The Lullingstone Roman Villa

THE LULLINGSTONE ROMAN VILLA SECOND INTERIM REPORT By LIEUT.-COLONEL G. W. MEATES, F.S.A., E. GREENFIELD and EDWYN BIRCHENOTTGH, M.A. INTRODUCTION THIS Second Interim Report deals with the results obtained during the 1950 and 1951 seasons of excavation and is concerned in particular with the northern part of the building where an important complex of ramps and stairways has been revealed. The Basement Room has also been excavated a further 5 ft. towards the east in the hope of reaching its limit in this direction. The limit has not been reached, however, and the presence of the road and underlying sewer defers further excavation. It is essential to obtain complete access to ah the evidence contained in the filling of the Basement Room, as without it not only are we prevented from working out its successive purposes, but, even more important, we cannot complete the unique Early Christian paintings that adorned the upper room and which are contained in the filling below. Both northern and southern walls of the fourth century house have now been obtained, and most of the western edge has been defined. As the work on the south and south-west has not yet been completed, the results of this have not been incorporated in this Report. Work has also been carried out on the short corridors north and south of the Mosaic Rooms to obtain more precise dating, and this work is being carried forward into the 1952 season. The evidence here is sparse, but what has come to light suggests an early to mid-fourth-century date for the mosaics themselves. Room 15 is dealt with in detail, and a transverse section is recorded, together with an architectural section west to east along the hne of the main flight of steps and extending through rooms 14, 15, and 16. The stratification in Room 15 provides a large quantity of evidence in relation to the neighbouring rooms and especially in relation to the Basement Room, whose chronology is rendered clearer, although its successive purposes, while suggestive, yet remain obscure ; its conceivable parallels may perhaps be found in the neighbourhood of the Rhineland. A second architectural section has been recorded from south to north throughout the house and passes through Rooms 11 and 15. This section, when compared to the west to east section, the plan, 26 THE LULLINGSTONE ROMAN VILLA and the diagrammatic sketch, will render the stairway complex understandable. In ah these matters our thanks are due to Mr. D. A. Broodbank, A.R.I.B.A., Mr. C. B. Mears, A.R.I.B.A., and Mr. Downs. As regards dating, no attempt has been made to obtain precision dating of the north stairway complex by interference with its construction. The evidence provided by the fillings that sealed it in Rooms 15 and 16 and supported by the state of wear of its tile facings place it probably in the second half of the second century, and perhaps later in that period. A suitable examination technique for these stairways is a complicated matter and their preservation undisturbed must first receive weighty consideration. Generally, it may be said that as a result of these seasons' digging, the dating and periods of construction of the house confirm those advanced in the First Interim Report (Arch. Cant, LXIII). It is sthl uncertain if the plan of the house before the period of abandonment at the end of the second century represents that of the main dwelling, or whether at that time this part of the building was a wing of a more extensive house whose plan we do not yet know. It is, however, sufficiently clear that towards the turn of the third-fourth centuries, the foundations of which we have knowledge were used for a smaller house, compact and symmetrical, the Mosaic Rooms later forming its central point. We have clear evidence that the Antonine house extended north, and this extension will be found entered on the plan. It is also continued towards the north-east, where a further building is known to exist; and this will be investigated during the 1952 season. Some very interesting work on the superimposed floors of Room 5 has been carried out; but beyond mentioning "that coins of Claudius II, Allectus and Constantine I I as Augustus, were found stratified in this room, report on this apartment and its relation to the remainder of the Bath Block must await complete excavation. It remains to thank all our workers who once again by their keenness and efficiency have made the'seasons under review so successful, and to acknowledge the continued interest and support both of the owners of the site and of archaeologists generally. We wish particularly to thank Miss M. V. Taylor, CB.E.,' M.A., F.S.A., and Professor J. M. C. Toynbee, M.A., D.Phil., F.S.A., for their helpful suggestions and provision of parallel data; to Dr. F. Oswald, D.Sc, F.S.A., for his detailed report on the Samian ; to Mr. B. W. Pearce, M.A., F.S.A., who again reports on the coins ; to Mr. Bernard Middleton, F.R.S.A., who again reports on the glass; to Professor A. J . E . Cave, M.D., D.Sc, of the Department of Anatomy, St. Bartholomew's Hospital Medical College, for his report on the infant burial in Room 8; to Dr. H. J . Plenderleith, M.C, B.Sc, Ph.D., F.S.A., Keeper of the Department of Research Laboratory, 27 THE LULLINGSTONE ROMAN VILLA The British Museum, for his report on the contents of the mortarium found in Room 16; to Mrs. B. L. Macrory for the pottery drawings; to Mr. A. G. Bell, I.S.O., F.G.S., who once again organized the site lectures, and to Mr. C D. P. Nicholson, F.S.A., F.R.Hist.S., F.S.A.(Scot.), F.S.G., for the great pains he has taken in assembling the painted wall plaster from the filling of the Basement Room, a project which by his patience and skill is taking shape with dramatic results, showing as it does that the plaster decorated a room where Romano-British Christians worshipped in the late fourth or early fifth centuries, perhaps 200 years before the arrival of St. Augustine. The two great chi rho monograms and the row of Orante figures facing the east sufficiently testify to the existence in the villa of the earliest place of certain Christian worship in Britain. A report on these wall paintings will form a separate publication in the future. The imperative protection of the site during the winters of 1950 and 1951 was supphed by the Kent County Council who generously defrayed the cost of necessary materials ; and a further grant of £50 by the Administrators of the Haverfield Bequest, is gratefully acknowledged. Finally, our thanks are due to Mr. John Matthew, the Secretary of our Organization, for his efficient work in the field of organization and administration. THE NORTHERN COMPLEX Room 11 (The Basement Room) As has been indicated, the presence of the road and sewer prevented any further large scale exploration of this room. It was possible, however, to excavate five feet onwards towards the road and to obtain an important section (PI. l a and Fig. 1). Two additional phnths were discovered, forming a rectangle with the two discovered last season (see plan, Fig. 2). AU four bore on their surfaces the remains of completely carbonized wood, probably horizontal plates to support uprights of some kind of framework fitting closely inside the walls of the room and supporting the room above. The discovery was also made, half-way between, and in line with, these two plinths, of a shaUow sump or water channel lined with shaped chalk blocks. This is a significant construction, as it seems to indicate that water was at one time present in the Basement Room. The bottom of this sump or channel, whose limit eastwards has not been reached, lies three feet below the present bed of the river, and the whole had become silted to within a foot of its top when the villa was finally destroyed. Reference to the section at Fig. 1 will show.the method of construction of this channel, and it is possible to give some chronplogy and to equate the construction with the deposit of the marble busts in the room. The sequence is as follows : 28 THE LULLINGSTONE ROMAN VILLA 1. The second-century concrete floor was roughly cut into and the sump or channel was dug to a depth of three feet below its level, being lined with shaped chalk blocks. 2. The clay floor was laid over the entire concrete floor and packed up against the box of chalk blocks. 3. During the second half of the third century, the bottom became filled to a depth of some twelve inches with clay and fragments of brick, and this was finaUy sealed by a thin tip of carbonized material. 4. Immediately above this ash tip was found a potsherd of dating importance. It is described at Fig. 8, No. 33. SimUar potsherds have been found at Colchester and SUchester, at the latter place being said to appear during the first half of the third century. This is the only datable object from the fining and from this single potsherd the construction of the sump or channel is placed tentatively about the middle of the thhd century. It was accompanied by a complete box-flue tile similar to those in situ in Room 15, which is a third-century construction, and a large portion of a patterned box-flue tile (Group 5 of " A study of the patterns on Roman flue tiles and their distribution " by A. W. G. Lowther, F.S.A.) which is, however, a relic of the secondcentury house. Bearing in mind the SVAVIS pot of mid-third-century date which appears to have been a votive deposit when the busts were placed in the room, it is reasonable to suppose that the deposition of the busts and the digging of the sump or channel were contemporary events. 5. Subsequently, the chalk lining was roughly repaired by the insertion of square bricks near the top, but by the time of the final destruction of the vflla this lining had become greatly dilapidated, much of the chalk having disintegrated. A group of five mortaria (see plan at Fig. 2) was found upon the surface of the clay floor around the sump or channel and sealed down by the carbonized destruction layers. These mortaria, unlike the mortarium in Room 16, showed no sign of contents, and theh disposition around the chalk lined sump or channel suggests that they were in use at the time of destruction, though there is nothing to show whether it was a domestic or religious use. AU five were standing complete at the time of the destruction, and all are fourth-century types, Nos. 39 and 40 being small colour-coated vessels, Nos. 41, 42 and 43 large heavy vessels. No. 41 rested in No. 42 and bore the brunt of the destruction ; it was a badly burned specimen and is too friable for reconstruction though its type is clear. These mortaria are illustrated at Fig. 9. 6. The destruction by fire was very evident from the quantity of completely carbonized planks and lengths of wood that were found lying upon the latest occupation surface. This fallen debris of the floor and room above averaged twelve inches in thickness and contained •chalk rubble, opus signinum, tiles and wall plaster. This plaster was 29 THE LULLINGSTONE ROMAN VILLA somewhat less in quantity than that recovered in 1949 ; the painted dado shows a turn into a doorway or window, while some new and richly coloured designs of great complexity came to hght. It is possible that we have here the first fragments to be recovered from the east wall of the Christian room above. 7. The first waU faU from the room above occurred soon after the fire, and a study of the section gives the impression of a north waU of flint and mortar up to the roof and a south wall of possibly wattle and daub to the roof, but resting on the south wah of the Basement Room, which is of flint and mortar. A cursory study of the waU plaster backings from the upper room shows that many fragments were fixed securely to daub which in turn carries the impressions of wooden laths or perhaps planks. An exhaustive examination of the plaster backings is a future task. Some roof tUes seem to have shd down at the time of this first waU fall, but they are largely concentrated in the southern part of the filling. An important circumstance was the recovery from the first waU fall of a large number of chalk blocks shaped as voussoirs of an arch. A section of this arch was found intact, though ah the blocks had shpped down from the upper level where they probably formed part of a doorway or window. An hon anvil and hammer were found together next the south wall in this layer. 8. Parts of both north and south walls remained standing for sufficient time to aUow a black earthy deposit to silt over the first waU fall. This formed a layer showing very few fragments of fallen wall debris. It averaged a foot in thickness and contained a scatter of fourth-century pottery, of which the most important sherds come from two bowls, colour-coated, rouletted and decorated with painted scrolls (Fig. 8, Nos. 34 and 35). Nothing of positive date came from this layer, but its deposit can clearly be placed very late in the century. 9. The remnants of the walls either fell or were pushed in, and the remainder of the roof tUes shd to rest on the debris. There is no evidence of post-Roman occupation in the black layer above, and the presence of large flints in it gives the impression that this part of the site was subsequently leveUed, but at what period there is no evidence to show. The only comment that can be made upon this further excavation of the Basement Room is that while more evidence has come to light of construction and active occupation, this is helpful in a very limited sense to the understanding of the functions of the room. These functions, on present evidence, are only suggestions, but the modifications, in construction are clear. Both may be summarized as follows : Period I. Probably late First Century. Excavation for the room having been made down to flood-plain level, the natural clay and flint was cut to form three ramps by which it might 30 THE LULLINGSTONE ROMAN VILLA be approached from the. higher level. The pah leading into it from the south were steeply cut and faced with rammed chaUi, and both carried wooden arrangements, the westerly being possibly a staircase. The ramp to the north was not faced with chant, and its gentle gradient suggests a wanting way. The room does not seem to have had any sort of floor at this period, and its depth suggests that water from the river may possibly then have been admitted. ROOM I I SECTION TTnT M O D E R N ROAD iTiriTiiTniTiTriTi'i'iTirriTi'iTrriTiTiTriTiTri'iTi'iTrrrrri'iTi'i'i'iTi'i © iiiiit o CHALK BLOCKS CHA BLOCKS NRON HAMMER AND ANVIL © BLACK WITH LARGE FLINTS. © FLOOD PLAIN GRAVEL. POTSHERD (3rd CENTURY) N9 33 ElG. 1 GINGERY MORTAR. (DECOMPOSED WALL FACING) WALL FALL. (FLINTS, MORTAR,TILE.ETC.) BLACK WITH FLINTS. (REPRESENTS PERIOD OF REST) DESTRUCTION LAYER. (CHARCOAL.PLASTER, ETC.) CARBONIZED WOOD. CLAY PACKING. CLAY FILLING WITH FLINT AND TILE. Period la. Probably late First/Second Century. The steep ramp at the south-east was blocked and a masonry staircase took its place. This staircase, like its wooden neighbour, does not descend to the bottom of the room, but stops short about two feet above it. Period II. Late Second Century. Both approaches on the south were blocked. The room was provided with a concrete floor, and the walking ramp on the north was 31 THE LULLINGSTONE ROMAN VILLA overlaid by a fourfold tiled stahcase. The walls of both room and stairways were given a painted decoration of panels and vines in hght colours, and the room probably functioned as a loggia or garden entrance to the villa. It is possible that the river, by changing its course, had provided by its silt the basis for such a garden, though this remains to be proved. ROOM II P L A N . POT I © No S4 WEDGE No 38. SVAVIS ^ No 23 N» 2»»27- BUST n PEDESTAL BSHR2 ^ POT ni © No 25. o X CAT SKELETON © ([No 41. <-No 42 CHAMMEL OR SUMP PLINTH No 39 PL NTH M O D E R N ROAD ^ Era. 2 Period III. Mid to Late Third Century. After a period of abandonment lasting from c. 200 to c. 250 the tiled stairs were rendered obsolete by the construction of the high level hypocaust on the north, involving the blocking of the stairs. The room then became a lodging for the marble busts. The sump or channel was dug, a clay floor was laid over the concrete floor, whhe two votive pots were sealed down beneath the clay floor (vide First Interim Report, Arch. Cant., Vol. LXIII). Period IV. Fourth Century. Two more pots were placed in the clay floor, which at this period received three coins temp. Constantine I, into its surface. The apartment 32 PLATE I a, Hiiwiwnt Koom, looking Kimt. "=• h. Basement Room. Busts in rilu, showing potter}' wedge (Xo. 38). LULLINGSTONE ROHAN VILLA. fare p. 32 PLATE II a. South-west Ramp, looking West. 6. South-east Ramp, looking North. LULLINGSTONE ROMAN VILLA. PLATE III iir—jTurwr" • - ^ ^ ft **~ W^^jjjt 1 * f* ""-« > > v ' ' . >^ 4 Y • C j - « S ^ M ^ * ^ _ J £ L ^ ^ M • * ft •r: - * •- "sSSl* m / -f. * i^gfr '/#.. -.2 Range of deep compartments, looking North (S.W. third-century hypocaust). Ramp, Basement Room, LULLINGSTONE ROMAN VILLA. PLATE IV * "' "*. *':"•'."y- x * " ^ ^ ^ \ . • v # * i M. Room 15, looking N.E. LULLINGSTONE ROMAN VILLA. THE LULLINGSTONE ROMAN VILLA above was reconstructed to provide a place of Christian worship, and the four plinths in the Basement Room were sited to carry vertical uprights to form a support for the wooden floor above. The terminal date of occupation is uncertain, but it must be at least towards the close of the fourth century. It is indeed possible that the final phase may have been enthely devoted to Christian worship, secular habitation of the house having ceased. And in that event, the terminal date may well be found in the fifth century. Room 15 A glance at the site plan (Fig. 18) will make it clear how the stratification of this room affects the dating of the steps in the Basement Room (Room 11). Before the insertion of the wah block between these two rooms and the wall block between Rooms 15 and 16, the four flights of steps formed a single unit, and in dating the steps in Room 15 we, therefore, date those in the Basement Room, which in turn whl aid us towards a time-scale for this latter room. It is convenient to treat this important Room 15 in phases (section and plan at Figs. 3 and 4). Phase I. The Antonine house became derelict c. A.D. 200 and remained so until the second half of the thhd century, during which a silting was deposited over the steps, tailing off in thickness towards the north and east. This silting was largely composed of fallen plaster from the stahcase walls and contained late Antonine pottery and Samian dated from c. 185 to c. 200, together with a coin of Antoninus Pius sealed weU down in it. This coin is an as and is dated to A.D. 144-45, its worn condition suggesting that it may still have been circulating at the end of the' century. The amphora stamp ACIRCI (Fig. 10, No. 44) included among the coarse pottery in this layer also confirms the Antonine period for the deposit {Richborough, IV, p. 242, where it is noted as of second century date). The top level of the layer coincides with a destruction level in the south wall, and confirms the proposition of a partial dilapidation of the house as a result of abandonment c. A.D. 200, of which contemporary political troubles may have been the cause. The tiles on the steps, which are worn by considerable use, were sealed beneath this layer and are, therefore, largely in situ, while those on the three treads leading on into the Basement Room were found to have been cut away, the inference being that the Basement Room remained open during the period of abandonment. Phase II. In the second hah of the thhd century, the site was again occupied and some rebuilding was undertaken. Room 15 and its pendant Room 16 were designed, the former to be heated by the hypocaust system and connected with a large brick arch with the latter, in which direction the furnace was intended to be sited. A thin layer 33 THE LULLINGSTONE ROMAN VILLA of clay was laid over the Antonine silt, being steeply channelled opposite the arch and round by the footing trenches of the north and west walls, which were the only ones to contain box-flues. These new apartments cut off the Period I ramp and the Antonine rooms and corridors at the north, leaving them robbed to theh foundations and exterior to the remodelled house. It is uncertain whether this hypocaust was ever put to use; carbonized material is absent and there is no sign of calcination of the flints next the arch. The clay layer or baffle showed no signs of heat, and supports for the floor, or any impressions of them, were lacking. A wooden floor supported on beams is possible, but there is no evidence of it. It would indeed seem that the room began to be filled with a layer of rubbish almost at once. This layer was greyish in colour and sharply contrasted with the yellowish Antonine layer sealed beneath it. It contained third-century pottery (Fig. 11), a little Samian dated c. 200 and later, which also occurred in the footing trenches, and a denarius of Alexander Severus, which is dated to A.D. 222-8. This coin is in fah condition, but it may well have been circulating for twenty or thhty years before it was dropped into this grey filling. A pah of ox skulls were deposited at this time, one in the channel immediately inside the arch, the other in the channel running along the inside of the north wall (PI. VIb). It is unlikely that any significance can be placed upon this deposit, though the ritual deposit of an ox's skull in the triangular temple at Verulamium may have some relevance. At Verulamium the re-edification of the temple, to which period the skuh there belongs, is dated on shght evidence to the end of the third century (Verulamium, p. 117), and the identity of period agrees both there and at Lulhngstone. At Verulamium the skull was deposited in a cavity roughly lined with roof tiles, whereas the skulls at Lullingstone were unprotected. It is to this phase also that the deposition of the damaged marble busts in the neighbouring Basement Room is ascribed (Arch. Cant., LXIII, p. 13) ; and the insertion hi the floor in front of them of the SVAVIS pot conforms in date with the same phase in this room (Room 15). Phase III. There seems to have been a pause in the occupation of this room until towards the middle of the fourth century, when intensive use of it again becomes apparent. Once more, no use seems to have been made of the hypocaust arrangement as such, but the surface of the grey third-century filling was adopted as a level for operations that requhed heat. A new west wall containing a doorway was set back from the earlier wall containing the box flues, and a small ramp of heavy gravel was packed against it on the inside to lead down to this level. The southern portion of the room was the site of burning or incandescence, and in this part the lamination of burnt levels of charcoal with layers of burnt clay and reddened sandy material indicate a 34 THE LULLINGSTONE ROMAN VILLA lengthy period of use. The charcoal was intermittently raked back into the northern part of the room where it remained in layers, a shallow gully running east and west being maintained between the burnt area and the raked back material. In the north-east corner was an area of rough concrete which perhaps formed a small standing, sealing a badly burnt coin, probably Constantinian, c. A.D. 324 plus. Thirtyone coins were recovered from the burnt and raked back zones. They date from A.D. 330-54, the terminal date behig fixed by a coin each of Magnentius and Conatantius Gallus lying with a minim upon the surface of the burnt layer. It is clear, therefore, that the room began to be used for some type of combustion a few years after A.D. 330, ceasing to be so used probably in the sixth decade of the century. A word may be said as to the possible activity that took place in the room during this period. A thin hard level was encountered over the major portion of the burnt area at an average depth of six inches below its surface, and this hard level contained sixty-seven post-holes (plan at'Fig. 4). They formed an irregular arc around an empty space next the southern wall, the centre point of this semi-circle being marked by a single post-hole. They were of two sizes, fourteen having an average diameter at the top of three inches, the remainder of one and a half inches. At first sight the pattern appears haphazard, but if a movable framework is postulated, this framework having been from time to time lifted and placed in different positions, but with its centre point unchanged, the legs of such a structure might well produce such an apparently chaotic pattern of post-holes. Signs of such a framework are suggested at Fig. 4 where PHs 1, 2, 3 and 4 seem to be centred on PH 5 and may show the main supports of a semi-" bee-hive " type of framework with its diameter lying along the wall. Had some such contrivance been in use, it may have been placed over incandescent material, for drying, airing and possibly bleaching cloth, the cloths being spread over the frame, confining the heat and excluding the ah. While it is not suggested that a fuller's establishment was then in existence at the vhla, it is a possibility that some such use was made of the room. It was not used for smelting or any sort of foundry work. The use of the post-hole complex having ceased, it became concealed by a further deposit of burnt matter, representative of a short period only. Phase IV. This use of the room having ended c. 350-60, the walls were repaired and a floor, probably of wood, was fitted, the waUs being embellished with a fine scheme of coloured frescoes including a zig-zag decoration and a large Chi Rho monogram set within an ornate wreath of leaves and buds. This recalls those that decorated the neighbouring Christian Room, which was decorated with a row of orante figures within a portico on the west wah and a second large Chi Rho monogram set within an even finer -wreath at the west end of the south waU. These 35 THE LULLINGSTONE ROMAN VILLA are the two rooms that were undoubtedly dedicated to Christian worship. The narrow builder's trench for this wall repah cut through the laminated burnt levels and contained a coin of Constantius II as Augustus, Fel. Temp. Rep., legionary spearing fallen horseman type, A.D. 345-61, thus dating this wah repah. That the floor fitted over the " drying " surface below was probably a wooden one is indicated by a pah of carbonized planks that were found lying on the " drying " floor. These planks lay the length of the room from west to east, parallel with each other, north and south of the guUy and midway between it and the north and south wahs. The absence of any carbonized cross-members ROOM IS • S E C T I O N G - H. CARBONIZED PLANK irri'i'i'i'i'i'i'iTi'i'i'i'l'i'iri'iTrT CARBONIZED PLANK \///U\ CLAY SEAL SURFACE OF 6th TREAD COIN'OF ALEXANDER SEVERUS TILE FALL. NATURAL CLAY ftND FLINT STEPS BEFORE SECTION LINE SHOWN DOTTED. I COIN OF ANTONINUS PIUS. SAMIAN cA.D. 2 00 No. S.I. WALL FALL. E ^ PAINTED PLASTER AND WALL FALL. ™ - BURNT LAYER, CONTAINING I/////1 COINS c.330 TO e.3SO. FEET FIG. 3 IIIMII 3rd CENTURY DEPOSIT. i'A'S'N ANTONINE DEPOSIT. or thin flooring may be accounted for by theh removal immediately before the final fire ; many hon nails were found, theh distribution being thickest on the line of the gully. These planks and the gully were sealed by the fallen painted plaster and this in turn was covered by the flint and mortar wah fall; the latter contained a little the, but there was not enough to indicate positively the type of roofing employed over the room. There was no sign of ceiling plaster. The length of use of the room in Phase IV as part of a Christian establishment is not known. It was ultimately destroyed by the fire that destroyed this north complex of rooms. A coin of Arcadius (A.D. 383- 408) was found in the top of the burnt level of Phase III. This coin 36 THE LULLINGSTONE ROMAN VILLA R O O M 15. PLAN OF FOURTH CENTURY LEVELS. ROOM It RAMP UP ° ° x ° s° ^PLANK \ TROUGH /PLANK/ 3RD CENTURY ARCHED + ! i - H" OPENING UNDER | O O M 16. VZft K CENTURY. EggiJ ZK CENTURY. ESS HZ CENTURY. IS33 GRAVEL. •'.-. POST HOLES. + COIN. 12 6 O I 2 3 4 S INCHES FEET ^ Ei«. 4 37 THE LULLINGSTONE ROMAN VILLA hes completely outside the date scale of Phase III, which is accurately attested by the thirty-one mutually supporting coins. It is clearly an intruder, and may weh have fallen between the boards of the floor in Phase IV. If this is so, then the existence of the floor as such in A.D. 383 plus must be envisaged, which would place the final destruction of the Christian estabhshment probably in the last decade of the century and not in A.D. 367 as was heretofore thought to be the ease. Room 16 As in the case of its neighbour Room 15, this room was formed in the second half of the thhd century by the insertion of the block containing the arch and the continuation eastwards of the north wall of Room 15. The second century fihing over the steps, and over that part of the second-century room that was included in Room 16, was allowed to remain undisturbed except for a channel leading eastwards on the axis of the arch. This channel seems to have been intended to convey heat from a point stih further towards the east, and as the ground has been cut away by construction of the road ah vestige of a furnace will have disappeared. The existence of this channel demonstrates that Room 16 was never the site of a furnace, and the system of heated apartments along a single axis is paralleled at South Darenth (Arch. Cant., XXII, PL G—The chalk ramps there may have influenced the draught, being similar in purpose to the clay baffle in our Room 15). The pottery recovered from the undisturbed second-century filling is illustrated at Figs. 12 and 13 (excluding Nos. 80 and 81). The contents also included much fragmentary wah plaster which, by its orange, red and green decoration, had clearly slipped into the fihing from the walls of the stair and corridor and is an additional indication of the period of abandonment of the Antonine house. The floor of the second-century room extended over much of the area up to the stah and corridor, and consisted of a layer of crushed brick. The north wah was built upon it, and it continued on the exterior of Room 16. It is clear that this floor is contemporary with the stair and corridor, and its Antonine period is founded upon the pottery at its level, which includes a datable Samian Form 67 (Fig. 6, No. S2). The remainder of the pottery is of native manufacture and necessitates a wide date-scale, but it does not appear to continue beyond the second century. The grey (thhd century) pottery-bearing layer, which was so conspicuous a feature of the stratification of Room 15, is absent from Room 16,' being replaced at a simhar level by a disturbed filling that seals the yellowish second-century layer. Modern post-holes have made this filling stratigraphically valueless. Near the west wah, at key-stone level of the arch, was an uneven floor of clay containing traces of the, representing the occupation level 38 THE LULLINGSTONE ROMAN VILLA in this room during the last years of the villa's existence. The keystone bricks of the arch were cut down, certainly after A.D. 345, and the four courses of flints running over the arch represent the base of the partition wah that was then remodehed. The mortar binding these four courses is strongly impregnated with pounded the, whhe for the whole height of the arch below it is white. The clay floor was laid down somewhat earher than the wah rebuild, as it is interrupted by the narrow wah trench ; but that it is a fourth-century construction is not open to doubt. The pottery recovered from its surface also conforms to that in use late in the century and is hlustrated at Fig. 13. The complete mortarium (No. 81), agreeing in type with Nos. 39 and 40 stratified and dated in the Basement Room, was found resting in an upright position on the floor and next the west wah. It sthl preserves its original contents (Appendix V) and had been repaired with rivets before its final use. The thick zone of charcoal representing the destruction is sealed by the flint and mortar wah fah, but Period IV wall plaster is absent, indicating that the walls of this room, unlike its immediate neighbours, remained undecorated. Above this wah fall the ground has suffered disturbance and no trace remains to indicate the type of roofing in use. As in the case of the Basement Room (Room 11) the presence of the modern road and sewer prevents complete excavation. Room 14 At the time when the stair complex was in use, this room may have formed the northern end of the west corridor of the Antonine house, a thick layer of rubbish of that date having accumulated outside the west wah and the north-west corner. The entrance was then in the west wah, its tread consisting of two horizontal wooden members butted against a longitudinal mortar core placed centrally (PI. Via). This entrance was made obsolete in the fourth century reconstruction by removal of the timbers and theh replacement by courses of heavy flints. New doorways were then made, one internally into Room 15, and the other externahy in the north wah (PL Via). The floor level was retained at the same height as the previous corridor floor, and by the insertion of a partition wah on the south, a small square apartment or vestibule was evolved, which continued in use throughout the remaining life of the house. The floor seems to have been surfaced with a thin spread of gravel, and it partially sealed the footing trench of the partition wall between Rooms 14 and 15. The interest of this room hes in its possible use, late in Period IV, as a vestibule attached to the Christian estabhshment. Its floor bore a thin scatter of coins indicating its use in the second half of the fourth century. The nine coins at floor level include three minims, two Fel. 39 THE LULLINGSTONE ROMAN VILLA Temp. Rep. coins, and one of the House of Theodosius (c. 379-95). These were associated with the final occupation of the house, and the remaining three coins are Oonstantinian and were probably in contemporary use. The floor ran over a refuse pit in the north-west corner, which may be dated by its sealed coins to a short time before the other pit referred to in the First Interim Report (Arch. Cant., LXIII, p. 11). The pit was dug after c. A.D. 337, and by the absence of Fel. Temp. Rep. ROOM 14 • SECTI ON J • K.. EDWARD t PENNY qrrflflTTrTITm M E D . A E V A L S U R F A C E •Mmm. ROMAN FLOOR LEVEL BOORWAY INCHES FEET RaaH YELLOW MORTAR AND CHALK (FALLEN WALL FACING) r r S l FLOOR LEVEL.GRAVELLY WITH TILE.CONTAINING MINIMS ^ " ^ AND COINS SECOND HALF 4 t h CENTURY. rrozg FOOTING TRENCH WITH MORTAR AND CHALK FILL, 1 3 3 2 3 CONTAINING WALL PLASTER. WM MORTAR AND SOIL MAKE-UP. m CLAY AND FLINT. (HILLWASH) . CLAY AND FLINT. (HILLWASH) , BROWN SANDY WITH ' CHALK. GREYISH SAND AND MORTAR. BLACKISH WITH CHARCOAL SANDY CLAY SEAL • i WOOD ASH. [ K M I BLACK SILT WITH FLINT AND TILE 12S!J CONTAINING COINS A.D.337-345. E«;°3 GRAVEL. |T»7J PRIMARYSILTING-DARK GRAVEL <>-*-* CONTAINING COLOUR- COATED ROSETTED WARE. © WALL FALL. (g, BLACK WITH POTTERY ^ 2niHo 4 t h CENTURIES. m SANDY CLAY WITH W TILE. / a NATURAL CLAY AND w FLINT. Eict. 6 coins, which are otherwise common on the site, it may be said to have been filled in and sealed over before c. A.D. 345. The coins, stratified beneath the final clay seal, are mutually supporting, except for a radiate of Tetricus II, which doubtless was a survival. The primary fill of the pit contained a sherd of colour-coated rosette-stamped ware, and its general sealed contents included three specimens of small thick cookingpots roughly handmade (Fig. 14, Nos. 82-4), indicating the use of hand-made pottery at this period. The presence also of fragments of 40 PLATE V — «v' t—ngjSa. Room 15, looking South, mtmjL^^f^^' rm. A b. Room 15. Post Hole complex. LULLINGSTONE ROMAN VILLA. [facep. 40 PLATE VI ' *&. ~w/^ Room 14, looking North. m b. Room 15. Ox skull in situ. LULLINGSTONE ROMAN VILLA. THE LULLINGSTONE ROMAN VILLA animal bones and skuhs showing knhe cuts, and also of the knives themselves, which wih be hlustrated with the iron work in a later report, suggest that the room may have been used as a kitchen or preparation room for the secular house before its transformation into a vestibule for the Christian estabhshment. The room seems largely to have escaped the final fire, and the presence of fallen tiles on its floor indicate the type of roofing. A few fragments of wah plaster were found in the footing trench next Room 15, and came from there ; Room 14 does not seem to have had decorated wahs. The two fourth-century entrances bear the imprints of the wooden treads, and the sides of the entrance leading into Room 15 are shaped to take massive doorway plaster and indicate that the door probably opened outwards into Room 14. By medieval times the area of the room had been made level with the truncated wah tops, and a shver penny of Edward I was found lying on the surface of the west wall. APPENDIX I THE SAMIAN POTTERY The analysis of Samian set out below is based upon finds up to the end of 1951. A total of ninety-three individual vessels have been identified, and they are analysed in the following table : A complete absence of the typical Flavian/Trajanic Form 27 is significant and gives the impression that the earhest cultural occupation must be a second-century one, though the fibulae and much of the coarse pottery are first-century types ; perhaps the earhest occupation was purely a native one. It will also be noted, that of the seventy-two dated vessels, thirty-four lie in the two decades c. 180-200. And on the assumption that the twelve examples of Form 37 survived into that twenty years, no less than 63 • 9 per cent, of the datable Samian lies therein. This may be regarded as indicating a late Antonine occupation ending c. 200, the three later examples probably being referable to the third-century occupation. Examples SI, S2 and S3 figured in this Report are rare and merit immediate ihustration. The sherds showing decoration are also figured, but a full range of illustrations is reserved for the final survey of the Samian at a later date. 41 Zf r-3 o CD W M CD H» CO o> 00 1-1 M 00 tO M M M M H1 CO h-» H* Ui to H* s1 & ff pi to M -» M M CO cn M M l-1 M b bo o I cn o 05 to t-l bo o o M - J M t-» )-* M cn to 1* b © © - J M M t-» b M CO oo cn cn b M I CO © H» M b OS I O H K b £ OS o M O tf* M M CO M b cn o CO CO ..... w s 1 - 3 to CO _.. K» M I ! ^ N h-» M *• M to M Period Total Form 30 Form 37 Form 15 • Form 18 Form 18/31 Form 31 Form 33 Form 35 Form 36 Curie 15 Form 79 Form 80 Form 44 Form 44/81 Form 81 Form 32 Curie 11 Form 38 Mortarium 43 Mortarium 45 Form 67 Barrel shaped tumbler VTTIA NVMOH aNOISONITim SOLL THE LULLINGSTONE ROMAN VILLA NOTES ON THE SAMIAN By DR. F. OSWALD, D.SC, F.S.A. In general, I think, the Samian falls mostly within the period of Niederbieber (A.D. 190-260), and of this, mostly between A.D. 190 and 200, Avith survivals, especially of the decorated fragments from the Rheinzabern kilns. There is also a complete absence of the cup Form 27, its place being taken by Form 33. 51 (Fig. 6). Form 67, with cut glass technique. Cf. Oelmann, Niederbieber, O. and P., LXXVII, 6 and Lud. VSb at Rheinzabern. Probably Rheinzabern ware, c. A.D. 200. Good glaze and high neck as at Rheinzabern. 52 (Fig. 6). Form 67. Very thin Motto beaker, showing in the central zone portions of letters CA.B formed by white barbotine dots, between two rows of similar dots. Cf. Oelmann, Niederbieber, with white barbotine, Taf. V, 23, 24 and see Behn, Romische Keremik, 10,4 and O. & P., LXXIX, 15. There is a very full collection of mottoes on wine vessels in Loeschcke's "Roman monuments of the wine industry on the Moselle, Saar and Rhine in the German Wine Museum at Trier, 1934." In this article the only motto which seems to me to meet the case, viz : . . .CA at the end of the phrase and the imperfect B (not I) at the beginning of the motto is on p. 44 viz.: A ME DVLCIS AMICA BIBE in the Wine Museum at Trier, and the white dotted spots occur on a small jug at Niederbieber (Oelmann, Niederbieber, Abb. 10-5, p. 33 and his Plate V, Fig. 25, and notably on the fine vase with the figure of TREBERIS, the city goddess of Trier, surrounded by a ring of these white spots with centre dot, found in the Trier potteries, his p. 5 Abb. 3). c. A.D. 200. 53 (Fig. 6). Ribbed barrel-shaped tumbler, the upper part with ribs or close-set horizontal mouldings in zones above and below the plain swollen centre of the vessel. Figured and described by J. A. Stanfield, " Unusual Sigillata ", Archaeological Journal, 1929, p. 130, Fig. 6, Nos. 29-33. Also recorded at Shchester and Corbridge by T. May, Pottery of Silchester, 1916, XXXI, p. 27. I found some examples of this rare vessel in the potteries of Lezoux in 1931, now in the Margidunum cohection of Nottingham University. Late Antonine date, c. A.D. 190-200, or may be somewhat earlier. 54 (Fig. 6). Form 37. Good glaze and neat execution. Ovolo of REGINVS ; his use of striated rods instead of beadrows, as on 37 REGINVSF, Knorr, Cannstalt, 1905, PL XXXVIII, 3. Eagle (Oswald 2166) used by REGINVS and many Rheinzabern potters, c. A.T>. 140-50. 43 THE LULLINGSTONE ROMAN VILLA J / ( \ N \ —1 SJOOOOOOO© © oo ooo sweooeoo • • L ) ooo / MHWyWCT. I Sib S20 PIG. 6 (J) 44 THE LULLINGSTONE ROMAN VILLA 55 (Fig. 7). Form 37. Intersecting circles, each arc containing an eight-rayed rosette with a central hollow used by REGINVS and a short stalked pinnate leaf, filling space between the intersecting circles, much used by REGINVS. Probably by REGINVS. 56 (Fig. 7). Form 37. Ovolo probably by REGINVS, c. 150-60. 57 (Fig. 7). Form 37. Medahion enclosing a small gladiator (Oswald 1027 and Ludowici 146 in his Katalog, V, p. 41), used by CERIALIS and other potters. The medallions are separated from each other by stout rods of fine pinnate leaves as on Ludowici No. 84, p. 87 of Katalog V, used by CERIALIS and some other Rheinzabern potters. This piece is probably by CERIALIS of Rheinzabern, c. A.D. 140-50. 58 (Fig. 7). Form 37. Medahion enclosing grapes on a stalk, used only by CERIALIS as on p. 94 of Ludowici's Katalog, V, No. 28. 59 (Fig. 6). Form 37. With ovolo of CERIALIS of Rheinzabern (Ludowici Katalog, V, p. 148, No. 100), c. 140-50. 510 (Fig. 7). Form 37. Rhomboid beads forming right-angled triangles, each enclosing an astragalus, used by CERIALIS and IVVENIS. Probably the work of CERIALIS, c. A.D. 140-50. 511 (Fig. 7). Form 37. Columns in series, exactly as on Form 37 used only by LVPVS (Ludowici Rheinzabern Katalog, V, p. 109, his No. 300). 512 (Fig. 7). Form 37. Protome of lion to left, exactly similar to Oswald 1494, even to the striations on the upper part of the right fore-paw (Lezoux, Dech. I, 720) as on a 37 mould stamped CARANTINIM. By CARANTINVS, Antonine, c. A.D. 150. 513 (Fig. 7). Form 37. Lower part of basket (by CINNAMVS of Lezoux). Rather thin ware. Antonine, c. A.D. 150. S14(Fig. 7). Form 37. Clearly the work of DIVIXTVS of Lezoux. The small medallions with a hohow central core occur on a 37 stamped DIVIX.F at Carhsle (May and Hope, Catalogue of Roman pottery in the museum, Tuhie House, Carhsle, No. 63), and the same medalhon with central rosette at Leicester (Jewry Wah Site). The two goat's legs belong to a Pan (Oswald 717A) and the Acanthus occurs on a 37 stamped DIVIX.F (May and Hope op. cit., No. 64) with the same small rosettes on the bead rows as in the medallions, A.D. 140-50. S15 (Fig. 7). Form 37. Two conjoined dolphins above basket. Probably by CINNAMVS. Sharply modelled head with hah forming a crest, not identifiable. Antonine. S16. (Fig. 6). Smah rouletted Form 30. Cf. O & P., LXXV, 15, Heiligenberg. Also made at Lezoux. It is probably part of a bowl Oelmann's Type 16 at Niederbieber, figured in his Taf. 1,16. c. Antonine, probably later, c. A.D. 200 45 THE LULLINGSTONE ROMAN VILLA 517. Form 15. Profiling similar to O. & P. XLIII, 37. Flavian, but it might be Trajanic, or possibly a survival. 518. Form 18. Height of wall c. 30 mm. Cf. 0 . & P., XLV, 15, good glaze. Possibly Flavian, or probably later ; date uncertain in spite of theh similarity. 519. Form 18. Part of rim. Perhaps Flavian. S20 (Fig. 6). Form 18/31. Height of wah 46 mm. Similar to 18/31 Wiesbaden Kasteh 0. & P., XLVI, 3, good glaze. Graffito, X, in chcle. Probably Trajanic or later. a S7 S II I S I* S IO cm ^3 S12 S)$ S 8 S 13 So PlG. 7 (i) 521. Form 18/31. Simhar to S20. Probably Trajanic or later. 522. Form 18/31. Height of side 33 mm. Cf. 0. & P., XLV, 15 at Gellygaer, Domitian-Trajan. Probably Dornitian or later. 523. Form 18 or 18/31. Fragment of base with double notches in chcle. Possibly Trajanic. 524. Form 31 (Sb). Cf. 0. & P., XLVII, simhar diameter, height of side 57 mm., thicker moulding. Antonine. 525. Form 31 (Sb). Cf. 0. & P., XLVII, 7, Rheinzabern. Antonine. 526. Large Form 31 (Sb). Of. 0. & P., XLVII, 5 and 7, but with thicker moulding, and -with bevel just below the moulding. Oelmann,, 46 THE LULLINGSTONE ROMAN VILLA Niederbieber Taf. I, IC. Cf. also Oswald, Samian of Margidunum (University of Nottingham), Pis. LI and LII, 31 (Sb) from the third century weh R, at 10 ft. Date c. A.D. 200. 527. Form 31 (Sb). Probably the same dish as S24, bevel just below the thick moulding. Rheinzabern ware, c. A.D. 200. 528. Large Form 31 (Sa). Rather simhar to O. & P. XLVII, 5, Rheinzabern, c. A.D. 180-200, and cf. Oswald, Samian Ware of Margidunum Pis. LI and LII, diameter c. 240 mm. in thhd century well R. 529. Large 31 (Sa) with internal offset, cf. 31 (Sa) in third century weh at Margidunum, c. A.D. 200-250. 530. Large 31 (Sa). Height of side 54 mm., thick moulding as at Niederbieber (Oelmann Taf. I, lb), c. A.D. 200-250. 531. Large 31 (Sa). Straight sided wall, c/. Niederbieber (Oelmann la) c. A.D. 200. 532. Large 31 (Sb). Similar to O. & P., XLVII, 7, Niederbieber, with a bevel below the thick moulding. Probably same as S26. c. A.D. 200. 533. Large 31 (Sb). Thick moulding, c. A.D. 200-250. 534. Large 31 (Sb). Diameter 276 mm. Cf. O. & P., XLVII, 5, smah moulding. Probably Antonine, c. A.D. 150. 535. Large 31 (Sb). Probably late second century, insufficient for exact dating. 536. Large thick 31 (Sb). Cf. Oswald, Samian of Margidunum, LII, 2, in weh R (10 ft. deep), but with thicker moulding. Date about A.D. 200. 537. Form 31. Large moulding, but insufficient for exact dating. Probably thhd century. 538. Large 31 (Sb). Probably the same as S26, with bevel just below the thick moulding. Rheinzabern ware, c. A.D. 200. 539. Form 33. No external central groove. Possibly Hadrianic. 540. Form 33. Height of side 62 mm. Cf. O. & P., LI, 10, with external groove, good glaze. Probably later than Trajanic. 541. Form 33. External groove centrally, cf. O. & P., LI, 10. Domitian-Trajan ; more probably second century, for there are no traces of the cup Form 27. 542. Form 33; Much flaked, good glaze and probably similar to S41. 543. Form 33. Insufficient for exact dating, but somewhat simhar to S41 and S42 in glaze and contour. Cf. O. & P., LI, 10. Perhaps Trajanic or later. 544. Form 33. Insufficient for exact dating. Perhaps Trajanic. 545. Form 33. Lower part, flaked on inside. Too fragmentary for identifying or dating ; perhaps Hadrianic. 47 THE LULLINGSTONE ROMAN VILLA 546. Form 33. Fragment of side, cf. 0. & P., LI, 14. 547. Form 35. Smah. High glaze, stalks of barbotine leaves. Perhaps Lezoux ware,, cf. 0. & P., LIII, 3, Newstead. End first century into second century. 548. Form 36. Smah. Stalks of barbotine ivy leaves on flange, internal groove. Uncertain date, possibly Trajanic. 549. Large 36, with barbotine ivy leaves, cf. Pan Rock 36, O. & P., LIII, 15. Late Antonine, perhaps A.D. 170. 550. Smah 36. Fragment of rim with stalks of barbotine leaves. Cf. 0. & P., LIII, 12, Treves pottery, Hadrianic. 551. Small 36, with stem of barbotine leaves, good glaze. Cf. O. & P., LIII. 552. Form 36. Imperfect, indeterminate. Probably second century. 553. Form 36. Part of flange, plain, no barbotine, probably second to thhd century. 554. Form 36, with rectangular footstand, flange with barbotine ivy leaves. Cf. 0. & P. LIII, 17, Niederbieber I, 46. c. A.D. 200. 555. Large 36. Probably Lezoux ware. Flange rim, plain. Probably thhd century. 556. Curie 15. Fragment of rim, smaher size, diameter c. 120 mm. Cf. O. & P., LVI, 4, Heiligenberg. Probably Antonine. 557. Quite insufficient to determine the nature and form of the vessel positively, but it is probably part of the base of a 79. Cf. 0. & P. LVIII, 1, or more probably like the 79 from Niederbieber (Oelmann, Niederbieber 5b, which he erroneously designates as a 32), but any exact determination is impossible in the absence of the rim. Date perhaps c. A.D. 200. 558. Form 44. Lezoux ware, figured by Curie (Newstead, XL, 20 in section, and in the group on Plate XLI, 1, p. 200), bright glaze, whilst the Newstead example had a dull red glaze, probably due to weathering. It is late Antonine, probably A.D. 160-70. 559. Large 44 or 81. With rectangular footstand, chcle of striations on inside of base. Cf. 0. & P. LXI, 8. c. A.D. 200. 560. Plain rim with very faint and indistinct moulding, possibly of a thin 32. Cf. 0. & P. LXIII, 10. Date, c. A.D. 200. 561. Flange of bowl Curie 11 with stalk of barbotine leaf. Possibly Flavian, cf, 0. & P. LXXI, 10. 562. Fragment of possibly flange of Curie 11. Possibly Trajanic or later. 563. Mortarium 43. Lower part, like 22 of Oelmann, Niederbieber, from which the fine grit of the interior has been rubbed off. The two fluted grooves are visible on the exterior. 564. S65. Mortarium 45. Four fragments of two different 48 THE LULLINGSTONE ROMAN VILLA examples, the thickest of the flanges shows part of a hon's tah in barbotine. End second century to thhd century. 566. Mortarium 45. Fragment of lower part of plain flange, no external groove as in most Lezoux examples ; a slight internal groove. End second century to thhd century. 567. Mortarium 45. Fragment of lower part of plain flange, no external groove ; but internal groove present. Good glaze, fine quartz grit in interior. Cf. 0. & P., LXXIV, 4, Niederbieber. End second century or beginning of third century. 568. Mortarium 45. Lower part of plain flange, quartz grit on interior as at'Niederbieber. End second century to thhd century. 569. Small fragment of base of an 18 with a circle of fine striations, the so-called " engine turning " ; graffito E underneath, undated. POTTERS' STAMPS 570. Partial stamp FL (ORVS F inside a chcle of striations, the so-cahed " engine turning ", as Niederbieber, on a Form 46. c. A.D. 200. 571. SIICA equals SIIDA(TVS F of Heiligenberg (the D is often rendered as C in illiterate stamps). This stamp is on a large 33. c. A.D. 130-40, perhaps later. 572. This belongs to the class of meaningless potters' marks. It is on a Form 33. i \I\IWI-\-1 which cannot be identified. 573. Stamped LIBERTI M on flat base, too wide for a 33, and therefore probably on Form 80, as at Rouen or Barentin. Although LIBERTVS is typically a Trajanic potter, he evidently worked on into Hadrian's reign. 574. Base of 18/31 stamped SACRILLI M. SACRILLVS of Lezoux, Antonine period, c. A.D. 150-60. This stamp is fairly frequent in Britain, e.g., it occurs on the Pan Rock (SACRILLIM), on the plain rim of Form 37 (Richborough Report, II, p. 66), and on the plain rim of a 37 by DOECCVS at Baldock. APPENDIX II THE COARSE POTTERY In Rooms 14, 15 and 16, the stratigraphical sequence is founded upon positive coin and/or Samian evidence, and the coarse pottery from these dated levels is hlustrated in groups. The pottery from Room 11 and the miscellaneous pottery is associated with the final destruction of the house, the series of mortaria in Room 11 being of especial interest, as it indicates the mortaria forms that were in use and that may be met with very late in the fourth century. The same 49 4 THE LULLINGSTONE ROMAN VILLA remark apphes to the two large storage jars (Nos. 91 and 92) which, together with No. 12 (First Interim Report, Arch. Cant., LXIII, p. 21), form a class of vessel which has not, it is believed, appeared previously in a dated series. From a study of this coarse pottery, three points emerge : (1) Native ware, often of Patchgrove type, tends clearly to persist towards the end of the second century. (2) Handmade cooking pots of rough construction and fabric were hi use, probably more widespread than is at present apparent, as early as the second quarter of the fourth century. (3) Fourth-century mortaria, and colour-coated ware both plain and with applied painted designs, continued in use without modification towards the close of the century, and there is no evidence of theh gradual substitution by coarser ware, though there is evidence of theh individual repah, indicating perhaps a difficulty in replacement (No. 81 is a case in point). ROOM 11 (THE BASEMENT ROOM). (Figs. 8 and 9.) 33. Bulbous beaker of fine, hard pink ware, with grey-green shp on interior and exterior. Decorated round the bulge diagonally en barbotine. Ref. May, Silchester Pottery, PL LII, No. 83, p. 122, where the attribution is to the first hah of the thhd century. From the sump or water channel. 34. Pedestal bead rim bowl with short upright neck and bulging under towards base. Round the bulge a band of scroll in white paint bordered with bands of rouletting. Hard weh baked pink ware with deep red shp. Second half fourth century. Ref. May, Silchester Pottery, PL LVI, No. 99, p. 128. From between wah fahs. 35. Pedestal bead rim bowl with short upright neck and bulging under towards base. Round the bulge a band of scroh in white paint bordered above and below with bands of rouletting. Hard well baked pink ware with hght red shp. Second hah fourth century. Ref. as for 34. From between wah fahs. 36. Conical necked bulbous beaker in fine weh baked pink ware. Greenish brown metallic glaze on exterior. Mid to late fourth century. Ref. Lockleys, Fig. 12, No. 6, p. 373 for form. From cohapsed debris of Christian room. 37. Large wide-mouthed two-handled flagon of hard red brown ware with orange red burnishing on exterior. Zones of horizontal grooves below base of handles, and vertical toohng on neck. Handles adjoin the lip. Mid fourth century. 50 LULLINGSTONE ROMAN VILLA 34 mwMWmm FIG. 8 ($) 51 THE LULLINGSTONE ROMAN VILLA "•"""" iV* :<>".o?. m * • "a •* P '* V • •«• o* ^ F b 4-3 FIG. 9 (I) 52 THE LULLINGSTONE ROMAN VILLA From collapsed debris' of Christian room. 38. Large store jar with heavy bead rim. Weh fired brown ware with grey Patchgrove type core. Dark grey shp on interior and exterior. Burnished on interior of bead and for three-quarters of an inch below bead on exterior. Single band of horizontal stabbing on neck below zone of burnishing. Zone of vertical burnishing towards base. Its Flavian date is confirmed by Mr. A. W. G. Lowther, F.S.A., who kindly examined it. Used as wedge for the pedestal of Bust II on step (PL lb). Group of five complete mortaria from floor of Basement Room at final destruction of house : 39. Mortarium. Smah hght red vessel. Quartzite grit on interior. Right angled flange with downward rounded edge. Mid fourth century. Ref. Richborough, I, PL XXVIII, No. 97. 40. Mortarium. Smah light red ware with blue core. Cream slip on interior and exterior, quartzite grit on interior. Top of bead and edge of flange decorated with shallow grooves. Right angled flange with downward rounded edge. Mid fourth century. Ref. Richborough I, PL XXVIII, No. 97. This vessel is a completed reconstruction of No. 32 (First Interim Report, Arch. Cant., LXIII, p. 20). 41. Mortarium. Badly burned specimen. Drooping flange, high bead out-turned at top. Quartzite grit on interior. Mid fourth century. Ref. Richborough I, PL XXVIII, No. 98. 42. Mortarium. Large pink ware with cream slip on interior and exterior. Quartzite grit on interior. Right angled flange, the downward sloping edge of which is decorated with a shallow groove. Mid fourth century. Ref. Richborough I, PL XXVIII, No. 101. 43. Mortarium. Large pink ware with cream slip on interior and exterior. Quartzite grit on interior. Right angled flange with downward sloping edge. High bead out-turned at top. Mid fourth century. Ref. Richborough I, PL XXVIII, No. 101. ROOM 15. POTTERY FROM LATE ANTONINE DEPOSIT. (Fig. 10.) 44. Amphora Stamp. ACIRCI. Upwards on curve of thick chcular handle tapering towards base. Ref. Richborough IV, p. 242, where it is second' century. 45. Outcurving rim of large olla of Patchgrove fabric. Light brown soapy slip on exterior. Dark grey on interior, with burnishing on inside of rim. Horizontal rilling and toohng on neck down to slightly angular shoulder. Flavian, probably persisting weh into second century. 53 THE LULLINGSTONE ROMAN VILLA 4* ^ 4 8 ^ 1 45 146 50 ¥b 47 7 49 O X *ts4 ^55 ^\56 H^7 58 FIG. 10 (i) 54 THE LULLINGSTONE ROMAN VILLA f bb J LtfX'Mtt'COGt'ttWl*. W>k'^ VS>«- FIG. 11 (£) 55 THE LULLINGSTONE ROMAN VILLA 46. Bead rim bowl, well fired grey ware. Soapy slip on exterior. Persisting Belgic type. 47. Rim of oha outbent and recessed to receive a lid. Hard well fired, coarse gritty cream ware. Ref. Richborough I, PL XXV, No. 53. 48. Pedestal bowl with out-turned rim. Hard weh fired dark grey ware. Horizontal double groove on neck with single groove above bulge. Persisting Belgic type. 49. Beaker with everted rim and indented sides. Burnishing on rim and exterior of neck. Double horizontal groove below neck and above indentations. Ref. Richborough III, XXXIX, No. 305. Group of shallow Pie Dishes, hard well fired dark grey ware : 50 and 51. No bead. 52 and 53. Bead appearing. 54 to 58. Distinct bead. ROOM 15. POTTERY FROM THIRD CENTURY DEPOSIT. (Fig. 11.) 59. Pedestal vase, with outcurled rim. Five indentations on sides. Hard weh fired dark grey ware. Burnished on top of rim, outside neck, and below indentations to base. Horizontal groove at base of neck and above indentations. Late thhd century. 60. Hemispherical bowl, hard weh fired ware with outcurling rim. Horizontal line of oblique stabbing to simulate rope design, and a wavy line below, with horizontal double groove beneath. Graffito (not identified). The shape of this vessel shows early characteristics, but it here appears in a thhd century context. 61. Small beaker of hard weh fired grey ware. Outcurled rim. Zones of burnishing on shoulder and base, with four horizontal burnished lines between the zones. 62. Oha. Hard weh fired light grey ware with reddish sandy surface on exterior. Cornice rim with slightly pointed downward edge. 63 and 64. Pie dishes or flanged bowls. Deeper than 54 to 58, and bead has been drawn out and down, making a flange with an inoipient rim. 65. 011a of Patchgrove type. Hard well fired grey orange ware with cordons on neck. A survival. 66. Storage jar of Patchgrove type. Heavy outcurling rim with zone of stabbing on shoulder. A survival. ROOM 15. POTTERY FROM THE LATEST FOURTH CENTURY LEVEL. (Fig. 11.) 67. Oha. Out-turned rim with external groove. Hard well fired red ware with brown shp on exterior. Zones of darker brown burnishing below neck and above base with a single simhar zone between. 56 THE LULLINGSTONE ROMAN VILLA 68 FIG. 12 (J) 57 THE LULLINGSTONE ROMAN VILLA J 73 ' V V79 FIG. 13 (£) 58 THE LULLINGSTONE ROMAN VILLA ROOM 16. POTTERY FROM LATE ANTONINE DEPOSIT. (Figs. 12 and 13.) 68. Oha. Patchgrove ware, hard well fired, hght brown. Large outcurled rim. Soapy surface on interior and exterior. Horizontal rilling on neck derived from cordons. Zone of burnishing on shoulder. 69. Oha. Patchgrove ware, with large outcurled rim. Well fired, brown. Soapy surface on interior and exterior. Horizontal rilling onTtieck. Burnishing on top of rim. 70. 011a. Patchgrove ware, with outcurled rim, somewhat smaller than 68 and 69. Soft, soapy, light brown ware. Horizontal rilling on neck with zone of stabbing on shoulder. 71. Cooking bowl. Patchgrove ware, with outcurled rim. Hard well fired brown orange ware with light grey slip. Rough on interior. Exterior soapy and burnished with tooling. A close group of six vertical lines, painted brown, at opposite ends of one diameter. Base has tooled cross on underside. A native type with early characteristics. 72. Cooking vessel. Bead rim recessed for lid. Hard well fired brown ware with sandy, red-grey surface on exterior. 73. As for 72. 74. One handled mug. Hard light grey ware. Zone of coarse tooled lattice lines on exterior, with horizontal tooled and burnished lines towards base. Handle roughly chcular in section. Ref. Richborough III, PL XL, No. 322. 75 to 79. Shallow pie dishes. Distinct bead. No. 75 with external tooled lattice, which is a rare feature at Lullingstone on these vessels. ROOM 16. POTTERY FROM LATEST FOURTH CENTURY LEVEL. (Fig. 13.) 80. Flanged bowl. Hard weh fired grey ware, with orange shp on interior. The interior is also decorated with an acute lattice design with parallel bands of horizontal lines apphed over the lattice. Second half fourth century. Ref. Lullingstone First Interim Report, Nos. 11 and 28 (Arch. Cant., LXIII, pp. 16 and 20). 81. Colour coated mortarium. Hard weh fired pink ware with red slip. Quartzite grit on interior. Square rim with high bead. Broken and repahed before final destruction of house indicated by three sets of two holes, two sets drilled in rim and one set drilled midway between rim and base. Second half fourth century. Ref. Richborough I, PL XXVIII, No. 99, p. 102. The contents of this vessel are discussed at Appendix V. ROOM 14. POTTERY FROM P IT DATED C. A.D. 337 to c. A.D. 345. (Fig. 14.) 82. Smah handmade cooking-pot. Coarse grey ware, well fired with rough tooling on exterior. Ref. Richborough III, PL XL, No. 333. Fourth century. 59 THE LULLINGSTONE ROMAN VILLA r 1 86 \ \ S9 JL 93 85 7 87 • ^ FIG. 14 (J) 60 THE LULLINGSTONE ROMAN VILLA ( 91 a V r 92 I 1 FIG. 15 ($) 83. Squat handmade cooking-pot. Coarse, hard grey ware. Carinated shoulder, very rough interior. Exterior smoothed by tooling. Fourth century. • 84. Handmade cooking-pot, roughly made, soft, brown ware, with burnishing on exterior of body. 85. Globular cooking-pot of hard gritty grey ware. Rough interior, interrupted zone of tooled lattice line on exterior. Probably wheel made, but simhar to 82, 83 and 84. Fourth century. 61 THE LULLINGSTONE ROMAN VILLA 86. Cooking-pot of roughly made soft brown ware. Everted rim, outcurling, with burnishing on hp. Slight cordon below rim. Probably wheelmade, but simhar to 84. 87. Beaker. Hard grey ware. Everted rim with one groove below neck and another below shoulder. Burnished rim and shoulder. 88. Flanged bowl. Hard weh fired grey ware. Smooth surface on interior, exterior rough. Tooled horizontal shallow grooves on exterior. Fourth century. 89. Flanged bowl. Hard brown ware. Smooth surface on interior and exterior. Wavy band of toohng on exterior. Rim drawn out at right angles to body and vertically. 90 Mortarium. Hard weh fired pinky cream ware, with quartzite grit on interior. Tooled groove on bead and above zone of grit. Fourth century. MISCELLANEOUS POTTERY FROM LATEST FOURTH CENTURY LEVELS. (Figs. 14 and 15.) 91. Dolium or large storage jar. Hard grey ware, weh fired, heavy bead rim. On shoulder, a band of combed feathering and a band of curvilinear feathering divided by a burnished zone. At girth, four zones of interlacing wavy lines, and a fifth immediately below bead. Ref. Alice Holt, p. 12. PL 1. Fourth century. 92. Dolium or large storage jar. Hard well fired, grey ware, heavy bead rim. Zone of vertical and curvilinear combing between two zones of burnishing. Bead rim as for 91 but a smaller example. Fourth century. Ref. as for 91. 93. Oha, with outbent moulded rim. Weh fired sandy buff ware with zone of regular horizontal rilling on body. Fourth century. Ref. Richborough IV, PL XCIII, No. 470, p. 269. APPENDIX III THE GLASS By BERNARD C. MIDDLETON, F.R.S.A. The yield of glass resulting from the work during the period under review has been no less rewarding than that of the first year, though numerically it was inferior. As before, the glass may conveniently be divided into two main categories, (a) vessels, (b) window-panes. The season's yield was 404 fragments of the former, and 422 of the latter ; also five beads and one ring(?) stone (No. 54). 62 THE LULLINGSTONE ROMAN VILLA No complete vessels were found ; neither were there fragments of mihefiori or other comparatively rare first century polychrome wares. Owing to the general lateness of the levels these could not be expected. However, a number of the fragments are interesting. Of these the most important is No. 45, an Antonine goblet with stem and base. This type may with some probabihty be cited as the prototype from which the wine glass with drawn stem developed. Several fragments are wheel-cut, some of the work being good, the remainder indifferent. This technique, together with apphed threads of the same colour as the vessels, and the moulded star-shaped device on the base of a bottle (No. 46 (4) ) are the only forms of decoration to be found in the collection. As is generally the case the majority of the glass is green, varying in tone from yellow-green to blue-green. The most important of the few blue pieces is No. 55, part of a pillar-moulded bowl—a common early form, but not hitherto discovered on this site. A fah number of fragments of colourless glass were found, most of these apparently belonging to bowls ; as is usual with colourless glass (or so I have noticed), most of it is suffering from " milky " weathering. Not all the rims and other interesting fragments are pubhshed in this Report, but those selected are intended to be representative of the whole. The greater part of the window-glass is comparatively thin (about xg- in.), often smooth on both sides and generally yehow-green in colour. The thicker, obviously moulded pieces are more often blue-green. A few pieces are colourless. One fragment (No. 44) is evidence of a triangular pane simhar to those found in the first year. These pieces are of interest because, as far as I know, nothing of a similar kind has been pubhshed in this country before. The dating of the examples is that of their dated levels or stratification. The date of theh manufacture or use cannot, at present, be more closely assessed. For the purpose of convenience and condensation the following abbreviations are used in the Catalogue : " ADR "—Approximate diameter at rim; "ADB"—Approximate diameter at base ; " ( a ) "— colour of glass ; " (b) "—degree of iridescence or type of weathering ; " (c) "—decoration ; " (d) "—technique of manufacture. Other abbreviations are self-evident. Rim measurements marked with an asterisk should be regarded as a closer approximation to the original dimension than those not so marked. The designation of some vessel forms must necessarily be arbitrary owing to the smahness of the fragments. I am indebted to my wife for her help with the work involved in the preparation of this short interim Report. 63 THE LULLINGSTONE ROMAN VILLA CATALOGUE 44. Window-pane, fr. of rim.,\va.. thick, (a) pale blue (a) moulded, with tool marks on upper surface. One edge chipped to an angle of 45 degrees from moulded rim. This piece is remarkably similar to fragment B, No. 43, Fig. 9 in the First Interim Report. (Not illustrated.) 45. Goblet, frr. of body, stem and base. ADB. 1|4 hi.* (a) tinted yeh.-gr. (b) milky weathering (c) apphed thread of same colour on under side of body. Late Antonine. (Fig. 16, No. 45.) 46 (1). Bowl, fr. of rim. ADR. 2 | in.* (a) tinted yeh.-gr. (b) milky weathering (c) series of shahow incised hnes starting § in. below rim. Late Antonine. (Fig. 16, No. 46 (1).) 46(2). Bowl, fr. of rim. ADR. 4 in. (a) colourless (b) mod. Many minute bubbles. Late Antonine. (Fig. 16, No. 46 (2).) 46 (3). Bowl, fr. of rim. ADR. 4j in.* (a) tinted yeh.-gr. (b) milky weathering. Late Antonine. (Fig. 16, No. 46 (3).) 46 (4). Bottle, prob. straight-sided, fr. of base, (a) mod. bl.-gr. (b) mod. (c) moulded marks, seemingly forming part of a four-pointed star and corner device. Late Antonine. (Fig. 16, No. 46 (4).) 46 (5). Bottle, straight-sided, fr. of side. Width of side approx. If in. Apparently tah, narrow variety of the " common square ". (a) mod. bl.-gr. (b) mod. Late Antonine. (Not ihustrated.) 47. Bowl, fr. of rim. ADR. 2 ^ in. (a) colourless (b) incipient. Thhd century. (Fig. 16, No. 47.) 48. Bowl, fr. of side, (a) mod. yeh.-gr. (b) incipient (e) fine, apphed threads of same colour, unmarvered. Inner surface crissled. Prob. fourth century. (Fig. 16, No. 48.) 49. Bottle, fr. of lip and neck. ADR. 1-j— in.* (a) mod. bl.-gr. (b) incipient. Tool marks on top surface of lip. First-second century. {Fig. 16, No. 49.) 50. Bead, roughly circular. Diam. -| in. (a) dark blue. Unstratified, prob. fourth century. (Fig. 16, No. 50.) 51. Bowl, fr. of rim and side. ADR. 2^. in.* (a) tinted yeh.-gr. (b) outer surface pitted (c) shahow, wheel-cut line. Unstratified, prob. fourth century. (Fig. 16, No. 51.) 52. Vessel, fr. of side, (a) tinted yeh.-gr. (b) mod. (c) apphed, unmarvered thread of same colour, cross-tooled. Unstratified. (Fig. 16, No. 52.) 53. Bowl,fr. of rim. ADR. 3 | in. (a) colourless (b) milky weathering. Fourth century. (Fig. 16, No. 53.) 54. Stone, prob. for finger-ring, (a) mod. bl.-gr. Well made. Fourth century. (Fig. 16, No. 54.) 55. Bowl, pillar-moulded, fr. of side, (a) dark cobalt-blue (b) heavy, with pitting (e) pillar of same colour ; pressure from outside has 64 THE LULLINGSTONE ROMAN VILLA 1 * 7 ^ 4-6W W(,W sod) S4-W ^ SI SS 4-6U) \ 4-6 « ffc'h '• **. y, *9 4-6<*> 4-8 r 52 53 56 W 57 580) \ 66(0 saw Ssf. I.- &7 ** 59 V 65(2) 6 f 7+o; 74-U) FIG. 16 Glass Vessels. Scale : f except where otherwise indicated. 65 THE LULLINGSTONE ROMAN VILLA resulted hi slight bump on the inner surface. Antonine. (Fig. 16, No. 55.) 56. Bowl, frr. of rim. ADR. 4 | in.* (a) mod. yell.-gr. (b) incipient. Rim ground flat. Fourth century. (Fig- 16, No. 56.) 57. Bottle, fr. of rim, neck and handle. ADR. 2 | in.* (a) mod. bl.- gr. (b) mod. Rim formed by pressing the metal outwards, and then folding it back on itself to give double thickness. Antonine. (Fig. 16, No. 57.) 58 (1). Flask (?), possibly part of one end, thefr. being a segment of an ellipse, (a) colourless (b) milky weathering and shght iridescence (c) apphed thread of same colour. Third-fourth century. (Fig. 16, No. 58 (1).) 58 (2). Bowl or cup, fr. of rim. ADR. 3 in.* (a) mod. bl.-gr. (b) mod. Fourth century. (Fig. 16, No. 58 (2).) 59. Jug, fr. of rim. ADR. 2 | in.* (a) mod. yeh.-gr. (b) incipient. Unstratified. Prob. late second-early fourth century. (Fig. 16, No. 59.) 60. Bottle, flat-sided, fr. of base, (a) mod. bl.-gr. (b) mod. Antonine. (Not hlustrated.) 61. Bottle, flat-sided, fr. of base, (a) mod. bl-gr. (b) incipient. Antonine. (Not illustrated.) 62. Bottle, flat-sided, prob. four-sided, fr. of neck and slwulder. Diam. inside base of neck 1| in. (a) hght bl.-gr. (b) mod. first-early second century. (Not hlustrated.) 63 (1). Bottle, flat-sided, fr. of base and side. Average thickness of metal •§ in. (a) mod. bl.-gr. (b) considerable. Fourth century. (Not ihustrated.) 63 (2). Cup,'fr. of base and side. ADB. 1 f-J- in. (a) light yell.-gr. (b) incipient. Fourth century. (Fig. 16, No. 63 (2).) 64. Bowl, fr. of side. Approx. diam. at lower thread 3-pg- in. (a) light bl,-gr. (b) mod. (c) applied, unmarvered threads of same colour. Fourth century. (Fig. 16, No. 64.) 65. Bowl or cup, fr. of side near base. Fr. .§ in. wide and high. (a) colourless (b) heavy (c) many shallow, parahel incised lines, certainly running horizontally. Unstratified. (Not ihustrated.) 66 (1). Bowl, fr. of rim. ADR. 7 | in.* (a) mod. yell.-gr. (b) considerable. Many small bubbles. Rim ground flat. Fourth century. (Fig. 16, No. 66 (1).) 66 (2). Bowl, fr. of rim. ADR. 6 | in. (a) tinted yeh.-gr. .(b) incipient. Patches of massed, minute bubbles. Rim ground flat. Fourth century. (Not illustrated.) 67. Cup, fr. of base and side. ADB. 3T 0B in. (a) tinted yeh.'-gr. (b) mod. Sides vertical at turn from base. Third-fourth century. (Not ihustrated.) • ' • • • • 66 THE LULLINGSTONE ROMAN VILLA O I A G P . A M A T I C S K E T C H OF * n> » M 3 Obverse and Reverse Obv. IMP. C. M. AVR. SEV. ALEXAND AVG. Bust r. laur. dr. Rev. IOVI CONSERVATOR!. J . s t . 1. cloak over r. arm holding thunderbolt and sceptre. Obv. IVLIA MAMAEA AVG. Bust r. diad. Rev. VESTA. V st 1. with palladium and sceptre. Obv. IMP. C. POSTVMVS P.P. AVG. Bust r. rad. dr. Rev. PAX AVG. P. st. 1. with branch and trans, sceptre JPJ Obv. IMP. C. CLAVDIVS AVG. Bust r. rad. dr. Rev. GEND7S EXERCI G. st. 1. with patera and comucopiae Indeterminate. Obv. Bearded head r. rad. dr. Rev. PAX AVG. type. Obv. C. P.E. TETRICVS CAES. Bust r. rad. dr. Rev. PIETVS AVGG. Pontifical instruments. Obv. IMP. C. ALLECTVS P.F. AVG. Bust r. rad. • cuir. Rev. PROVIDENTIA AVG. P. st. 1. holding baton and cornucopiae; at foot globe S1 P CL (Camulodunum) Obv. IMP. C. ALLECTVS P.P. AVG. Bust r. rad. Rev.'PAX AVG. P. st. 1. with branch and tr. sceptre. Indeterminate. Rev. SPSS ? Obv. CONSTANTINVS AVG. Bust r. diad. dr. Ref. 141c 1 360 1 318 1 and note 48 1 1 100 1 255 1 111 1 33 1 1 Cohen 20 2 69 50 THE LULLINGSTONE ROMAN VILLA Emperor Urbs Roma Constantinopolis Hybrid Overstrike Helena Theodora Date A.D. 330-37 330-37 1 Coin >» yy yy yy yy yy M yy »> (small) yy (small) Obverse and Reverse Rev. BEATA TRANQVILLITAS. Globe on altar insc. VOTIS XX 3 stars above. .PTR, .STR. Obv. CONSTANTINVS MAX AVG. Bust r. diad. cuir. Rev. GLORIA EXERCITVS. One standard. Obv. as above. Rev. Two standards TRP. Obv. CONSTANTINVS PF AVG Bust r. laur. cuir. Rev. PROVIDENTIAE AVG. Camp gate Obv. CONSTANTINVS AVG. Bust r. laur. Rev. SARMATIA DEVICTA. Victory r. with captive. Obv. DIV CONSTANTINVS P.T. AVGG. Bust r. veiled. Rev. Emp. in quadriga hand in aky. Usual type. TRS., TR.P., TR.S., w PLG., SLG.? Usual type. TRP., TR.P., TRS., TRSV. Constantinopolis—Urbs Roma. Obv. Bust r. diad. on Rev. Constantinopolis. Rev. EEL TEMP REP legionary on head helm- 1. Obv. FL IVL HELENAE AVG. Bust r. dr. Rev. PAX PVBLICA. P. st. 1. with br. and tr. sc. TRP. Obv. EL. MAX. THEODORAE AVG. Bust r. dr. Rev. PIETAS ROMANA Empress nursing child PLG. Ref. 250 1 255 1 452 1 487 1 760 1 11 5 1 1 4 2 4 2 70 THE LULLINGSTONE ROMAN VILLA Emperor Crispus Constantine II as Caesar Date A.D. 317-26 317-37 Coin M 3 >> >> t> >t >j tt »» .» t> >t Obverse and Reverse Obv. IVL CRISPVS NOB. CAES. Bust 1. laur. cuir. Rev. BEAT. TRANQLITAS .STR. Obv. CRISPVS NOBIL. C. Bust 1. laur. cuir. Rev. as above. PLON Obv. CRISPVS NOB. CAES. Head r. laur.. Rev. CAESARVM NOSTRORVM VOT T R / / / Obv. CRISPVS NOB. CAES. Bust r. laur. cuir. Rev. as above VOT TRS. X Obv. CRISPVS NOB. CAES. Bust 1. laur. cuir., spear and shield. Rev. VIRTVS EXERCIT. Standard with two captives. S| HE rsisc Obv. IVL. CRISPVS NOB. CAES. Bust 1. laur. cuir., spear and shield. Rev. as above. Obv. CONSTANTINVS D7N. NOB. C. Bust r. laur. cuir. Rev. BEAT TRANQLITAS T 1 P PLON As above, but N.C. Bust 1. rad. cuir. PLON Obv. as above, head r. laur. Rev. CAESARVM NOSTRORVM VOT X or VOT PTRV STR (2) As above, but bust r. laur. dr. cuir. Obv. as above, but bust 1. laur. cuir. Ref. 22 29 36 42 165 174 5 8 38 40 165 1 1 1 1 1 1 1 1 5 1 5 71 THE LULLINGSTONE ROMAN VILLA Emperor Constantius 11 as Caesar Cons tans as Caesar Constantine 11 as Augustus (probably) Constantius II as Augustus Date A.D. 330-7 325-37 333-7 337-40 337-61 Coin M3 yy i» JJ >» yy JJ JE 4 M 3 Obverse and Reverse Rev. PROVEDENTIAE CAESS. Camp gate, PTR6,STRV(2),STR. Obv. as above, bust r. laur. cuir. Rev. GLORIA EXERCITVS 1 standard ?R.P., 1 *SLG. Obv. as above, bust r. laur. cuir. Rev. GLORIA EXERCITVS. 2 standards TRS. Obv. CONSTANTI)NVS rVN. N.C. Bust r. laur. cuir. Rev. 1 standard PLON. Obv. PL. IVL. CONSTANTIVS NOB. C. Bust r. laur. cuir. Rev. GLORIA EXERCITVS. 2 standards TR-P, TR-S. Obv. legend as above, bust 1. laur. dr. cuir. Rev. PROVIDENTIAE CAESS. Camp gate. QAURL, TRS. Obv. PL. IVN. CONSTANTS NOB. C. Head r. laur. Rev. GLORIA EXERCITVS, 1 standard •TRP. Obv. CONSTANTINVS MAX. AVG. Bust r. diad. cuir. Rev. GLORIA EXERCITVS. 1 standard •TRP. Obv. CONSTANT)NVS. AVG. Bust r. laur. dr. Rev. GLORIA EXERCITVS. 2 soldiers guarding one standard. Obv. D.N. CONSTANTD7S P.P. AVG. Bust r. diad. dr. Ref. 114 122 113 104 167 250 47 3 2 1 3 2 1 1 1 5 72 THE LULLINGSTONE ROMAN VILLA Emperor Constans as Augustus Date A.D. 340-5 340-61 337-50 Coin M 2 M 3 J» JS >> (small) yy >y M 3 tt tt Obverse and Reverse Rev. EEL. TEMP. REPARATIO legionary spearing fallen horseman. CPLG. Obv as above. Rev. Emperor on galley. Obv. as above. Rev. Phoenix on rocks. ?TRP. Obv. as above. Rev. GLORIA EXERCITVS. 1 standard. Obv. CONSTANTIVS P.F. AVG. Bust r. diad. dr. Obv. FL. IVL. CONSTANTIVS P.F. AVG. Obv. CONSTANTIVS. P.F. AVG. Bust r. diad. dr. Rev. VICTORIAE D.D. AVGG. QNN. 2 victories facing. ^ TRS. Obv. D.N. CONSTANTIVS P.F. AVG. Bust r. diad. dr. Rev. SPES REIPVBLICE. Emp. st. 1. with globe and rev. spear. Obv. • -TP7S. Bust r. Rev. FEL TEMP REPARATIO. legionary type. Obv. D.N. CONSTAN (S. P.F. AVG.). Bust r. diad. dr. Rev. FEL TEMP REP. Emp. on vessel. Obv. as above. Rev. GLORIA EXERCITVS. 1 standard. ?TRP. Obv. as above. Rev. VICTORIAE D.D. AVGG. QNN. 2 victories. D , G TRS TRS Ref. 34 1 58 1 1 99-100 1 93 2 293 1 188 1 1 10 1 65 1 2 73 THE LULLINGSTONE ROMAN VILLA Emperor Constantius Gallus Magnentius (or Decentius) House of Constantine Valentinian I ' Date A.D. 351-4 350-3 364-75 Coin M 3 M 2 M 3 >t tt Obverse and Reverse Obv. D.N. CONSTANTIVS NOB. C. Head r. bare. Rev. FEL TEMP REPARATIO. Legionary as above. D | / / C ON Obv. D.N. MAGNENTIVS P.F. AVG. Head bare bust dr. Rev. GLORIA ROMANORVM. Emp. on horse spearing fallen Dacian with broken spear and shield. Obv. as above. Rev. VICTORIAE D.D. N.N. AVG. ET CAES. Barbarous. Obv. ? AHIIGVII. Head Rev. IONNVDITC (for D.D. N.N. AVG. ET CAE). 2 victories holding shield. FEL. TEMP. REP. Legionary sp. f. h. CPLG, ?ESLG. GLORIA EXERCITVS. 2 standards. GLORIA EXERCITVS 1 standard, one may be the labarum, one TRS. Two victory type. Obv. D.N. FL. VL. . . . Bust r. diad. dr. Rev. GLORIA EXERCITVS. 1 standard. Obv. D.N. VALENTINIANVS P.F. AVG. Bustr. diad. dr. Rev. GLORIA ROMANORVM. Emp. with labarum dragging captive r. OIFIIS, OIFII LVGS LVGS Ref. 52 2 21 1 68 2 1 8 1 10 2 1 12 3 74 THE LULLINGSTONE ROMAN VILLA Emperor Valens Gratian House of Valentinian Valentinian II Victor (probably) Arcadius Honorius Date A.D. 364-78 367-83 375-92 383-8 383-408 393-423 Coin M 3 yy M 4 M 3 M 4 M 3 M 4 tt tt tt tt Obverse and Reverse Obv. D.N. VALENS P.F. , AVG. Bust r. diad. dr. Rev. SECVRITAS REIPVBLICAE. Victory moving 1. SCON, ?SMAQP, ?Aq. mint Obv. D.N. VALENS P.F. AVG. Bust r. diad. dr. Rev. GLORIA ROMANORVM. SCON. Obv. D.N. GRATIANVS AVGG. AVG. Bust r. diad. dr. Rev. SECVRITAS as above. Obv. D.N. GRATIANVS AVGG. AVG. Bust r. diad. dr. Rev. GLORIA NOVI SAECVLI. Emp. st. front, with labarum shield. GLORIA ROMANORVM SECVRITAS REIPVBLICAE Obv. D.N. VALENTINIANVS P.F. AVG. Bust r. diad. dr. Rev. SALVS REIPVBLICAE. S. with trophy dragging captive £ | AQP Obv. (D.N. FL. VICTOR P. F. AVG.). Bust r. Rev. (SPES ROMANORVM). Camp gate. Obv. D.N. ARC(ADIVS P.F. AVG. Bust r. diad. dr. Rev. VICTORIA AVG) GG. Victory moving 1. As above. Obv. D.N. HO)NORI( VS. P.F. AVG. Bust r. diad. dr. Ref. 47 8 2 1 13 2 1 3 33 1 1 1 1 1 75 THE LULLINGSTONE ROMAN VILLA Emperor House of Theodosius Indeterminate Post Roman Edward I Date A.D. 1302-7 Coin M 4 J> JJ )» M 3 Minims Minimissimi AR Obverse and Reverse Rev. SALVS REIPVBLICAE. S. dragging captive 1. and carrying trophy, probably Rome mint. Rev. VICTORIA AVGGG. 2 victories facing. Rev. VICTORIA AVGG. 1. victory Rev. SALVS REIPVBLICAE. Indeterminate. Penny. Rev. CIVITAS LONDON Ref. 1 2 2 5 14 11 2 1 THE COINS—SUMMARY AND PERCENTAGES Roman Second Century Third Century, early Third Century, late Constantine I House of Constantine House of Valentinian House of Theodosius Minims and Minimissimi Total identifiable Roman indeterminate 1949 Nil Nil 6 22 29 6 3 6 71 10 Totals 1950-51 2 3 . 8 59 49 20 15 13 169 14 1949-51 21 i\ an 27681/ 18 I . 18j 240 24 Percentages 7-9 66-3 25-8 100-0 76 THE LULLINGSTONE ROMAN VILLA APPENDIX V THE MORTARIUM CONTENTS (ROOM 16) By DR. H. J. PLENDERLEITH, M.C, Ph.D., F.S.A. (Keeper, Department of Research Laboratory, British Museum) We have examined the dark grey material from the mortarium and find that it consists of lumps ranging in size from coarse powder up to about 2-5 cm. X 1-5 cm. X 0-5 cm. The colour is due to a surface coating mainly organic in nature—dust and/or earth. A freshly scraped or ignited fragment has a grey green colour and it consists of a blue pigment embedded in white matrix. The matrix was shown to be calcium and magnesium carbonates probably derived from a dolomitic limestone. The blue substance is a frit, shown by spectrograph to contain copper, silicon, calcium, tin, lead, hon, nickel, and manganese—clearly a sample of paint such as might be used in fresco painting. APPENDIX VI THE INFANT BURIAL (ROOM 8) During the cleaning down of the L-shaped wah in Room 8, a shallow hole was discovered in the surface of the northern arm of the wah. This hole had been cut through the Period I I concrete floor and the Period IV opus signinum floor above and contained certain human bones which were submitted for examination by the kindness of Professor Zeuner to Professor A. J. E. Cave, whose report is appended below. The interest of the find lay in the objects accompanying the burial. They consisted of four minims, two minimissimi, a fragment of decorated cast bronze, probably from a vessel, a sherd of undatable coarse pottery, and a few grains of carbonized wheat. One of the minims was of Fel. Temp. Rep. type, and certainly places the burial in the second half of the fourth century, though probably later than earlier in that period. And as Christian worship is now known to have been flourishing in the house at that time, this evidence in a neighbouring room of the persistence, or perhaps recrudescence, of such burial ritual is doubly interesting. It should be stated, however, that no certain evidence exists that the burial did not occur after the destruction of the house ; but as the date of this event is not yet known, the point must for the moment remain unanswered. 77 THE LULLINGSTONE ROMAN VILLA NOTE ON BONES FROM ROMAN VILLA AT LULLINGSTONE, KENT By PROFESSOR A. J. E. CAVE, M.D., D.Sc. (Medical College of St. Bartholomew's Hospital) These bones were received for report from Professor Zeuner on 5th December, 1951. They were found beneath the flooring of a room in a Roman vhla dated to the late fourth or early fifth century A.D. The material comprises: (1) Non-human remains: one complete bone and one mere fragment of animal (?ungulate) origin. (These items have been marked " A " and require identification by a professional zoologist.) (2) Human remains of, probably, a newborn chhd. The skull is represented by the frontal, parietal (right and left), occipital and right temporal bones (none perfect), the right maxilla, the incomplete sphenoid and an ethmoidal fragment. In addition there are present the shafts of three ribs and the shafts of the right humerus and ulna. These baby-bones afford no clue as to the sex or the precise age of the individual represented. The age may he anywhere between the last month of pre-natal life and the end of the first month of infancy. Most probably, however, the bones represented a neo-natal stillbirth, for they agree in anatomical particulars with corresponding bones of newborn babes with which they have been compared. The bones present show no signs of ante-mortem injury or disease. The archaeological evidence of floor-burial is characteristic of the Roman period ; such procedure seems to have been a common practice in disposing of still-bhths and, less certainly, of the bodies of very young infants. 78 THE LULLINGSTONE ROMAN VILLA I P e r i o d I la V////S/////A „ 2 BW9999WM •> 3 ©I I ® 2ZJ53 \u 1- 4-j '; j • L • * I H • » li RG. 19