Little Farningham Farm, Cranbrook, Revisited

LITTLE FARNINGHAM FARM, CRANBROOK, REVISITED NEIL ALDRIDGE The site of Little Farningham Farm, Cran brook (Fig. 1) is well known to scholars of Romano-British archaeology as being one of the locations that has produced examples of the stamped tiles of the Classis Britannica.' These were discovered in the late 1950s during the excavation of a Romano-British iron working site. This is the furthest inland that these tiles have yet been found. 2 No other site in the Kentish Weald has yet produced similar tiles although a number of Llttlo Pam.Ingham Parm .f. 1 To Rocl>c$1ber lranc.h • DellowsM pot ,·-·-·-·-·-·-·-· . • I : ! ' ·-·---•-·---·-·' oi:::i•,,,•s:::, 3 5, ;:is .,,.==,,,;,' d.' SCALI! IN Ml!TRl!S Fig. 2 Plan of Excavations, c. 1960 based on that by Mrs M. C. Lebon. Letters A to G indicate areas discussed in this paper. 140 LllTLE FARNINGHAM FARM, CRANBROOK, REVISITED PLATE II Areas D and G, facing north-east, c. 1959 (M. C. Lebon) channel possibly connected with an iron working process and used as some kind of quenching pool. From the excavated areas detailed above came mainly 'native type', grog-tempered pottery wares in grey or black/red fabrics, together with a few fragments of Sarni an and other wares, all of late first century to second century AD date. Because of the difficulty encountered with identifying trenches and the associated finds it has proved impossible to link the pottery finds to any individual areas. It has been necessary to treat the ceramic finds as a whole for dating purposes; certainly there is a complete absence of any material from the third and fourth centuries. The excavation also produced a number of fragments of window glass (further pieces were found during the most recent fieldwork (see below), presumably originating from the buildings excavated or from a neighbouring structure. Many of the various types of tiles were found to be stamped with CL BR. Indeed by the close of the I 957 season some fifty examples had been recovered with five or six different types of stamps. No clear evidence of tilemaking was found on the site (apart from baked clay 141 NEil., ALDRIDGE bosses/saggers presumably used to keep tiles apart while firing) despite the great variety of tile and tile fragments found. 12 Most of the tiles used in the drain [G] were CL BR-stamped types. Also the pedalis found in area [A]. The majority of the other stamped examples appear to have been found right across the 1950s' excavation area. All of the stamped tiles were of clay from the Guestling area, Peacock's Fabric 2. (No examples of stamped tiles were found during the recent investigations - see below.) Although it was stated in the original site notes that some of the pottery may have been Iron Age it would seem more probable that the type of native wares that were largely continuing to be used into the Roman period were those of the pre-conquest period. 13 Certainly the recent programme of field walking around the perimeter of the scheduled site has not identified any pottery that could be earlier than first century AD. This suggests that the settlement at Little Farningham may not have been founded until after the Roman conquest, although it would not completely rule out earlier 'Belgic' iron working somewhere on the site. There was only a brief note published in Archaeologia Cantiana for 1957; however by 1958-9 it appears that a proposal had been made to schedule the site. The official reason was to ensure that any future work on the site would be carefully planned and directed and that an adequate labour force would be organised. No further areas were to be opened up and all work should be concentrated on 'going over' the areas previously dug 'insufficiently in earlier seasons' .14 lt seems that one of the major problems faced by the excavators was how to differentiate between the disintegrated walls and the apparent 'floors' of unmortared stones. Some of the walls appeared to have been robbed of their stone at some time in the past. The interpretation made at the time was that there seemed to be several possible rooms with narrow passages between them. This could also be interpreted as spaces between walls of disparate dates, it is not possible to speculate further owing to the absence of any further dating evidence. In the report published in the Journal of Roman Studies 15 it is stated that two rooms were identifiable, i.e. [A] and [BJ referred to above; and a further room identified from a small test hole associated with a deposit of plain and painted wall plaster (CJ); the extent of this third 'room' was not, as has been stated earlier, established further. Adjoining area A to the north was a 'pit', [F] in Fig. 2, which was filled with many pieces of clay tuyeres, the nozzle tips used during iron smelting to protect the end of the bellows. It was stated that the associated pottery in this pit was Iron Age but this may have been deceptive in view of the 'native' character of the majority of the other 142 LITTLE FARNINGHAM FARM, CRANBROOK, REVISITED pottery found on the site. The report for 1959 states that deeper digging had produced further pieces of window glass. From what area these came is not clear. A hearth like structure [E] was also found lying to the north-east of [B]. This was 1.6m in length and 60cm wide; the site records state, 16 that it was rounded at one end, although the plan depicts both ends as being rounded. The description says that it had a raised border of burnt clay with a floor of concave appearance. The interior was filled with burnt wood and iron slag. 17 It should be noted that a worked iron bloom was found somewhere in this area.18 Across the excavated area were a number of areas of charcoal and iron slag stratified at the level of the Roman features. In fact slag and cinder were found across the greater part of the site. The records do not properly record or interpret the site stratigraphy and there is no means of knowing the relative dating of the two 'rooms' with their stone walls or the possible third 'room' suggested by the deposit of plaster debris found in [C]. The only details of a 'typical section' appear as a general description: 26cm of topsoil, 23cm of soil containing much rubble, tiles etc., 15cm of crumbly darker soil containing Romano-British potsherds and iron slag. Iron slag is also present in the upper layers but less densely so. Finally, underlying the site was the pale grey natural clay into which features such as wall foundations and pits had been cut. The question that needs to be asked at this point is what was the purpose of the two, possibly three rooms? The discovery of window glass, plain and painted wall plaster and flooring of opus signinum suggests some type of domestic structure, possibly part of a villa or perhaps a bath complex? However despite the presence of tiles suitable for use in a hypocausted heating system none of this material had been used for such a purpose. 19 Generally the site had the appearance of a rough shanty type area. There are a number of features common to both Little Farningham and the Romano-British site partially explored at Bodiam (12km to the south) in the 1960s. 20 Both had re-used materials of tile and brick, many bearing the CL BR stamp. The Bodiam excavation was severely hampered by the fact that only narrow trenches could be opened up between the rows of a hop garden that was being laid out. It was thought that there had been at least eight successive periods of occupation on the Bodi am site. The four earlier phases had been destroyed by fire, including what seemed to have been the most substantial structure during phase 4. This building of timber, with possibly a tiled roof, had a rough paved floor comparable to some of the features recorded a few years earlier at Little Farningham. The whole of the 143 NEIL ALDRIDGE construction was however somewhat 'rough and ready' and materials made for other purposes had obviously been used in a haphazard way. A significant proportion of the pottery found was of native manufacture which pointed to the site being used by the local population. The few coins found at Bodiam were of Trajan and Marcus Aurelius which suggests a probable main period of occupation around the mid-second century. A direct date parallel, suggested by the coinage, can therefore be drawn with Little Farningham. It is most unlikely that the iron working would have been undertaken contemporaneously with the assumed domestic functions of the other features found in areas [A], [B).and [C]. It would appear therefore that the principal features are of two different periods and functions. Without reliable stratigraphic relationships between the features it is very difficult to postulate which might be the earlier. The position of the features close to a watercourse would suit both a domestic and an industrial purpose. The site would not have been suitable for a substantial villa as it lies at the base of a relatively steep slope and in close proximity to the watercourse. A unique ceramic artefact was recovered shortly before the site was scheduled in 1960, as Mrs Lebon describes: .... a large pot, believed to have served as bellows in the first or second century AD,21 was somewhat casually discovered during a visit to inspect the condition of the scheduled site in March 1960. Pieces of it had long been visible in the sides of a trench, [marked 'p' in Fig. 2] and had been mistaken for portions of imbrex tiles. Enough had now weathered out to show something of its form: the thick flat cut rim, tall neck with faint groove at its base, and the bulbous body, pieced by three significant holes of I inch diameter, spaced equidistantly round the widest zone of the pot. It was immediately recognized that such a vessel would answer for bellows, which we had hoped to find in association with the numerous tuyeres. The pot, 11 inches in height, has now been skilfully reconstructed at the Institute of Archaeology. This vessel may have been used with three sets of bellows attached in order that a continuous supply of air could be maintained to the smelting furnace. The tuyeres, of short life, would have been simple and inexpensive to replace them. Mrs Lebon describes the finding of a small gem of red jasper close to the bellows pot. It had an intaglio possibly depicting Pan and is now in Maidstone Museum together with some other material from the site. THE RECENT FIELDWORK AND EXCAVATION A preliminary resistivity survey was carried out at Little Farningham 144 LITTLE FARNINGHAM FARM, CRANBROOK, REVISITED during August 1999 during which several areas of higher resistance on the south, west and north sides of the fenced-off scheduled site were noted. Subsequent observation after ploughing of the land around this area revealed quantities of broken tile on all three sides with the greatest concentration on the northern side. The concentration of tile debris would possibly account for the higher readings encountered during the resistivity survey. Large fragments of tile, identical to the pedalis used as flooring material in part of the 'room' in area A, were abundant in the ploughsoil abutting the fence on the northern side. This would seem to suggest that they were originating from some unexcavated feature adjoining [D] and the area of the tiled drain. The field adjoining the western side of the fenced site appears very black with soil containing much waste from iron working. This includes tap slag, cinder, fragments of hearth bottoms, charcoal, pottery sherds and clay furnace lining extending out for a distance of at least 30m westwards. This area also corresponds with a zone of slightly higher resistance noted from the geophysical survey. The area has all the appearances of being a waste heap and indeed, as can be observed from the site plan (Fig. 2) this material had been noted as having extended into the western edge of the original 1950s' excavations. The Excavations Although the survey failed to provide evidence of extensive building it did suggest a possible alternative course for the Roman road, the location of which had been postulated by Margary,22 lying approximately 80m west of Ommerden stream. As a result of this new evidence, combined with the results of a surface artefact study carried out during October 1999, two trenches were machine-excavated, one across the line of the alternative course of the Roman road and a further shorter trench to test for any features that might lie close to its eastern side (Fig. 3). The limited excavation took place after the harvest during August 2000, timed to be carried out within the limited window of opportunity between the cereal harvest and the subsequent cultivation of the field. Trench A, 21. 7m in length and Im wide, was set out at right angles to the line of the road with a second, Trench B/C, 13.5m long and Im in width, laid out across its eastern terminus. It was envisaged that Trench A would bisect the postulated line of the Roman road as it was possible to discern an 'agger' running southwards from the crossing of the watercourse at Folly Gill to within a few metres of the northern side of this trench (see below). At the western end of Trench A, ex tending for 10.6m in width was a compacted 'road' surface lying at 0.30m below the surface of the field. The surface consisted of clay interspersed with natural iron 145 NEU. ALDRIDGE w.11 of TI"""" 81,ll!hfl1t9 0 .., C ,,o.. . Po■thrlle.■ 2,S. Fig. 3. Plan of Excavation, August 2000. nodules, cinder, charcoal, burnt clay, and pottery sherds. There was a definite limit to this surface at its western and eastern ends, although there were no ditches to delineate the edges. This feature was not the iron slagged surface that had been expected, however. The metal detector survey had suggested that the iron slag along the line of the road was not continuous. No iron slag was present in the topsoil in the area of Trench A so it appeared unlikely that it had been removed by earlier ploughing.23 At a position I 0.9m from the western terminal of Trench A, on the eastern edge of the assumed road, was an occupation surface with a I O- I 5cm thick spread of domestic/industrial material. Finds included sherds, burnt clay, iron slag/cinder, together with two iron nails. At I 3.2m from western end of Trench A was the base of an iron smelting furnace (Furnace 1 ). The north end of the furnace protruded from the baulk by some 0.7m with an average width of 0.5m. Four sherds of Roman pottery and a fragment of tile show that this feature was contemporary to the other features recorded here. It was comparable to others of similar furnace-base pattern excavated previously by the writer at Little Poplar Farm, Ulcombe which range in date from the Middle to Late Iron Age. The ground surface on which this furnace lay was 45cm below the topsoil and only the furnace-base survived. Cutting into this surface some 17m from the west end was a gully containing a large quantity of domestic material (Fig. 3). The fill con- 146 LITI'LE FARNINGHAM FARM, CRANBROOK, REVISITED tained some 376 pottery sherds, 6 nails and 2 pieces of tile. This gully averaged 32cm in depth and extended to 19.8m from the western terminus of Trench A. Abutting the eastern edge of the gully was a deposit of clay, 0.3m wide x 0.9m in height, which may represent the material originally excavated from the gully. The ditch fill contained the bulk of the pottery recovered during the excavation and included the major part of a large domestic vessel. The pottery is first- and second-century AD and much of it is grog-tempered native type material. It only proved possible to excavate the entire width of the gully at the eastern side where it was 0.9m. A posthole (PH l), 30cm dia x 12cm deep, was sited 0.4m from the southern edge of the gully. No further features were visible in the remaining l. 7m of Trench A. The southern portion of Trench B contained the bases of three further iron smelting furnaces of similar type to Furnace 1 in Trench A. Only the eastern ends protruded from the face of the trench. Furnace 2 was 0.7m wide at 0.4m from its east end it was associated with 2 pieces of tile. Furnace 3 was of similar dimensions and contained pottery sherds; Furnace 4 was 0.5m wide at 0.5m from its eastern end and contained a few pieces of iron slag. The surface of the trench was 0.3m below the topsoil as in Trench A. Fragments of furnace lining and slag were present on the surface of the trench and in the ploughsoil. Trench C produced evidence of a probable timber building of domestic type. This came from a series of postholes extending across the machine-excavated surface, together with a clay oven identified on the west side of the trench. An alignment of postholes was situated across the southern end of Trench C extending NW-SE. The surface of this deposit was some 0.45m below the topsoil. It consisted of a compacted surface with pottery sherds, iron nails, glass, tile fragments and burnt clay. This deposit is the interior floor of a probable domestic building of the late first-second century AD. The wall of the timber structure was represented by 2 postholes, PH 2 and 3. Further postholes existed north of the wall; PH 4 was the largest found during the excavation with a diameter of 38cm and a depth of 22cm. It is possible that it contained a supporting timber for the upper structures of the building. The structure appears to have been aligned NW-SE. A well-preserved oven of burnt red clay was exposed in the western side of Trench C. The rear of the oven sloped back from an inner chamber that measured approximately 0.5m square. The walls of the oven survived to a height of 0.8m above the floor of the building. The pottery from this trench was largely stratified on the trodden clay floor surface and clustered around and within the larger postholes, particularly PH 4. This also contained large fragments of clay oven material similar to the extant oven located some 0. 7m to the north 147 NEU.. ALDRIDGE within the building. Further pottery sherds were associated with the oven and these made up a total of 161 sherds from Trench C. A total of 7 pieces of Roman tile were stratified in Trench C, the largest fragments served as packing within PH 4. Further tile fragments were part of the somewhat deeper fill, some 0.25m which lay at the extreme northern end of Trench C. A total of 13 nails were also recovered from this trench. Although it was only possible to excavate a small part of the timber building within Trench C it is probable that part of its eastern exterior wall was exposed. PH 4 may have contained one of the main supporting posts of.the structure as it was sited barely 60cm from the inner edge of the assumed outer wall. The oven was situated 1.1 Orn from this same wall. The ditch excavated in Trench A was probably contemporary with the timber building and was being utilised as a refuse dump for domestic material. The single posthole located in Trench A, PH 1, was not part of the timber building and represents part of another feature which remains unexplored. The position of the timber building is significant as it lies immediately adjacent to the Rochester-Bodiam Roman road, on the east side. Although the western end of the structure was not found, the oven which presumably would have been sited close to the centre of the building, assuming it was contemporary with the road, would have been barely 10m from its eastern edge. Other Investigations The necessarily limited excavation was undertaken primarily to locate the exact position of the Roman road where it passed close to the earlier excavations. The course of the road had been suggested by the earlier fieldwalking and metal detector survey, which produced clear evidence of significant metalling from the road which, despite being spread north-south across the arable field by ploughing action, was still largely concentrated within a 20m wide track. This material was made up largely of waste material from iron working, principally bloomery tap slag. Along the amended road line a thin scatter of Romano-British pottery extended for some 350m north, with Samian sherds being found as far as the crossing of the stream at Folly Gill, and for some 150m to the south towards Chittenden (Fig. 4). There was a much greater concentration of pottery close to the new line of the Roman road some 110m west from the scheduled area. The spread of pottery in the plough-soil here included a number of sherds of Samian ware. The surface of the Roman road produced a sherd of Samian ware, a Form 27 bowl, dated to 50-170 AD. Augering along the line of the road indicated a horizon of dark 148 LITTLE FARNIN'GHAM FARM, CRANBROOK, REVISITED I • I • • • '* I . 1 • I I I I I I I I j f /F .UU.lU.l ' 3 ; • I "* I I ' •' I I I I ' I . ' ' I l ' ' .. * 4 B C o 200m c:::==::::c=:::::::c:==::::::5 Fig. 4 Plan showing Little Farningham Oast [A]; actual course of RochesterBodiam Roman Road (BJ; course as postulated by Margary [C]; I 956-60 Excavation area [D]; 2000 Excavation site [E]; Earthwork marking possible northern limit of Roman site [F]. Coin find spots are indicated by asterisks. 149 NEIL ALDRIDGE occupation material, 9m east of the road. It was noticeable that occupation material did not extend across the line of the road, lending further support for the road's presence at this point. Within the occupation zone was the timber structure of posthole construction together with a gully of contemporary date (as described above). The discovery of the four smelting furnaces was unexpected. It is not clear whether this area was contemporary with the structures found in the 1950s' excavations, although it would seem likely that they were part of the same complex. Certainly the timber building would have been in a more suitable position for any domestic settlement associated with the metal working. It was positioned such that the prevailing wind would normally have carried toxic fumes away from the settlement zone. Also it was adjacent to the road and on higher ground above the stream lying to the east. The domestic nature of the structure is suggested by the quantity of pottery within the building and in the adjacent gully together with the well preserved domestic clay oven within. A number of fragments of tiles were present on the floor of this building, and re-used as packing for some of the postholes. However, there was no suggestion that the building itself had been roofed with them; indeed, as there was not a continuous layer sealing the floor it appears unlikely that it had. Observations made during fieldwalking showed that there were further areas with concentrations of Roman tile and other occupation material. The first was in the area between the scheduled site and the road (Fig. 4, Area 1); this extended westward for 30m from the site of the earlier excavation and also contained large concentrations of waste from iron working. Indeed the soil bore all the hallmarks of being the principal dump for the material from the industrial activity taking place within the area immediately to the east. This area produced a sherd of Samian ware, a Form 38, AD 125+. Area 2, lying on the northern side of the scheduled site, contained large fragments of tegulae and pedalis. This area also produced high readings during the resistivity survey. Area 3 (Fig. 4 ), closer to the site of the Aug 2000 excavation and thus on higher ground, produced large concentrations of Roman building debris. This area appeared to be associated with a prominent earthwork bank which extended out from the Roman road and was traceable for some 80m east towards the stream (F on Fig. 4). The building debris appeared to respect this feature as it did not continue beyond the bank. It is possible that the bank represents the northernmost limit of the settlement. Area 4, 40 m south-west of the 1950s' excavation, produced a further concentration of pottery, Roman tile, and brick. 150 LITTLE FARNINGHAM FARM, CRANBROOK, REVISITED The metal detector survey resulted in the recovery of four Roman coins and other finds. Only two coins were identifiable: coin I, a denarius of Trajan, AD 98-117 details uncertain, base silver; coin 2, a brass dupondius of Marcus Aurelius as Caesar, AD 139-161. The find spots of these coins, together with coin 3, a bronze dupondius or as of the first-second centuries, and coin 4, of similar date and type, are shown in Fig. 4. Two other finds were also made; a lead seal or weight, probably Roman, found along the line of the road, and a copper alloy nail (Fig. 4), again quite close to the road. The Course of the Rochester-Bodiam Roman Road in the Vicinity Fig. S shows the line of Margary' s road and the correct route obtained as a result of the fieldwork undertaken in connection with the Little Farningham site. The road appears to pass through the village of Sissinghurst [marked as A] just to the west of the Bull Inn. Margary' s alignment heads down the hill to the west of the present Chapel Lane (where iron slag can be seen in the fields). It was also seen and recorded near the crossing of the Crane Brook in January 1936 by Major E. Clarke. The road there was found to be 10 feet wide with large stones at the sides to assist drainage. Slag and iron cinders were used for metalling here. The road was also seen in the garden of Golford Oast and Margary states that there was a sandstone block erected there to mark it [see B]. From this point it now seems that Margary's postulated line is incorrect. He felt that the road veered slightly south-west and joined a sunken lane heading south down to Folly Gill (part of the old road, pre eighteenth-century, from Benenden to Cranbrook which fell out of use after the construction of New Pond Road). It now appears much more likely that the road continues on a more south-easterly course from Golford until it reaches the crest of the ridge (TQ 8010 3610), to the west of Tollgate Farm [C on Fig. 5]. Here the road must have been laid out to the next elevated point which is now occupied by Benenden School. The alignment was also adjusted so that the road descended the slope at the easiest point. From the top of the ridge down to the stream much iron slag can be seen in the ploughed fields. Folly Gill is crossed at TQ 8010 3557. Prom here the road (having crossed the continuation of Margary' s line) turns slightly to the south-west and enters the field in which lies the Romano-British iron working site. There is a hollow close to the southern bank of the stream at Folly Gill with a concentration of iron slag indicated by the metal detector. An agger is discernible across the field to where the road was sectioned during August 2000 [BJ. 151 NEIL ALDRIDGE 0 6oom ------------J Fig. 5 Plan showing the amended course of the Rochester to Bodiam Roman Road. Margary's postulated alignment is shown by dotted line. [A] Sissinghurst; [BJ Golford, where the correct route diverges from Margary's; [CJ the sighting point for the next section to Benenden; [DJ site of original excavations; [EJ site of 2000 excavations. 152 UTILE FARNINGHAM FARM, CRANBROOK, REVISITED (Margary' s line lies further to the east and nearer to the site of the 1950s' excavations[D].) It can be traced on either side of this point. This is particularly noticeable when the field has been newly sown with a cereal crop. This was also confirmed by the metal detector which found heavy concentrations of iron slag in an alignment, averaging 20m wide, up to and beyond Folly Gill. The lack of metalling in the excavated section suggests that the road only needed surfacing in areas that became impassable in wet conditions. FINDS FROM THE RECENT EXCAVATION AND FIELDWORK Pottery A total of 456 sherds were found during the excavation of Trenches A, B, and C. An analysis of the fabrics showed that there were six basic types: grogtempered Patchgrove type ware, red thin ware, grey thin ware, orange sandy ware, white colour coated, and Samian ware. Mortaria and amphora were also represented in the assemblage. Grog-tempered wares, having a soapy texture, made up a total 70 per cent of the total. This type of pottery appears to have continued in use for a much longer period in the Weald than elsewhere in Kent. In the Sussex Weald, where this type is referred to as East Sussex ware, the same appears to be true. If found without any other more typical Roman fabrics it would probably be reasonable to assess the site as pre-Conquest in date. 24 It might be advisable to refer to this pottery type as East Sussex ware even though it is now also found in the Kentish Weald. It represents a preRoman native pottery tradition that seems to have continued in the Wealden region well into the third century AD. The pottery found during the fieldwalking survey reflects a similar pattern of fabric types and date. There was no evidence of any occupation on the site earlier than the first century AD apart from a few struck flint flakes of probable Neolithic or Bronze Age date. Building Materials The excavation produced less tile than might have been expected given the earlier archaeological history of the site; but no CL BR-stamped types (nor indeed any tiles of the very distinctive clay of Peacock's Fabric 2). A total of 18 small fragments were recovered from Trenches A, B, and C. Three were found on the surface of the road in Trench A, a further three were from the occupation surface between the road and the gully, two were from within the gully, and six were in Trench C. The remaining two tile fragments came from the smelting hearths. The largest fragment of tile was in the upper fill of PH 4 in Trench C. The majority of this material appears to be of a relatively crude fabric possibly of local origin unlike that found on the site of the original excavation, and during the recent fieldwalking survey. A total of 48 examples of tile were found during the latter. Out of these a total of 5 fragments of 153 NEIL ALDRIDGE tegulae and imbrices were present in area [l], the 'industrial' zone; in [2], 18 tegulae, l imbrica, 6 pedalis and I brick; in [3], 6 tegulae, 5 imbrices, 2 flue bricks and 2 other types of brick. Finally in area 4 there was one example each of fluebrick and plain brick. Metal Objects A total of 23 iron nails of varying lengths were found during the 2000 excavation: Trench A (surface of road): I x length 32 mm (fragment). Trench A (gully fill): 2 x 52mm; I x 62mm; 2 x 70mm; I x 135mm. Trench C: 3 x 30mm; I x 31mm; 1 x 38mm; Ix 44mm; 2 x 48mm; 1 x 50mm; 1 x 52mm; l x 53mm; 1 x 62mm; Ix 75mm; Ix 80mm. A nail of copper-alloy, 50mm in length was found during the metal detector survey west of the site of the 1950s excavations. A probable lead weight was found during the metal detector survey along the newly discovered section of Roman road. The Romano-British site at Little Farningham was certainly producing quantities of smelted iron obtained from the iron-rich local outcrops of Tunbridge Wells sandstone during the late first through to the mid second century AD. A number of shallow quarry-like areas still exist on the eastern side of the Ommerden stream, and these may well date from the Roman period although no positive evidence has been found to confirm this. Most of the waste material from the iron working would presumably have been utilised for surfacing parts of the Roman road. It is unlikely that the material would have been transported any great distance so the plentiful nature of this material in the vicinity suggests that a number of other, sizeable iron working sites might await discovery in the area. It appears unlikely that any major stone structures are present in the immediate area of the site. This is in spite of the notable presence of the CL BR stamped tiles. As has already been remarked, no substantial stone buildings were found either at the similar Bodiam site when this was briefly examined in the 1960s. There is thus no evidence to support the suggestion of a villa or bath-house on either site. The structures here probably had a short period of use and would most probably have been mainly of timber construction. At both Little Farningham and Bodiam flue tiles were being used for drainage purposes. All of the CL BR stamped tiles from this site were manufactured 154 LITTLE FARNINGHAM FARM, CRANBROOK, REVISITED from a distinct type of clay which occurs principally in the area between Guestling and Fairlight. 25 The precise location of a Roman tileworks is as yet unknown but a site somewhere to the north of Guestling or Westfield in East Sussex would perhaps be most likely, having easy access to the sea via the Brede valley. How the CL BR tiles came to be at the inland sites is a matter for conjecture, possibly simply as a result of the contemporary road transport not wishing to return to the iron working sites empty. The occupation of the Little Famingham site appears to have ceased by the second half of the second century. No evidence was found for any later Roman settlement there. The scheduled area within a fenced enclosure is now covered by a thick growth of trees and shrubs. Some of the trees are mature specimens which are probably causing considerable damage to the remaining archaeology there. ACKNOWLEDGEMENTS The writer would like to acknowledge the assistance of the following organisations and individuals for their help during the course of the fieldwork and related research. The Kent Archaeological Society and the Fieldwork Committee for their support, both financial and otherwise; Ted Connell, the Chairman of the Fieldwork Commitee, for his help; also Wayne Coomber for his work with the metal detector during the field survey and excavation and for showing the writer common sense when he became bogged down in the complexities of the site; Friday's, the landowners, for allowing the fieldwork to take place. Several members of the Cranbrook Local History Society assisted with the resistivity survey as well as other work at the site. Betty Carman of the Cranbrook Museum helped greatly when the writer was working on the site records in the archives there. Benenden School staff provided other help when working on the documents in their library. As well as contributing this paper to the memorial volume, the writer also would like to dedicate it to the memory of the late Cecily Lebon who first recognised the significance of the site, and would have been pleased to see this fuller account in print. He is grateful for the encouragement she gave to his interest in the archaeology of the Weald. NOTES 1 Salway, P., 1993, The Oxford Illustrated History of Roman Britain, Oxford, 444. 445: Detsicas, A. P., I 983, The Cantiaci (Gloucester), 173: Cleere, H. and Crossley, D., 1995, The Iron Industry of the Weald, Cardiff, 58-65. 155 NEIL ALDRIDGE 2 Peacock, D., 1977, 'Bricks and Tiles of the Classis Britannica: Petrology and Origin', Britannia, 235-248: Salway, op. cit. (see note I), 445: Brodribb, G., 1969, 'Stamped Tiles of the Classis Britannica', Sussex Archaeological Collections, cvii, 104, 109, 111. 3 Drewett, P., Rudling, D. and Gardiner, M., 1988, The South East to AD 1000 (London), 239-240: Cleere, H., 1978, 'Roman Sussex -The Weald', in Archaeology in Sussex to AD 1500, CBA Research Report, No. 29, 59-63: Brodribb, op. cit. (see note 2), 108-109, 111: Peacock, D., op. cit. (see note 2), 236. 4 Geological Map, 304. Institute of Geological Sciences, 1974. 5 Archaeologia Cantiana, 71 (1957), 224; 72 (1958), !x-lxii; 73 (1959), xlvi. 6 Scheduled Monument No. KE 169. 7 Cranbrook Museum, file of papers relating to Little Parningham Farm Roman site including correspondence and site notebooks. See letter dated 29/5/59 to Professor Ian Richmond. 8 Cranbrook Museum letters and other documents relating to Roman site dated 29/5/59; ! 2/8/59; 8/2/60. 9 The writer had previously assisted Mrs Lebon during her investigation of the Roman ford at Iden Green, some 3km to the south of Little Farningham; see Archaeologia Cantiana, IOI (1984 ), 69-81. 10 Peacock 1977; Brodribb 1969, op. cit. (see note 2 above). 11 Cranbrook Museum, file of documents relating to the Roman site. 12 Archaeologia Cantiana, 72 (1958), lx-lxii. 13 This mirrors what has emerged from recent fieldwork undertaken by the writer in the area around Headcorn and Ulcombe in the Low Weald where grog-tempered pottery traditions seem to have continued well into the third century AD. 14 Archaeologia Cantiana, 73 (1959), xlvi. 15 Journal of Roman Studies, 1960, interim report, 235. 16 Cranbrook Museum, site notebook 1958-59, M. C. Lebon. 17 This feature is virtually identical with those interpreted as iron smelting hearths (dated to the late Iron Age and Romano-British periods) recently excavated by the writer in Ulcombe and Headcorn (reports in preparation). 18 Cleere and Crossley I 995, op. cit. (see note I), 47-48. 19 The site had the appearance of a rough shanty type area. It has erroneously been stated elsewhere that 'a substantial stone-built structure with a hypocaust system' was found; this can now be shown to be incorrect. Ibid., 296. 20 Lemmon, C. H. and Darrell Hill, J., 'The Romano-British Site at Bodiam', Sussex Archaeological Collections, I 04 (1966), 88-102. 21 Archaeologia Cantiana, I I 5 (1995), 458-460. 22 Margary, I. D., Roman Ways in the Weald, 1965, 220-222. 23 The surface was similar to the sections of road recently excavated at the RomanoBritish settlement at Westhawk Farm, Kingsnorth, to the south of Ashford. It is of similar dimensions to that at Little Farningham and despite passing through a settlement with significant iron working activity it also appears not to have been surfaced with any significant quantity of slag. (Observations by writer during site visit, September 1998.) 24 At a site in Head corn excavated by the writer a Romano-British farmstead produced a total of 1058 sherds, of which 577 were soapy grog-tempered fabrics. 25 Peacock 1977, op. cir. (see note 2). 156