread on the surrounding fields as a top dressing to improve crop yields in an area where the subsoil is somewhat acidic, a process known as 'chalking' or 'marling'. The garden backs onto an area of Joydens Wood which has mediaeval field systems and associated deneholes which pre-date the square earthwork (now destroyed by housing), located a few hundred metres to the west. Excavations in the earthworks by P. J. Tester and J.E. L. Caiger found the footings of buildings dated 1280 to 1320, and are thought to be part of the lost mediaeval manor of Ocholt. 1 The date of construction of the denehole under discussion is therefore probably before 1280 and a date of around 1250-1260 is suggested. The denehole was filled and made safe shortly after the survey. R. F. LEOEAR 1 'Mediaeval buildings in the Joyden's Wood Square Earthworks·, Archaeologia Ca11tiana, LXXII ( 1958), 18-39. 415 RESEARCHES AND DISCOVERIES PROBABLE SE VENTEENTH-CENTURY WITCH BOTTLE: BJDDENDEN This find was originally made at some time during the l 960s at Tanner Farm, Biddenden (NGR 8294 3856) to the south-east of the Three Chimneys public house. Workmen engaged in repairs to the house discovered a glass bottle beneath a stone slab which formed the base of the principal inglenook fireplace in the property. The bottle is typical of an 'onion type' wine bottle of c. 1650. Owing to the effects of heat from the fire above, the glass has largely Jost its colour except on the base where it can be seen to have been olive green originally. The height of the bottle is 117mm, the base 98mm in width whilst the neck extends some 53mm above the top of the base. It was empty when found. A direct comparison can be made with an almost identical type of container found at Reigate (Massey, 2000). The Surrey bottle is of the 415 RESEARCHES AND DISCOVERIES same type and date but still contained the original contents, which mainly comprised bent pins and urine. The belief in the use of these anti-witch devices reached a peak during the seventeenth century. Acknowledgment: The writer is grateful to the present owner, M. Coomber, for drawing his attention to this object. NEIL ALDRIDGE Massey, A., 2000, 'The Reigate Witch Bottle', Currenr Archaeology, 169, 34-36. 416 RESEARCHES AND DISCOVERIES CHALKWELLS: HARTUP AND L YNSTED On 26 November 2000 a chalkwell shaft subsided in the grounds of a commercial cold storage site at Spade Lane, Hartlip. (NGR TQ 8356 6504) The site manager, Mr S. Coad, kindly informed the writer and an investigation was made by members of the Kent Underground Research Group on Sunday 3 December. A roughly elliptical hole 2.0 by 2.8m had suddenly appeared in a shallow grassed bank to the east of the cold store buildings. Very little of the original shaft profile could be seen as most had collapsed when the chalkwell reopened. From what remained it was possible to estimate an original diameter of 2.0m. The remains of the shaft widened out to 3.9m in rough diameter at a depth of 8.1 m. Chalk was exposed at a depth of 5.9m below the surface. The overlying strata was noted as Thanet Sand with a metre or so of poor Brickearth just below the surface. Three small chambers radiated from the base of the shaft, the largest measuring 4.4m long by 2.0m wide and 3.3m high. It was not possible to ascertain the true depth of the excavation as the debris from the collapse had covered the floors of the chambers to an unknown depth. The lowest attainable point was found to be 11.6m. No tool marks were discernible as the chamber walls and roofs have suffered from minor falls and spalling of the chalk since mining operations ceased. The overall layout of the excavation and the poor mining techniques employed suggest that this chalkwell was dug by farm labourers when other work was slack. A date of late eighteenth/early nineteenth century would be appropriate for this particular example. The chalkwell has now been filled and made safe. In December 2000 a deep hole appeared near an old field boundary at Hole Street Farm, Lynsted and was first reported to the writer by KAS member Mr Michael Peters of Sittingbourne. The Grid Reference has been withheld from publication at the farmer's request. 416 RESEARCHES AND DISCOVERIES A visit and survey by KURG members in January 2001 revealed a typical chalkwell type of denehole. Three chambers, average size 8.0m long by 5.lm wide and 4.2m high, had been excavated from the base of a 2.0m diameter 15.6m deep shaft, which had passed through 1.9m of Clay-with-Flints before entering the upper chalk. Two somewhat larger chalkwells in the local area are known to have been supplying chalk for lime burning in the late nineteenth century. The above example was dug at an earlier period, probably in the eighteenth century to provide raw chalk for use as a top dressing on the adjacent fields. Approximately 380 tons of chalk had been excavated from this example. The usual application rate was around 60 tons per acre. R, F. LEGBAR 417 RESEARCHES AND DISCOVERIES AN EARLY NINETEENTH-CENTURY ENGRAVING OF MALLING ABBEY TOWER Ruins, therefore, of all kinds, are at first deformed; and afterwards, by means of vegetation and of various effects of time and accident, become picturesque. 1 In the era before photography began to make an impact as a popular documentary tool (i.e. before c. 1860) drawings, and the engravings made from them, are sometimes the only means by which one can make judgements about the vicissitudes of edifices - most notably in the absence of precise documentary evidence. This short paper aims to address a common and (to some) obvious problem of the representation of historic buildings through an analysis of J. Storer' s 1809 engraving of the tower of Malling Abbey (Plate I). The engraving of the Abbey's famous Romanesque tower was made from a drawing by J. Hawksworth and was published alongside a description of the conventual buildings in the Antiquarian and Topographical Cabinet.2 A popular early nineteenth-century publication, the Cabinet specialised in 'elegant views of the most interesting objects of curiosity in Great Britain' (often medieval) and it would be fair to say that the texts accompanying the engravings are unremarkable. 3 They offer little more than a concise account of the history of each building, with very little aesthetic appreciation. Indeed, in referring to the tower at Malling Abbey, the Cabinet can offer little more than that it: ... affords in its present state an interesting specimen of Norman 417 RESEARCHES AND DISCOVERIES PLATE I View of Malling Abbey Tower from the Antiquarian and Topographical Cabinet. Vol. VI. London, 1809. Copper plate engraving by 1. Storer from a drawing by J. Hawksworth. 8.7cm x 6cm. (Photo: Centre for Kentish Studies, Maidstone) 418 RESEARCHES AND DISCOVERIES architecture; it has on either side a small tower that contains three tiers of rounded arches, some of which are exquisitely ornamented. 4 The contribution of the Cabinet's artists, however, reveals much more about the way its contemporary audience was expected to relate to the 'curiosities' contained within and Storer's view of the tower is a good example. He was helped in his task, of course, by the very fact that the tower was (and to some extent remains) an enigma. It is clear that Storer' s engraving is a product of the application of the theory of the picturesque - adherence to which dominated the representation of ancient buildings (and their surroundings) from the mid eighteenth century and well into the nineteenth. In this respect, it is interesting to note the style of the engraving in relation to the context of its publication in the Cabinet. Storer' s view presents us with what appears to be, at first sight, an attempt to accurately describe the notable medieval structure for a discerning early nineteenth- century antiquarian audience. Little contextual information related to the site is given (i.e. its relationship with the surrounding buildings) and the edifice completely dominates the field of vision. Scale is provided by three figures of largely indeterminate age (though the two nearest the central door could be children engaged in animated conversation). Contextual details are limited; the meandering stream in the foreground, some rough terrain on the right hand side, a rustic fence to the left with, above it, part of a tree. The mouldering textures of the structure itself are clearly emphasised, with indications of weed growth in the masonry. Typically, for this form of representation, the ruined north wall has been given a filigree edge. It is clear, in addition, that every effort has been made to draw attention to the variety of stonework, both ancient and modern, to be found at that time. This adds a great textural vitality to Storer's representation which must certainly have been present in Hawksworth's drawing and is accentuated by the engraving process. The overall impression is of a patched-up, if not neglected, timeworn structure, dramatically lit so as to emphasize the linear qualities of the Romanesque arcades as they divide up the composition. The ruinous outline of the later fifteenth-century addition, which gives rise to its current status as a tower, is clearly emphasised. A. M. Oakley confirms this status in her book on Malling Abbey, whilst Newman in The Buildings of England contends that the tower was once a west front modelled on Rochester Cathedral after a fire of 1190 had destroyed the earlier fa9ade. 5 Both sides of the argument have some merit, but for this discussion it is the achievements of the AlQ RESEARCHES AND DISCOVERIES London banker Frazer Honeywood, who bought the estate and conventual buildings in 1740, that are most important (the Cabinet, however, states that he 'succeeded' to it). Clearly, in his response to the ruinous state of the structures, Honeywood was interested in making the most of its picturesque potential. In true eighteenth-century fashion his 'restorations' worked to achieve an artful effect of controlled decay and this is the impact we feel in Storer's engraving. Honeywood was in an enviable position in that he did not need to work too hard to achieve the ruinous effect he was no doubt desirous of achieving. The vicissitudes of history combined with the work of nature avoided having to build a ruin from scratch - such a popular pastime in the eighteenth century, and one of which William Gilpin (1724-1804 ), one of the pioneers of picturesque theory, did not approve.6 Much has been written of the impact of theories of the picturesque on eighteenth- and early nineteenth-century art and architecture. Along with Gilpin, the two other principal protagonists of the picturesque, were Richard Payne Knight (1751-1824) and Sir Uvedale Price (1747-1829). Amongst other things, all three strove to explain that peculiar mixture of sensations that can be felt when viewing either landscape or, better still, landscapes with ruins containing figures. In going to great lengths to account for this, they cite many different qualities of scale, texture, light effects, and so on which contribute to the understanding of what we would now call 'visual dynamics' or, more simply, variety. Complete products of the early Romantic movement, the works of Gilpin, Payne Knight and Price were hugely influential, principally in the way that they summed up the preoccupations of previous generations of antiquaries who had found vicarious inspiration in ruins. Honeywood was certainly one of these and must have been in the vanguard of those who longed to have their imaginations stirred by the romance of ruins. Theorists of the picturesque were at pains to point out that their object of study lay somewhere between the sublime (in itself more than capable of being expressed in a ruin, as admirers of Piranesi would point out) and the beautiful, both of which had been famously explored by Immanuel Kant and Edmund Burke in the middle of the eighteenth century.7 Their analyses of aesthetic experience attempted to explain the full gamut of emotions ranging from terror and astonishment (the sublime) to calm contemplation (the beautiful). Both pointed out that these states, as experienced in nature or art, resulted in pleasure. 8 The effect of these ideas on poets, artists and architects from the mid eighteenth century onwards has been profound and it is interesting that Price et al were trying to 420 RESEARCHES AND DISCOVERIES describe, in the picturesque, an aesthetic category that was principally concerned with variety, irregularity, tonal contrast and texture. This was removed from both the savage and awe-inspiring qualities of the sublime and the serenity and mildness of the beautiful. An outline of some of the principal characteristics of the picturesque as defined by Price, for example, will help in considering the aesthetic intentions of Storer' s engraving. In his famous Essays on the Picturesque (first published in 1794) Price outlines the principal characteristics of Gothic architecture and, significantly, describes how medieval ruins lend themselves perfectly to the theory. According to Price, 'ruined abbeys' in particular display: ... the triumph of the picturesque; and their charms to a painter's eye are often so great, as to rival those which arise from the chaste ornaments, and the noble and elegant simplicity of Grecian architecture. 9 The ennobling of the medieval ruin to a position of equality with that of classical architecture is significant. Price's mild corruption of Johann Winckelmann's famous dictum that Greek architecture displayed 'noble simplicity and calm grandeur' demonstrates his wish that the picturesque be taken seriously as an aesthetic pursuit. 10 Indeed, Price had already established this view earlier in the Essays when considering the picturesque potential of the Gothic ruin and is at pains to point out one of its principal attributes: Gothic architecture is generally considered as more picturesque, though less beautiful than Grecian; and upon the same principle that a ruin is more so than a new edifice .. and among the various causes of the superior picturesqueness of ruins compared with entire buildings, the destruction of symmetry is by no means the least powerful. 11 Here we must note how Storer's view of the tower, looking at it from the north-west, seemingly makes every effort to avoid emphasizing its symmetry and places great visual emphasis on the ruined outline which is most visible from that angle. In this respect, the engraving is a typical example of the general absorption of ideas on picturesque theory which, as Price's account testifies, was originally intended to be utilised by garden designers but which was quickly adopted by artists and, one surmises, by their patrons. Devotees of the Cabinet were, therefore, served up a suitably engaging version of Malling Abbey tower at a time when the fashion for the picturesque in architecture and engraving was at its height, surely sustained by the continuing popularity of Price's Essays. Although points of difference 421 RESEARCHES AND DISCOVERIES between Price and, in particular, Payne Knight have been pointed out, the picturesque was a well-understood aesthetic category in the late eighteenth and early nineteenth century. 12 A further significant point in the reading of Storer' s image is the fact that it is an engraving on copper. Although it is now a truism that 'the medium is the message', the engraver of picturesque views was certainly helped by the medium at his disposal - particularly in the form of the copper plate. Here, the difference between copper and steel engravings (the latter having supplanted the former technique for small plates from c. 1820) is worthy of consideration. It has been noted that with the introduction of the steel plate: ... a shimmering, pale grey became for the first time a possibility in line engraving, and it is this that provides the most recognizable characteristic of steel beside the heavier and warmer mood of copper.13 Expressed crudely, the increased number of lines per inch capable of being engraved on a steel plate has an aesthetic result that often takes away a certain directness of feeling. In other words, steel engravings have a tendency to 'even out' textures resulting in a noticeable Joss of contrast and 'roughness' - an effect central, of course, to the aesthetics of the picturesque. Clearly, Storer employs his medium very effectively in rendering intelligible the particular mood which readers of the Cabinet would have enjoyed. Malling Abbey may have rightly been considered to be a 'curiosity' by the publishers of the Cabinet, but its subscribers were presented with a view of the edifice entirely in keeping with their expectations. J. Storer' s engraving is, therefore, artfully constructed to serve a particular purpose and, it can be argued, tells us more about the way the past was viewed in 1809 than the true state of the preservation of the tower in that year. Starer's engraving was produced just a few years before Thomas Rickman' s highly influential classification of British medieval building styles brought about a re-appraisal of the way these structures were to be judged. Through the publication of his Attempt to Discriminate the Styles of Architecture in England of 1817, Rickman instigated a more rigorous approach to the analysis and interpretation of medieval buildings. A case can be made, therefore, that the engraving was published at a time when the concerns of its audience lay more with artful effects than with archaeological accuracy through a clear adherence to the theory of the picturesque. SIMON BLISS 422 RESEARCHES AND DISCOVERIES 1 Price, U., Essays 011 The Picturesque, Vol. I, Gregg International, 1971, 199. The work was first published in 1794. This is a facsimile reprint of the 1810 edition, published by J. Mawman of London. 2 Antiquarian and Topographical Cabinet, Vol. VI, London, I 809. Volumes are unpaginated. Descriptions of buildings and accompanying illustrations are grouped according to county. The publishers of the 1809 volume are listed as W. Clarke, New Bond St; J. Carpenter, Old Bond St; C. Chapple, Pall Mall; J.M. Richardson, Cornhill; and Sherwood, Neely and Jones, Paternoster Row. 3 The authors of the descriptions are not credited. 4 Antiquarian and Topographical Cabinet (see note 2). 5 See Oakley, A. M., Ma//in1-: Abbey 1090-/990, Malling Abbey, 1990, 60; and Newman, J., The Buildings of En1-:land: West Kent and the Weald, Penguin, 1980, 602. 6 Gi !pin's distaste for artificial ruins is briefly discussed in Macaulay, J ., The Gothic Revival 1745-/845, Blackie, 1975, 18-19. 7 Kant's Observations 011 the Feeling of the Sublime wul thi• Beautiful was first published in 1764. Burke's Philosophical Enquiry into the Origin of our Ideas 4 rhe Sublime and the Beautiful was first published in 1757. 8 An account of Kant's and Burke's approaches to the sublime and the hcautiful can be found in Goldthwaite, J. T .. 'Translator's Introduction', in Kant, I. (trans. Goldthwaite} Observations on the Feeling of the Sublime and the Beautiful, University of California Press. 1960, 1-38. 9 Price 1794, op. cit. (see note I), 54. 10 The corruption is from Winckelmann's famous Thou1-:hts 011 the lmita1io11 of Greek Art in Painting and Sculp/llre, first published in 1755. A modern translation can he found in Holt, E. G., A Documentary History of Art: Vol. II, Doubleday Anchor, 1958, 335-351. 11 Price 1794. op. cit. (see note I), 52-3. 12 For a short account of the theoretical differences between Payne Knight and Price. see Ross, S., 'The Picturesque: An 18th Century Debate'. The Journal o( Ae.11/retit·.1 and Art Criticism, Winter I 987. 275-277. 13 Gascoigne, B., How to Identify Prints, Thames and Hudson. 1986. section 55g. The volume is unpaginated. 423