Excavations at Eccles 1968

INTRODUCTION EXCAVATIONS AT ECCLES, 19681 SEVENTH INTERIM REPORT By A. P. DETSIOAS, M.A., F.S.A. E:XOAVATIONS were undertaken by the Eccles Excavation Committee for a seventh season, beginning in early April and concluding at the end of October 1968, at the site of the Romano-Bdtish villa at Rowe Place Farm, Eccles, in the parish of Aylesford (N.G.R. TQ 722605; O.S. 6-inch Sheet TQ 76 SW). A continuous fortnight's work was also carried out in August, during which period a training course in archreology was based on the site. Permission to continue with this work wa.s again readily granted by the landowners, Messrs. Associated Portland Cement Manufacturers Limited and the Reed Paper Group Limited. I am greatly in the debt of the tenant farmers, Messrs. A. A. and A. C. Southwell, for welcoming us once more on their land and easing the burdens of our work by their unstinted kindness. Financial assistance for this work was given by an anonymous donor, the Kent Archreologioal Society, the Society of Antiquaries of London, the Haverfield Bequest of the University of Oxford, and other public contributors. The major burden of the actual work fell upon many volunteers, amongst them several members of the Lower Medway Archreological Research Group, without whose help this exca.vation would have been impossible and whom space considerations only preclude from individual mention. I must, however, make an exception in the case of the following to whom I am especially grateful for their sustained support throughout the season's work: Mrs. M. E. Davies, B.A.; Mis& M. B. V. Webster; and Messrs. I. J. Bissett, R. W. Chapman, A. C. Harrison, B.A., F.S.A., T. Hetherington, T. Ithell, B.Eng., W. A. Knowles, R. Lowson, C. E. J. Martin and P. Thornhill, B . .A. I am also particularly indebted to Mr. I. J. Bissett, for continuing to draw the pottery; Miss D. Charlesworth, M.A., F.S.A., for reporting on the glass; Mr. R. G. Foard, for the monochrome site-photography and the prints 1 Arch. Oant., lxxviii (1963), 125--41; lxxix (1964), 121-85; lx:i..-x (1965), 69-91; hc:iod (1966), 44-52; lxxxii (1967), 162-78; lxxxiii (1968), 39-48. Professor S. S. Frere, M.A., F.S.A., has kindly ren.d this report i n draft form and has made several suggestions for its improvement and, in acknowledging my indebtedness, I should !llso like to place on record my personal appreciation for his praotical support and lllterest throughout the several yeats of this excavation. 93 EXCAVATIONS AT ECCLES, 1968 illustrating this report; Mr. A. C. Harrison, B.A., F .S.A., for occasionally acting in my place; Mrs. K. F. Hartley, B.A., for reporting on the mortaria; Dr. J. P. C. Kent, B.A., Ph.D., F.S.A., for identifying the coins; and Mr. R. P. Wright, M.A., F.S.A., for reacting and reporting on the graffiti. Fina.Uy, I must place on record my appreciatfon to my wife for the initial processing of the pottery. THE Exe.av A TION This season's work was planned to continue the examination of the villa's living quarters beyond the points reached in 1967. Period I, ? to A .D. 43: Ditch I A further trench was cut across the line of this ditch north-ea.'!t of the area explored in 19672 but, apart from confirming the alignment of this ditch in this pa.rt of the site, no fresh evidence was forthcoming nor did it prove possible to carry out more extensive work in this area. Periocl-s II-III, to c . .A.D. 65: Ditches IV-VI To these periods are tentatively assigned for the time being three rectilinear ditches found at the very end of the season's work. As these ditches are built over by the fir&t house, they are clearly earlier than Period IV; this is independently supported by the few sherds found in Ditch VI. On the other hand, it must be noted that no pottery was recovered in either of the other two ditches. However, as Ditch IV runs approximately parallel to Ditch VI, into which runs Ditch V, it is clear that Ditches IV and V cannot have been cut much later than Ditch VI, even if it can eventually be shown. that all three are not strictly contemporary. Ditch IV was a.ppro:rimately 2 ft. 6 in. deep and about 3 ft. wide, if its sides are projected to th􀃷 top of the Romano-British ploughsoil; it has been traced so far for some 74 ft. and its profile is roughly V-shaped. Ditch Vis rather wider (3 ft.), but of the same depth and outline; so far only some 19 ft. of its course have been examined, from its junction with Ditch VI towards the eastern part of the site. Ditch VI is by far the largest of the three, with a width and depth of 8 ft. and 3 ft. 9 in. respectively; it is U-shaped in outline a.nd its course has been traced for about 70 ft. What distinguishes these ditches is that Ditches IV and V are cut mainly through the Romano-British ploughsoil and barely penetrate the subsoil, hence their filling contained a very smaJl proportion of Gault Clay, whereas Ditch VI is cut well into the subsoil and its filling consisted mainly of re-deposited clay subsoil over some silt accumulation. Clearly the subsoil upcast from this ditch must have been piled up on its western (1 outer) lip and later a Arch. Ocmt., lXJCXii (1967), fig. l; lxxxiii (1968), 40. 94 EXCAVATIONS AT ECOLE$, 1068 used to backfill the ditch when the site was levelled for the purpose of building the earliest villa. The function of these ditches is at present uncertain though, as Ditch V runs into Ditch VI, it is safest to assume that they were intended for drainage.8 Periods III-IV, c. A.D. 55-120 Examination to the north-east of the villa's ma.in range of rooms has shown that some sort of a building, perhaps a small outhouse, was located in this area before the building of the rear corridor; all that survived of this presumed structure is a short length of wall, built of ragstone and yellow mortar and about l ft. 6 in. wide, running approximately parallel with the villa and turning to south-west. The whole area to north-east of the outer wall of the rear corridor showed evidence for a compacted layer of yellow mortar, cut immediately outside the wall by the trench of the water-pipeline passing under Rooms 86 and 88, which is more likely to be the result of mixing mortar at the time of the building of the corridor rather than a floor, as it rests directly upon the Romano-British ploughsoil without any make-up layers beneath it; this mortar layer was recorded as far as the southeast limit of the length of wall mentioned above and covered some loose ragstone laid in a shallow trench which may belong to foundations extending that wall towards the south-east. Two fresh trenches were out in the area of Room 94 which was partly explored in 1967.4 One of these established the line of the south-west wall of this feature, and the other was out astride both walls so as to give a continuous section. AB a result, it is now known that this feature was 12 ft. 9 in. wide; its full length is not yet traced, 34 ft. 3 in. only have so far been examined. As previously reoorded,4 the whole feature had been methodically demolished: its north-east wall had been completely removed, and only the cutting of its construction trench into the sub-soil indicated where it had been built; the south-west wall had fared rather better as some loose foundation material survived in situ and suggested that the demolition had been carried out from a northeasterly direction. No traces whatever of the floor of this feature were preserved in these new trenches; instead demolition had penetrated 1 ft. 9 in. deeper than the level thought last year6 to be that of the floor, and the back-filling consisted of alternating layers of building debris and subsoil, the latter clearly deriving from construction trenches dug elsewhere on the site at the time of the demolition. The depth of this back-filling suggests that the floor of Room 94 must · 3 This ditch seems to run practically parallel with a similar reotilinear ditch explored in previous years at the western limit of the site; cf. Arch. Oant., lxxxii (1967}, fig. 11 and pp. 163-4. 4 Arch. Oant., lxxxiii (1968), 40-1. 6 Ibid., 40. 95 EXCAVATIONS AT ECCLES, 1968 have been lower than 4 ft. 6 in. from the present surface; it is scarcely deep enough to indicate the presence of a cellar. On this basis, what was thought in 1967 to be the floor-level could be tentatively re-interpreted as the top of the uppermost of a number of steps such as would be found inside a bath. Support for this suggestion can be found not only in the fact that the debris filling in the 1967 trench dropped below the opus signinum then considered as the bathfloor but also, and much more to the point, in the opus signinum rendering on the inner surface of the north-west wall, which had survived, and the painted wall-plaster debris present in the backfilling, which would have rendered the wall above the opus 8igninum, i.e. above the water level. The alternative possibility that the room was first robbed of its materials and then excavation continued at depth for the purpose of obtaining clay can be discounted on the grounds that clay could be had more easily elsewhere and that it formed much of the back-filling. Furthermore, the subsoil material used to back-fill this feature argues that the area occupied by it was being levelled. Immediately beyond the line of the north-east wall in the larger of the two new trenches was found a layer of whitish mortar deposited directly upon the subsoil; its full extent is not known but, as it did not appear in the 1967 trench further to north-west, it is unliltely that it represents anything more than mortar being mixed at the time of the construction of Room 94 . .A large part of a small, carinated reeded bowl, normally datable from Flavian times to the early second century A.D., was found embedded in this mortar deposit. It seems now clear that this feature was not built as early as provisionally suggested on the previous report and that the early-Flavian samian found below the level of the opus signinum step must be considered as rubbishsurvival. The function of this long feature is still fa.r from certain, though the suggestion put forward above that it may have been an open-air bat,h does fit all the known facts of the excavation. Its pro;ximity (about 28 ft.) to Room 93, the fronting corridor of the villa, makes it unlikely that it would have been retained beyond Period IV; it certainly had been demolished before Period VI a,s the cobbling, extending to south-east from the outer wall of Room 15, the corridor leading to the baths of that period, had been carried over the line of the feature's south-west wall. Period IV, c. A.D. 65-120 Examination of the central range of rooms in this period of the villa's occupation continued and two further large rooms were added to its plan. 96 i---i I u-7 j I i' i i !_ _ __ J JS JJ 􀂻 Periods HNV 􀂼!Periods //MV 1-·-·-·-·-·--·- 0 􀀂] 􀃀PerlodV.Robbed/lnltrred WJl Period VI 􀂽IPerlods H//V,R oilburing 11 ft. by 11 ft. 6 in., provided:the heat for the channelled hypocaust under Room 95 through a flue, 1 ft. 6 in. in width, in its north-west wall; both partition walls of the furnace-room were 2 ft. wide and built of ragstone set in yellow mortar. Examination of their construction trenches, however, confirmed the observed fact that these two walls were not of contemporary construction; for not only did the north-west wall of the room look far more solidly built than the south-east one but also its ragstone and mortar construction had continued to the very bottom of its bedding trench, whereas the soμ.th-east wall had been founded on a deposit of loose ragstone and debris containing painted wall-plaster and mosaic fragments, as already noted for the partition walls of Rooms 95, 92, and 103. Originally, the furnace-room was not floored but the fire had been laid directly on the subsoil, except for the area directly in front of the mouth of the flue where amongst the debris filling were found three fragments of sandstone which have not originated in the Wealden area; they could have come from the Millstone Grit of South Wales or the Midlands or, perhaps, even from abroad.lo All these fragments are slightly altered by heat, and it would thus seem that steps had been taken to ensure that at least the mouth of the :flue was floored with durable materials: Kentish ragstone has a calcareous cement which, turning to calcium oxide on heating, would cause the stone to crumble, whereas these pieces of sandstone have a siliceous matrix which would not decay in the heat, of the furnace. Two very worn coins, both of Tetricus I and Tetrious II, were found embedded in the subsoil within this furnace. At a later stage during the life of Room 96 a reconstruction apparently became imperative because excessive heating of the northwest wall was beginning to affect it, and a pair of cheeks was constructed in order to reduce the area of direct contact; that these cheeks were not part of the original plan is clearly shown by the thin deposit of soot and ash found below them on top of the subsoil and extending to the face of the north-west wall. Both cheeks measured 4 ft. by 3 ft. 6 in. and were built of ragstone bonded with clay which had hardened to a brick consistency as a result of firing in the furnace; at the same time, tiles were laid o n the subsoil, again sealing some soot and ash below them, to serve as a floor, though they were very badly cracked and disturbed in the course of later robbing. The entire area of the furnaceroom was filled with the usual demolition debris and, above the remains of the tiled floor, a fairly thick la.yer of ashes and charcoal; this deposit u I a.m indebted to Mr. I. F. Mercer, of the Institute of Geological Sciences, and to Dr. R. P. S. Jefferies, B.A., Ph.D., F.G.S., for the identification of these fragments and useful comments. 102 EXCAVATIONS AT ECCLES, 1968 yielded a scatter of 65 coins, mainly of fourth-century A.D. date,20 clearly dropped in the course of tending the fire. No direct evidence survived for an entrance into this room, but it is virtually certain that this would have been sited through the north-west wall and at about the east corner of the furnace. Rooms 96, 97 and 101 occupy the area, as far as examined, of the earlier Room 104, i.e. the north-eastern section of the villa's corridor beyond the south-east wall of Room 95 where the tessellated pavements stopped and yellow mortar was used as flooring material. Though once again no direct evidence could be secured, owing to the complete demolition in this area and the robbing of the walls, it is clear that the partitioning of this corridor here into smaller rooms is not contemporary with the partitioning of the north-western area of the corridor as the construction trenches cut for both partition walls of Room 101 are much shallower than those of the pai-tition walls beyond the furnace. room. Room 97 measures 11 ft. 6 in. by 17 ft. 3 in. and Room 10111 ft. 6 in. by 20 ft. 6 in.; both these rooms were floored with yellow mortar, about 2-4 in thick, laid on a make-up layer of re-deposited subsoil.21 Room 100, a small apsidal recess opening off the north corner of Room 101 (Plate III, B), is shown on the plan (Fig. 1) as belonging to an earlier phase than Rooms 97 and 101; in fact, because robbing has again removed the evidence, this distinction may not be valid, and it is more likely that all these three rooms belong to the same reconstruction of this part of the rear corridor. The apse measured 7 ft. 6 in. in length and had a maximum width of 3 ft. 9 in.; its floor com,isted· of opus signinum, about 4 in. thick and la.id on successive make-up layers of loose mortar rubble and building debris over a foundation course of unmortared ragstone resting on the subsoil. The junction of the floor and the wall of the apse had been sealed with a strip of opus signinum quarter-round moulding, which indicates that the apse probably contained water. The wall of this room had been built on a foundation of loose ragstone laid on the subsoil below a layer of yellow mortar rubble and tile debris; immediately above this deposit were two courses of mortared ragstone and then the wall of the apse which consisted of coursed bonding-tiles set in yellow mortarfive of these courses of bonding-tiles survived below the ploughsoil. The apse had clearly been abutted on the north-east wa.11 of the corridor and, though the latter had been completely removed, the apse's wall ao Included in this scatter were coins of Gratian (1), Constantius II (3), Urbs &ma, (1), Theodora (1), Magnentius (3), Constans (4), an a,ntoninianue of Ota.cilia Severa., House of Constantine (1), Decentius (1), 'Constantinopolis' {l), Va.lens (1), nine illegible coins, probably of late-Roman date, and 37 barbarous FEL TEMP REP .A.RATIO types. 11 See note 7, above. 103 EXCAVATIONS AT ECOLES,.1968 and opus Bigninum quarter'0round moulding were preserved · hi.tact along the whole course of their abutment on the north-east wall. Obv:iously, a doorway and, perhaps, a step into the apse must have existed through the wall at the north corner of Room 101, but robbing has remo-ved all direct ev:idence. DATING Once more all the ev:idenoe recovered .iii. this season of work has confirmed t,he prov:isional dating sequence proposed .iii. earlier reports. On the other hand, the recovery .iii. sealed deposits in Rooms 95 and 96 of coins and pottery enables the closing date of Period VII to be established at about the beg.irining of the fifth century A.D., rather than the prev:ious suggestion of merely post A.D. 290. SUMMARY AND Drsouss10N After seven years' work at this large site, it is now possible to attempt a tentative summary and evaluation of the history of the villa (Fig. 2).22 The ol'iginal house (Villa I, Period IV, c. A.D. 65-120) consists, at present, of at least six rooms, three of which were certainly laid with tessellated floors, ranged behind a fronting corridor; clearly, this is still only part of the villa's earliest plan, but it now looks probable that it did not possess any wings and recalls in its essentials town striphouses and other villas,23 though it is likely to have been larger and was certainly built earlier t,han most other villas. The bath-house belonging to this first dwelling-house has already been fully examined and described;24 both its size and furnishings require a comparable degree of Romanization in the house itself which should be established when the whole villa is fully excavated. Villa I and its baths form a substantial establishment, showing that 'exceptional Romanization was actiye in this area',25 within a generation of the Roman conquest; the influence of military baths on the architecture of the Eccles first baths has already been referred to.20 Apart from the implications that this raises, the size of the baths is inordinately large for the occupants of the -villa and their attendants alone; certainly these ba,ths could act as 'some sort of social focus'27 for the estate workers and, probably, othPrs in the neighbourhood. Thls, however, implies organized large-scale farming, i.e. a large farming community, bailiffs, slaves (a, length of 22 The pre-Roman occupation of the site is still incompletely explored. is E.g. Lockleys, Park Street, Frocester Court, etc. " Arch. Oant., lxxx (1965), 71-85; for its re1ationship to the villa, ibid,, lxxxiii (1968), fig. I. " S. S. Frere, Britannia, 1967, 269-70. " Arch. Oant., l:xxix (1964), lxxx (1965), 89. 11 S. S. Frere, op. cit., 270. 104 ECCLES ROMAN VILLA JS J6 J4 JJ J4 JJ 79 J6 JJ I /OZ III IS llll r·· ,, I. "Ii ' ' ' :::::::::: 1· -􀀉 J , ■■ .. I ; . . . • l ---------- IV IS FIG. 2, 98 99 98 99 /04 98 99 9J 96 97 IOI 98 99 9J [facep. 104 EXCAVATIONS AT ECCLES, 1968 chain found in 1965 may be a pointer here), craftsmen, etc., very early in the life of the province; if so, this should mean that the owner of this estate was of considerable social importance and wealth. Only a person of this description, with obvious philo-Roman. leanings, could wield sufficient influence to be cultivated by the Romans for their own purposes,28 and possess the wealth needed for the construction and appointments of the earliest house and baths-there can have been very few persons fitting this description in late Iron Age Britain, but it would be idle to speculate further on this point. The destruction by fire of the first baths29 seems to have brought about not only the building of a new ba-th-houses0 but also some additions to the dwelling-house itself (Villa II, Period V, c. A.D. 120-180); these consisted of the construction of the north-west wing of the villa, a small complex: of servants' quarters, and the lengthening and reconstruction of the fronting corridor of the villa. The implication of a thriving establishment is further underlined in the next stage of reconstruction (Villa III, Period VI, c. A.D. 180-290); a rear corridor was now added to the house, new servants' rooms and kitchens were built off the north corner of the building, a new and substantially larger bath-wing was erected.Sl Certainly, this is not 'a period of decay'S2 at Eccles as might be argued for other villas;ss and, if it can be accepted that the Lullingstone villa may have been confiscated as a result of its owner's involvement in the political upheavals of the period,84 one wonders why the Eccles villa seems to have escaped unscathed. Moreover, this cpntinued prosperity of the estate was not affected by the.troubled times of the fourth century A.D. (Villa IV, Period VII, c. A.D. 29()-400). The rear corridor was now partitioned into small rooms, a channelled hypocaust was inserted, other rooms were added;35 nor was the life of the villa apparently disturbed, let alone interrupted, by the upheavals of A.D. 367, if the coins in the hypoca.ust's furnace are taken into account. On the other hand, evidence is gradually accumulat-ing that the centre of gravity of the villa's life may have shifted from the central range to the unexcavated south-east wing. The size of the villa., its early foundation and continued exparu.ion 18 It is not impossible tha.t a military engineer was responsible for the building of this bath-house; Arclh Oant., lxxix (1964), 135. 20 Arch. Oant., lxxx (1966), 86. so Il>ii