Excavations at Eccles 1973

EXCAVATIONS AT ECCLES, 19731 TwELl!Tll INTERIM REPOltT By A. P. DETSIOAS, M.A., F.S.A. lNTRODUOTION ExoAVATIONS were resumed, for the twelfth successive season, under my direction and on behalf of the Eccles Excavation Committee, at the site of the large Romano-British villa. located at Rowe Place Farm, Eccles, in the pariah of Aylesford (N.G.R. TQ 722605; O.S. 6-in. Sheet TQ 76 SW) and at another nearby site (N.G.R. TQ 718605; O.S. 6-in. Sheet TQ 76 SW); this work was carried out at weekends, from the end of March until the end of October, 1973, and during a fortnight in the summer when a training course in Romano-British archaeology and excavation techniques, jointly sponsored by the Kent Archaeological Society and the Eccles Excavation Committee, was conducted on the site. I am, once more, greatly in the debt of the landowners, Messrs. Associated Portland Cement Manufacturers Limited and Messrs. Reed Paper & Board (U.K.) Limited, and in particular to Mr. A. W. Baker, their Chief Engineer, for readily allowing this excavation on their property; I am also grateful to their tenant farmers, Messrs. A. A. & A. C. Southwell, for the wholehearted support of our work and many acts of kindness. Financial assistance for this excavation was again granted by the Kent Archaeological Society, the British Academy, the Haverfield Bequest of the University of Oxford, the Society of Antiquaries of London, and private donors, without whose generosity this work could not be undertaken. This excavation was carried out entirely by volunteer labour to whom I acknowledge my debt, though space limitations preclude individual mention; however, I am glad to be able to make- exceptions in the case of the following, for their sustained support throughout a long campaign: Mr. A. C. Harrison, B.A., F.S.A., who again shouldered much of the supervision of this work; Mrs. P. M. Winzar; Misses C. E. 1 Arch. Oam., lxxviii (1963), 125-41; lxxix (1964), 121-35; Jxxx (1965), 69-91; lxxxi (1966), 44-52; Jxxxii (1967), 162---78; lxxxiii (1968), 39-48; lxxxiv (1969), 93-106; lxxxv (1970), 55--60; lxxxvi (1971), 25--34; lxxxvii (1972), 101-10; and lxxxviil (1973), 73-80. I am very indebted to Professor S. S. Frere, M.A., F.B.A., F.S.A., for his continued interest in this work a.nd reading through this report in draft to its great benefit. 119 A. P. DETSICAS Cockram and R. C. Symes; and Messrs. A. J. Daniels, O. K. Hales, T. Hetherington, T. Ithell, B.Eng., W. A. Knowles, C. E. J. Martin, L. Thomsen and P. Thornhill, B.A. I am also grateful to Miss D. Charlesworth, M . .A., F.S.A., for reporting on the glass; Mrs. K. F. Hartley, B.A., F.S . .A., for reporting on the mortaria; Mr. M. W. C. Hassan, B.A., F.S.A., for reading and reporting on the graffiti; Mrs. S. C. Hawkes, M.A., F.S.A., for reporting on the Anglo-Saxon material and arranging for its conservation at the Institute of Archaeology, University of Oxford; Mr. M. R. Hull, M.A., F.S.A., for reporting on the brooches; and Dr. J.P. C. Kent, B.A., Ph.D., F.S.A., for identifying the coins. I am also particularly indebted to Miss R. C. Symes, for undertaking much of the surveying on the site, and Miss S. J. Marsh, for much help in the compilation of this report. THE EXOAVATION Excavation this year wa.s continued beyond the points where it was suspended in 1972, in the area of the south-west (baths) wing and another site close to the villa; the results of this work are reported below within the chronological period to which each feature is assigned. I. LATE htoN Ao:m/RoMANo-BRITISR Period I, to c. A.D. 55 Ditch IX (Fig. I) This ditch was sectioned south of the point reached in 1972 ;2 its width remained constant at 7 ft. 6 in. (2 · 25 m.) and its traced length to 179 ft. (53 ·70 m.). Ditch X was traced further to east, below the Anglo-Saxon cemetery (see below, 129-30), and its known length increased to 192 ft. (55 · 10 m.); in its back-filling were found sherds from pottery manufactured on Site D (see below, 128-9). It was observedS that this ditch is reminiscent of military work as, apart from its V-shape, it had at its bottom a shallow cleaning channel, not unlike a rudimentary box-gutter. Ditch XIII (Fig. 2) This feature, first exposed in 1972 below the water basin at the centre of the villa's co¥tyard; 4 was again sectioned in trenches olose to the south-west wing; its width was 7 ft. (2 ·20 m.), a.a before, and the total length so far proved reaches i74 ft. (52·20 m.f. The dating of all three ditches is not definitely established but, as they have been found cut below structures known to belong to Period II, they must be part of the initial settlement of the site or earlier. To this early period of occupation is a.lso provisionally assigned until further work establishes its shape and purpose a short length (c. 15 ft. 6 in., 4 ·65 m.) of a narrow (2 ft. 2 in., 0·65 m.) curving gully exposed a Arch. Oant., btXl!:.vili (1978), 74, 1 Ibid.,'174. 'Ibid.,􀀉75. 120 /'. LCCLES jf--: i ; ,-.. - I iI r-􀀃 9J L.􀀄 J ·- -·- ___ ; I 􀀣----·--, r---·---·--" ·I Ill r l:JO I'; 􀁆Pwl'iod IV B,\IPttlOd IV1 Robl,tt.d Or lh I'.· IIDP..l'iod .,,R􀁃•d onnf•,,•d Iii PtrJod VI 􀁄Pulo•dorint,r,·td 􀁅 /IWr/, r·--1 j i i I, 􀁇riod .,l/rRo/>bfHI or iflltrr•d􀁈Adodt Vl•V 1F = ==-,: 􀀊---􀀋􀀌i.. ;;;􀀍- -, r·-·-·, I i I i L ____ J p; £, H#Orlh 711--! I r--7 ' I ®p;,: I , ' 1973 ' ' ,.., i i I I i I I Dire Si1c f) / r-7 I I I Mtditt1ol 􀁄;,􀁅 1 I ·-c _______ .ll·"·i ;Ditch ·"J/ , I ! I L- !"le:-::rs:::i151!'1.. ----􀀊􀀋-5;;;==-"" =-==,====􀀌􀀍===􀀎 L_ti1i!J i A14P'1' Frn. I. I.fore JJ. 120 ECCLES L 1972-73 l I I ! j L_j j i \ I ----- -- ,;.,11, I j -·- - - - =·=- -_ = =- =-=·=---=:: _ J -, I I i i I I r--- 1. i i i L L. __ j i- - --􀀏--􀀐-􀀑 􀀒-=-_-:-::-_:! COURTYARO AREA %::I:: s,._,-,Mo•.inl•,r•dL -.J -------- - - - ,M,_,,,􀀲 􀀶 lt ,.,.;􀀳 .. -•••􀀴􀀵,- ··- __ :J' -- ------- , 􀀶 ., J FIG. 2. - __ _ jFw. 3. [/4<'• J>. 121 .... ___ _; -·-· ----•-, _____ ,_; f EXCAVATIONS AT ECCLES, 1973 just beyond the north-western lip of Ditch X; it had been dug into the subsoil and came to an end in the area occupied by the Anglo-Saxon burial ground. If it continued, after a gap, further to west, it would have been removed by the deeper excavation required for the construction of Room 121.5 Period II, c. A.D. 55-65 In the courtyard area (Fig. 2}, the mechanical trenches in 19726 were extended to north-west and the two furthest to east proved abortive; occupation evidence, however, was found in the area contained between the two furthest to west. This evidence consisted of an area, to south-east of the villa's later (Period V) south-west wing, which had been laid with a floor of ye1Iow mortar, averaging 1-2 in. (0 · 025-0 ·05 m.) in thickness and placed, at a depth of c. 1 ft. 6 in. (0 ·45 m.), directly on the Romano-British plough. soil; further trenching, established the north-east, south-east and southwest limits of this floored area but, because of much disturbance at depth during medieval times, it was impossible to prove its north. western limit which may not have extended quite to the line of the south-west wing. Furthermore, though the limits of this mortared area were clearly defined, it proved impossible to establish the outlines of the sleeper-beam construction trenches which, on the basis of the partitioning sleeper beam, delimited the whole area. If the floor did reach the line of the south-west wing, it would have occupied a total area of 41 ft. 6 in. by 15ft. (12·45 by4·50m.). This floored area was divided by a wooden partition, constructed almost at its centre. This has survived as the construction trench for its sleeper beam out through the mortar floor; it did not, however, divide the entire width of the floored area as it turned to north-west in order to allow for a narrow (3 ft. 6 in., l ·05 m.) conidor-like passage. The two larger compartments communicated through a gap, clearly a doorway, which interrupted the sleeper-beam trench and measured 4 ft. (l ·20 m.) in width. The purpose of this structure is problematic; its flooring and construction are not substantial enough to indicate a dwelling-house, and it seems more likely that it may have been a workshop or outbuilding. Its dating, too, is debatable as no finds at all were recovered from below the mortar floor; however, it fits more easily with the excavated evidence to consider this building as belonging to an early phase in the occupation of the site. For the whole layout of the villa's courtyard, with an ornamental water-basin at its centre7 and, •8 Ibul,., lx=vi (1971), 29. Ibid., Ixxxvili (1973), fig. 2. 1 Ibul,., 77. 121 A. P. DETSICAS probably, flower-beds and paths leading to the main range of rooms as well as to the two wings of the house, makes it very unlikely that such a flimsy structure, whatever its purpose, would have been allowed to stand in this courtyard area when the villa was reconstructed to face to south-west.8 It is more likely, and logical, that this outbuilding was in use at the rear of the villa, i.e. very probably during Period II, certainly not later than the close of Period III when the major rebuilding of the living accommodation of the house must have brought about the demolition of this outbuilding. Baths Area (Plate IA) Re-examination was also undertaken in the baths area of the villa (Fig. 3). In 1962, work was suspended at the level of the hypocaust underfloors of the third baths belonging to Period V;ll however, when excavation was resumed in 1963 and 1964 to north-west of the previous areas,1° it was soon found that earlier structures had been present below those later hypocausts, but their examination was deferred until the rest of the villa had been completely excavated. This year, the backfilling and the hypocaust underfloors were mechanically removed and the excavation completed to subsoil depth.11 As a result of this ree:x: amination, several new features and details can now be added to the published plans, and a partial reconsideration of the dating proposed in previous reports becomes necessary. Period Ila. It is now clear that a small structure, of unknown purpose, was first built in the area to be occupied by the later (Period V) Rooms 20 and 2312 against whose north-west wall was later abutted the caldarium (Room 43) of the first (Period III) bath-house; the foundations for the south-east wall of this early building were exposed, below and slightly protruding to south-east of the north-west wall of Room 21, and later out by the drain of the second baths (Period IV). This building would have mee.sured 13 by 14 ft. (2·90 by 3·20 m.) internally; it was subsequently superseded by a much larger structure, built partly over the original building and partly incorporating its walls. Period Ilb. This new building was first recorded lll 196412 as a. fairly substantial wall preserved below Rooms 44 and 20 (Period V). Only foundations (Plate IB), however, survived below the Period V hypooausts, and the plan of these foundations indicates a. large room, a Ibid., 1.xxxvi (1971), 33. t Ibid., lxxviii (1963), fig. I, 132-5 and 138-40. 10 Ibid., lxxix (1964), fig. 1, 71. 11 J am indebted to Mr. A. C. Harrison, B.A., F.S.A., for the general supervision of the work in this area, and to Miss R. O. Symes, for its reoording. u Arch. Oant., 1.xxx (1965), fig. 1, 122· EXCAVATIONS AT ECCLES, 1973 measuring internally 45 by48 ft. 9 in. (13·50 by 14·625 m.) and flanked to south-east and south-west by what appears to be a corridor below Rooms 22 and 133 (Period V); this large room had a doorway leading into it, as shown by a gap in the foundations at its south corner. The foundations were laid in construction trenches, c. 2 ft. 4 in. (0·70 m.) wide, and consisted of lumps of undressed ragstone bonded with yellow mortar and opus signinum: the solidity alone of these foundations suggests load-bearing walls of some height. Period llc. A reconstruction, which may prove contemporary with the building of the first bath-house (Period III), next took place, and this meant the demolition of the second building in this area. The south-east wall of the large room and the corridor were cut by the construction trenches of four sleeper-beams belonging to a small granaryls (Plate IB). The north-east sleeper-beam of this granary could not be located as it must have lain directly underneath the partition wall between Rooms 21 and 133 (Period V) whose construction trench is certain to have been dug much deeper than the level of the sleeperbeam and, consequently, removed all evidence for it; likewise, the south-west sleeper-beam must either have been located below the substantial south-west wall of the Period V baths or beyond it, hut this has not yet been confirmed.14 On the other hand, evidence a.gain was found, as a layer of soot and ash directly on the subsoil, for burning,15 which may be contemporary with the destruction of the first baths (Period ID),16 though this will have to be confirmed by a fuller study of the associated finds. In the north corner, below Room 21 (Period V) there was found a solid length of masonry, consisting of a foundation layer of ragstone under two courses of bonding-tiles, the whole construction set in stiff yellow mortar and opus signinum and cutting through earlier foundations; abutted on to this masonry was another fragment of wall, consisting of a foundation of ragstone and three courses of bonding-tiles, similarly set in opus si,gninum and presenting a. curved outline. It is not possible to interpret this ()()rtstruction satisfactorily: the bedding trenches for it did not continue beyond the points shown on the plan (Fig. 3), a.nd the solidity of the entire construction suggests a probable buttress, but for what purpose it is impossible to say. Perioa, IV, c. A.D. 120-180 The re-examination of the hypocausted area of the third baths (Period V) (Plate IA) has resulted in additions to the structural details 13 Imd., fig. 1, 70-71. u The former hypothesis was confirmed in 1974. This gra.no.ry need not, of course, have either north-east or south-west sleeper-beams. 15 Arch, Oattt., bcxx (1966), 71. U[bid., 85. A. P. DETSIOAS of the second baths (Period IV), which were obscured in 1962-63 by the presence of baulks at crucial points. It can now be shown that, in the original plan of the baths, Rooms 11 and 12 had formed one unit, served by a hypocaust; though this had been completely removed in subsequent demolition, the trowel-marks made by its builders on the wet opus signinum underfloor for the centring of the pilae were plain to be seen in 196217 and this year. The hypocaust underfloor was 3--4 in. (0 ·075-0 · 10 m.) thick and laid on a make-up layer of ragstone resting on the subsoil, and the hypocaust of the single room was served by flues through its north-west, north-east and south-east walls. There wa.s also an opening through the south-west wall, 1 ft. 8 in. (0·50 m.) in width, near the north-west corner of Room 11, which cannot have been a hypocaust flue; for there was no evidence at all to south-west of this opening for a stokehole outside the main wall of the baths, nor does it belong to the later baths (Period V) as the opening had been blocked by the thickening of the south-west wall when the latter was incorporated into the new bath-house. It can only be suggested that this opening may have served for the cleaning of the hypocaust or as a means for increasing its air-flow. Period IVb. At a later date within this building period, the area occupied by the single room was divided into the two Rooms 11 and 12 by the construction of a partition wall of bonding-tiles; though this partition had later been completely robbed, the imprints of the lowest bonding-tiles were clearly visible on a thin layer of opus signinum laid on the existing underfloor for the purpose of bonding this partition to the floor. That this partition wall is a later addition is further demonstrated by the fact that, at the points where it abutted on the north-east and south-west walls of these rooms, the face of the walls was still smoothly finished; moreover, opus signinum had been used to seal the junctions of wall and underfloor, on both sides of the partition, and smoothed round inside the corners thus formed. The original north-west wall of Room 11 was demolished and a new one built 4 ft. (l ·20 ni.) further to north-west and to the same width of 2 ft. (0 -􀀃o m.); a g􀄅p. 4·ft. (l ·20 m.) wide,.w.as 1'.ecorded in the foun􀄆a.­ tions for this new wall survi'ving its demolition when the south-east wall of Room 42 (Period V) wru:i built ·partly on top,· and the opus signinum underfloor had been carried through into this space. This gap in the foundations is much too wide for a single flue but of suitable width for a double flue with a short length of intervening wall; if so, subsequent demolition has removed direct evidence. As a result of this reconsideration, the size of Room 12 remains 17 lbw., lxxviii (1063), 134. 124: EXCAVATIONS AT ECCLES, 1973 unaltered, but the length of Room 11 is extended to 15 ft. 9 in. { 4 · 725 m.).18 Room 132 belongs to this reconstruction, too. It was a small compartment, measuring internally 5 ft. 3 in. by 4 ft. 6 in. (l ·575 by l ·35 m.), excluding the corners reinforcing its abutment on the southwest wall of the baths, and very solidly built. Its north-west wall was 2 ft. 5 in. {0·625 m.) thick and the south-west one 3 ft. 3 in. (0·975 m.), with a 9½ in. (0·24 m.) offset; the south-east wall was overbuilt by the Period V baths' wall, but examination at depth showed that it was of the same width as its north-western counterpart. This room was floored with opus signinurn, which had been ca;rried up to the face of its walls; this floor was 5 in. (0 · 125 m.) thick and laid on a solid foundation of mortared ragstone resting on the subsoil (Plate IIA). During the construction of this room, the south-west wall of the bath-house was breached to allow for the insertion of an arched opening through it, and the new foundations went deeper into the subsoil than before; this opening, rather more than 1 ft. (0 ·30 m.) wide and I ft. (0·30 m.) high, was corbelled with bonding-tiles laid well below the level of the room. The opus signinum was continued into this opening and smoothed upward at its junction with the underfloor of Room 11 (Plate IIA); the floor in Room 132 was some 10 in. (0 · 25 01.) lower than the hypocaust underfloor in Room 11. The thorough demolition in the area of this room, which resulted in the loss of essential details, presents many problems of interpretation, and the suggestions put forward below cannot, therefore, be regarded as conclusive. It is not doubted that Room 132 waa intended to contain water: the solidity and thickness of its walls, clearly intended to withstand considerable internal pressure, their surfacing with opus signinum and the thickness of its floor show this quite clearly. What are debatable, however, are the purpose of this room and the nature of its water supply. The opening into Room 11 is obviously functional as is shown by the care taken in its construction. It was thought, at the time of its excavation, that the downward smoothing of opus signinum from Room 11 to Room 132 was an indication that water was intended to flow from Room 11 into Room 132; but this is not necessarily so as it is more likely, as mentioned above, that the opus signinum was smoothed wpward to seal the junction of the floors below the arched opening. That it is not likely that water flowed freely between Rooms 11 and 132 is further supported by the presence of soot and ash,19 deposited 18 In the definitive report on the site, the opport1UUty will be taken to renumber these rooms. 19 Arch. Oant., lxxviii (1963), 134. 125 A. P. DETSICAS ag􀃛inst the inner face of the south-west wall of Room 11; if water had been circulating, this obviously would not have occurred. The most likely interpretation compatible with the excavated evidence is that Room 132 was a small plunge-bath over a hypocaust, used in conjunction with Room 11, a tepidarium. No direct evidence was found for base- or pila-tiles for the suspension of this hypocaust, which must have been completely removed during the building of the subsequent baths in Period V; likewise, no evidence survived for any mason's marks on the underfloor for the siting of the pi'lae. However, in view of the small size of Room 132, such marks would have been superfluous as very few pilae would have been needed at the central part of the room to support the hypocaust's bridging-tiles, resting on the buttressed corners of the north-east wall and on the offset of the room's south-west wall. Room 132 would have been supplied with hot water from the furnace, Room 7,20 by means of the partly demolished opus signinum channel leading from this furnace to the west corner of Room 9 where it turns at right angles and points directly towards Room 132. In 1962, it was recorded that this channel 'outfiowed just beyond the west corner of Room 9' ;21 but, in the light of this year's new evidence, from this point onward the water supply could have been carried by pipes across the 7 ft. (2 · 10 m.) width separating this 'outflow' from the south-east wall of Room 132. On reaching this wall, the piping would have continued through the south-east wall of Room 132 and ducted into the plunge-bath. If this interpretation, based on the excavated evidence, of Rooms 11 and 132 as a tepidarium and a plunge-bath, respectively, proves acceptable, there remains the problem of water disposal. It is not impossible that water was drained out of Room 11 through piping across Room 8 and towards the main drain passing outside Rooms 4 and 2;22 on the other hand, even indirect evidence for the draining of Room 132 is lacking. Logically, it should have been drained into the substantial drain known to be passing underneath Room 19 (Period V),23 by means of a channel or, more probably as the subsoil between Room 132 and the drain was undisturbed, piping through the room's southwest wall; the demolition of this wall may have removed any direct evidence for such piping, though an indication supporting its probable existence is the fact that the top of the drain under Room 19 (Period V) was found collapsed at the point where the drain changes direction, i.e. ·at the point nearest to the west corner of Room 132. In any case, Room 132 could not be completely drained because its floor level was ao Ibid., fig. 2. 21 Ibid., 132. 21 Ibid., fig. 2. ta Ibid., 128-9. 126 EXCAVATIONS AT ECCLES, 1973 ·a.bout 1 ft. 6 in. (0·45 m.) deeper than the highest remnants of its south-west wall surviving slightly above subsoil level. The main drain (Plate IB) was completely cleared from the east corner of Room 19 (Period V) to the point where it cut through the south-east wall of the early building (Period Ila; see above, 122) under the later (Period V) Room 21; its construction was confirmed as recorded in 1962.24 The drain channel had been built of roofing-tiles laid on the subsoil, with their flanges uppermost and built partly over by the sides of the channel; this was 9 in. (0 ·225 m.) wide and internally faced with opus signinum. The walls of the drain were 5 in. (0 · 125 m.) wide, bedded in a construction trench and consisting of four courses of ragstone alternating with three courses of bonding-tiles set in yellow mortar. Period, V, c. A.D. '180-400 Resulting from the re-examination of this area, it is now established that Room 20 was not, as reported in 1963 and 1965,26 a single unit, measuring 6 by 24 ft. (l ·80 by 7 ·20 m.), but two separate rooms (Rooms 20 and 133), divided by a 2-ft. (0·60 m.) partition wall, which was completely masked in 1962 by its coincidence with a baulk (Plate IB). Room 133 measured 6 by 11 ft. 4 in. (l ·80 by 3 ·4-0 m.) and Room 20 is now reduced t o 6 by 10 ft. 8 in. (I · 80 by 3 · 20 m.). Further exploration below the level of the hypocaust underfloor in Room 42, apart from locating the line of the second baths' wall (Period IV}, has shown that the flue from the stokehole (Room 25)26 narrowed down from a width of 2 ft. 1 in. (0·625 m.) to 1 ft. 7 in. (0 ·445 m.), as it reached the subsoil; it had been laid with yellow mortar burnt to the colour and consistency of brick. Re-excavation was also undertaken of the corridor (Room 18) in the area to south-east of Room 132 (Period IV). This space was found filled with a large deposit of painted wall-plaster fragments, some of which had been incorporated in the building of the corridor's north-west wall on top of the Period IV wall, serving as a make-up layer below the floor of the corridor; underneath this deposit, there was another layer, resting on the Romano-British ploughsoil and consisting mainly of domestic refuse. In 1962, in a restricted space mostly occupied by the east corner of Room 18, part of this deposit was excavated and considered to belong to a rubbish pit dug through the corridor floor;27 it is now clear that, though only traces remained of the mortar beneath the tiling of the floor, this rubbish deposit dates the building of the baths, not its destruction, as previously thought . .A detailed study of the u Ibid., 129. 15 I/:!id .. fig. 3, 139: Jxxx (1965), fig. 1, 86-7. ae Ibid., Ixxix (1964), 131. 17 Rubbish Pit 0, Arch. Oant., 1xxviii (1963), fig. 2, 140. 127 A. P, DETSICAS pottery contained in this rubbish must await the final publication of the site; however, it is obvious that a re-dating of the close of Period IV and the beginning of Period V is inevitable. This coITected interpretation of the rubbish layer has, however, solved one problem: for, on the earlier consideration that the third bath-house (Period V) had been destroyed c. A.D. 290, it was assumed that yet another baths building may have been erected elsewhere on the site, though several following sea.sons' work had failed to locate any such structure. The re-examination of this area has now shown beyond doubt that no other bath-house was ever built on the site and that the Period V baths may have remained in use until the end of the Romano-British period in the early fifth century A.D. as the villa is known to have continued in occupation until that date by pottery and late coins found stratified in the main range ofrooms and the south-east wing.28 Site D, pre-Fl,avian This site (N.G.R. TQ 718 605) was discovered in 1971 whilst watch was being kept on the laying of a water pipeline for the Medway Water Board in a field close to the main site; it was first seen as a dark band exposed in the 60-ft. (15 m.) contour in an area almost completely disturbed since the middle of the last century by the excavation of large pits for the disposal of industrial waste. This site, on the east bank of the Medway, lies a.bout one-quarter of a mile (400 m.) west of the villa and about the same distance from the disused Burham Cement Works. On closer inspection, it was established that the dark band consisted almost exclusively of kiln debris and building materials, dipping from south to north towards the river, and all that had survived was a narrow plateau in the middle of the industrial excavations; as mechanical stripping of the area was intermittently undertaken by the landowners, a small-scale excavation was carried out late in 197220 and a systematic examination of the surviving area was completed this year. The area measured approximately 40 by 60 ft. (12 by 18 m.) and was mechanically cleared of its modern overburden beneath which it was found to be entirely occupied by a thick and compacted layer (c. 1 ft. 6 in., 0 ·60 m.) of Romano-British wasters (Fig. la}; industrial excavation had cut into the site from all directions and partly disturbed a structure which was identified as a medieval tilery (see below, 130-31). As a result of the excavation of the site, there is no doubt that largeas H, when the whole evidence is considered in detail, it is concluded that the third baths was out of use before the end of the villa's occupa,tion, it may be necessary to consider whether the partitioning of the villa's rear corridor into small rooms and the insertion of a channelled hypocaust beneath one of them (Arch. Oant., lxxxiv (1969), fig. 1, 100-104) was, in fact, the flnal bathing unit of the villa. Several sections were a.lao cut across the line of Ditch VIII (Arch. Oant., lxxxvii (1972), fig. 2, 106), located beyond the area. of the general plan (Fig. 1). u Ibid., lxxxviii (1973), 78-9. 128 PLATE I I')JOol. : A. P. Dttsirng A. 81\ths III (Period V): General View at Level of H_vpocaust Underflcors Photo: .t. I'. D('(siras B. General Yiew ucross Roo111s 10 and 13:l. (Period \") Hoon, 1\1 (jorei,ro11nd). •howing earlier Dmin ( Period I\·). Construction Tl'cnch of earl􀄛· Building ( Period llb) Rnd Slecpcr-bcurn Slot (Period I le): Room 133. showing Fo1111dnt ions of earl.,· Building (Period llb). with I>oorwa.v. and 􀄜l<'Pt>er-beum Slots (f'.-riocl llc-). I fart p. 12...: Photo: A. P. n,t,ica• A. Hoorn I 32, showi11i:z arched Flue through north-eastern \\'all Photo: A. /'. Dt/titOI B. :Site D: <:Pnt>r!\I \'i􀁧"- of ,nedio,·al Tiler.,· EXCAVATIONS AT ECCLES, 1973 scale pottery manufacture was carried on nearby; the site is conveniently situated for river transport, and the clay subsoil provided excellent raw materials. The detailed study of the enormous amount of waste material recovered has not yet been undertaken, but a preliminary examination shows that the greatest part of these wasters consisted of various forms of Hofheim-type flagons; also produced on the site were early mortaria,, colour-coated wares imitating imported Continental prototypes, sta.mped native platters imitating terra nigra and several other forms. This debris also contained wasters from a type of unusual vessel, with triangular openings through its wall and ba.se, which were also found in the back-filling of Ditch X (see above, 120), a potter's punch for the production of small plaques, probably for application on vessels, and one-half of a roulette for the decoration of pottery with horizontal bands of striations, which matched with wasters found in this deposit.so This industrial activity may well have begun very soon after A.D. 43, and the pottery types suggest that manufacture ended c . .A..D. 65-70; this is further supported by the total absence of Flavian and later coarse-pottery types as well as by pre-Flavian samian ware found stratified with this deposit. It is, of course, quite possible that potterymaking may have continued elsewhere in this area; it is, however, so badly overgrown and disturbed in recent times that it will need a major excavation to attempt to establish this possibility. Il. POST-ROM.AN, c. A.D. 400-650 Two ditches, Diwh XIV and Diwh XV,31 were further cleared below the Anglo-Saxon cemetery, and it was found that Ditch XIV merged into Ditch XV which continued to east and beyond the excavated area; the traced lengths of these two ditches are 57 ft. 6 in. (I 7 · 25 m.) and 64 ft. (19·70 m.), respectively. Ditch XV cut through a late Romano-British pit, filled with domestic rubbish, and it is now clear that both these ditches, as well as Ditch XVI which was partly explored in 1972,32 were cut after the end of the Romano-British period of occupation and remained open as late as c. .A..D. 650 when some of the earliest Anglo-Saxon burials were inserted in these ditches. Ill. .ANGLO-SAXON The Cemetery, later than c . .A..D. 650 Work continued in the area of this burial ground beyond the trenches cut in 19721 s2 and a further seventy burials came to light, as 80 Antiq. Journ., liv (1974), forthcoming, 31 Arch. Oant,, lxxxviii (1973), fig. 1, 77-8. SI Ibid., 78. 129 A. P. DETSICAS well as many scattered bones which could not be associated with any individuals. As previously, the majority of these inhumations had been buried very close to the bottom of the ploughsoil, laid approximately east to west, without grave goods and on top of other earlier burials; some of the Iatt.er had personal effects buried with them, including a complete vessel placed very close to and to right of the skull, which confirmed the initial dating of this cemetery to the middle of the seventh century A.n.ss Indications were also found that the north-eastern and south-eastern limits of the burial ground may be close to the excavated trenches, but this remains to be confirmed; to north-west and south-west, its area is delimited by the ruins of the villa. · IV. MEomvAL Site JJ A small medieval tilery (Plate IIB) was excavated in the middle of the Romano-British pottery waste debris (Fig. la, and above, 128-9). This tile-kiln, which measured overall c. 5 ft. 10 in. (l ·75 m.) square, had originally been constructed as a rectangular cutting into the clay subsoil serving, on the three surviving sides of the kiln, as its·wall and reddened by fire to a depth of 4 in. (0 · 10 Il1.) all round. Within this square cutting had been constructed a central spine crossed at right angles by the kiln's four flues whioh were all badly worn, collapsed and compacted by the weight of the overburden. The central spine had been built almost entirely of clay-bonded large lumps of ragstone and boulders and Romano-British roofing- and bonding-tiles; it was c. 1 ft. (0 ·30 m.) thick. The flue walls were constructed of clay-bonded medieval tiles, with some Romano-British material as well, and had an average thickness of 8 in. (0 ·20 m.); each flue wall had been pierced by two arched flues to allow for the transverse circulation of hot air and gases, each arch on either side of and springing directly from the central spine. These arches, too, had been built of medieval tiles arid were c. 9 in. (0·225 m.) wide; soot was found on the subsoil beneath these arches, as well as in the passages between the flue walls, under a deliberate back-filling of clean clay and debris. Little survived in situ of the kiln's floor, but there were indications that this, too, had been made of medieval tiles. Though the location of the tilery's stokehole is not in doubt, little had survived industrial excavation; this is doubly unfortunate as it, probably, removed any pottery contemporary with the use of the kiln, which would have confirmed its dating. On the other hand, it is clear, aa I am indebted to Mrs. S. C. Hawkes, M.A., F.S.A., for this information. 130 EXCAVATIONS AT ECCLES, 1973 from the fact that both the arches and the passages had been deliberately filled-in to prevent their total collapse, that at least one later kiln must have been constructed on top of the original oven after the latter was beyond further repair. A narrow, gully-like trench was recorded, out into the subsoil and filled with Romano-British debris, projecting beyond the east corner of the tilery and o. 1 ft. (0·30 m.) wide by 6 ft. 6 in. (l ·90 m.) long; its purpose is not clear, but it may have resulted from industrial excavation rather than served a function in the tile-making process. Two postholes were also recorded in the same area; they were both 1 ft. 4 in. (0·40 m.) in diameter and 6 ft. 9 in. (2·025 m.) apart. Neither the purpose nor the date of these post-holes could be established in the restricted area surviving, though they could belong to a shed or similar temporary structure connected with the industry on the site. As mentioned above, no direct dating evidence survived, and this tile-kiln can only be dated by the medieval tiles used in its construction; these were the typical medieval tiles, with rounded peg-holes, which first appear in use about the middle of the thirteenth oentury.34 This tilery must, therefore, date from about that time, a dating which accords well with much medieval pottery of this period found at the main villa site; on the other hand, it is impossible to estimate how long this industry may have continued though, clearly, it could have been prolonged. l(iln In the courtyard area of the villa, enclosed by its south-west and south-east wings (Fig. 2), was found a circular pit cut into the subsoil and immediately against the inferred line of the Period V bath-house; in fact, part of the baths' south-ea.st wall slightly projected into one of the trenches cut here. This circular pit had a diameter of c. 9 ft. (2 · 80 m.) and a maximum depth of 2 ft. 8 in. (0·70 m.) from the present surface; it had been lined with a 2-in. (0 ·05 m.) thickness of yellow mortar, laid directly on the subsoil, and was found filled-in with lumps of ragstone and burnt tufa. There was also a thick deposit of soot and ash at the north-west side of this pit; the truncated baths' wall was badly burnt at this point, which clearly indicates a stokehole pit beyond the excavated area. No dating evidence was recovered from the back-filling of this pit, but its location and cutting of the bath-house wall obviously demonstrate post-villa use; its purpose is not conclusively established either, but the presence of the burnt material strongly suggests that this pit may have been used as a furnace or kiln. If so, the only date acceptable for its use must be the middle of the thirteenth century when it 34 I owe this information to Mr. P. J. Tester, F.S.A. 131 A. P. DETSICAS has been established beyond doubt muoh of the villa's surviving structures was robbed for building materials. Medieval Site35 Medieval pottery had, in earlier seasons of work, been found mostly in the baok-filling of various robber trenches, and this suggested the period during which took place the wholesale robbing of the foundations of the villa; also, the fact that most such robbing tended to be concentrated at the north-east and south-east corners of the villa indicated the direction from which robbing had begun. However, a medieval site was found this year as a result of mechanical trenches cut to south-east of the villa's perimeter wall (Fig.1).30 Two long trenches were cut in this area by machine, some 20 ft. (6 m.) apart, and, following the finding of archaeological material, the space between these two trenches was excavated. The whole area had been covered with rough cobbling, more like a treading surface, consisting of Romano-British tile debris, lumps of ragstone and cobbles and laid directly on the subsoil. The southern limit of this cobbled area appears to have been a short length (12 ft., 3·60 m.) of insubstantial foundations built of two courses of clay-bonded ragstone to a thickness of c. 1 ft. 3 in. (0·375 m.); work, however, south of these foundations was oarried out muoh below the topsoil level, and it is not impossible that this rough cobbling continued beyond them. To west, the foundations ended within the excavated area; to east, they were cut by a deep ditch, which was 5 ft. (l ·50 m.) wide, U-shaped and had fairly vertical sides. The beginning of this ditch was'not traced as it continued under the rough cobbling, which wa,<;1 not removed, but the ditch could not have extended very much farther to north because of the various pits cut in that area. To east of this wide ditch had been cut another and narrower d itch, which measured 23 ft. 3 in. (6·975 m.) long, 2 ft. 6 in. (0·75 m.) wide and had rounded ends; at its northern end, it partly overlapped a third ditch of similar width but unexcavated length. Immediately to west of these ditohes was cleared a shallow, circular pit, 4 ft. 3 in. (l ·275 m.) in diameter. The rest of the exposed area was occupied by four successive, rectangular cess-pits, cut deeply into the subsoil. Of these cess-pits, nos. 1 and 3 were the earliest, as no. 2 had partly removed the side of no. I and no. 4 had been dug up to the lip of no. 3. No. 4 measured, at the top, 4 ft. 9 in. by a maximum 4 ft. 5 in. (l ·425 by l ·325 m.), no. 3 was rather larger at 6 by 4 ft. (1· 80 by I · 20 m.), and nos. I and 2 have 36 I am grat.eful to Miss R. C. Symes, for the surveying of this site. 31 Arch. Oamt., bocxviii (1973), 76: this pit may also have been a cess-pit belonging to this period of occupation. 132 EXCAVATIONS AT ECCLES, 1073 not yet been fully cleared; in all cases, these cess-pits had sloping sides ,vhich ma.de them rather smaller at the bottom. All the cess-pits had been back-filled with grey-brown soil and building debris, which included residual Romano-British pottery as well as medieval sherds; on the other hand, the layer sealed by this back-filling was uniformly grey and organic and contained mainly medieval pottery. Examination of the medieval pottery found in these cess-pits and on top of the rough cobbling makes it clear that this occupation of the site began about the middle of the thirteenth century37 when it would seem that some sort of impermanent building may have been erected to the south of the length of foundations at the extreme limit of the excavated area, probably connected with the stone-robbing activity on the villa site. The purpose of the ditches is not yet clear, though the two overlapping ones could be interpreted as part of a boundary for the occupied area. SU'l\fl\LIBY A.ND DrsoussroN At the end of twelve seasons' work at this site, some of the questions posed in earlier years have been answered, others still await further excavation. The major result this year has undoubtedly been the excavation of Site D, for two principal reasons: (a) the discovery of potteries on the Medway, so far inland from the pottery sites on its marshes, producing early wares, some of which have hitherto been thought to be mainly Continental imports, in such quantity, as indicated by the enormous amount of wasters, so early in the Romano-British period that it is very unlikely that this massive production wa.a intended to satisfy local demand alone. It is much more likely that this manufacture owes its volume, if not its origin, to the satisfaction of the specific and large demand represented by the invading Roman army; the preponderance ofHofheim-type flagons in the waste deposit is at least a pointer to the possibility of an army contract, which can only be proved by the eventual mineralogical comparison of Eccles flagons and others found in military contexts; (b) if a case can be proved for such an army contract, it will not alter the fact of the great wealth possessed by the original builder of the Eccles villa, in whatever manner this wealth was amassed, but it will help to explain the involvement, however slight and indirect, of the Roman army in the building of the first bath-house which, as suggested long ago,ss appears to have been based on a military model. The excavation of Site D and the evidence it supplied for pottery manufacture can now be considered in conjunction with the road which 37 I owe this information to Mr. A. 0. Harrison, B.A., F.S.A. u Arch. Oant., bcrix (1964), 135. 133 A. P. DETSICAS must have linked the villa with the main road from Rochester to Hastings on Blue Bell Hill. Clearly, this road cannot lie to the south-east of the villa; for, apart from ditches, evidence for Anglo-Saxon and medieval occupation in that area, Romano-British debris, so plentiful and close to the ploughsoil on the slightly rising ground upon which stood the villa, is noticeably absent. On the other hand, there is much Romano-British debris, and another probable building to be examined in 1975, at Site C, close to the main site, where a length of wall, likely to be the west and north-west perimeter walls of the estate, and other evidence have been found and date from as early as the second half of the first century A.n.39 This site is centred, so far as known to date, at a point between the potteries (Site D) and the large Gault Clay pit at the rear of the villa where some evidence has already been found for a road in a section exposed on the face of the pit.4° It is not improbable, therefore, that a roa-d is yet to be found linking Site D with the Blue Bell Hill road and the villa.il Furthermore, the Romano-British cemetery, which is still to be discovered, may be located in the vicinity of this presumed road. The Anglo-Saxon settlement connected with the cemetery also remains to be established though, if the limits of the burial ground have now been reached, it cannot lie too far away to east or south-east of the site. As for the medieval occupation at the south of the site, it is doubtful whether this may prove of great extent or significance, even if it is found that it was not of a temporary nature and connected entirely with the robbing of the villa for the building of the priory at Aylesford; in any case, the area available for further examination is very restricted by a large pond. 3o Ibid., lxxxvii (1972), 108. •0 V.O.H. Kent, iii (1932), 146, 163. u Cf. also, above, p. 100 and fig. 1. 134