A Roman Building Remenant at Thurnham- Excavations 1933

A ROMAN BUILDING REMNANT AT THURNHAM: EXCAVATIONS 1933 PAUL ASHBEE INTRODUCTION The first notice of the Thurnham Roman building was by S.C. Lampreys (S.C.L.) in his Brief hi storical and descriptive Account of MAIDSTONE and its Environs (Maidstone, 1834, 67). He wrote: 'In a field not far from Thurnham church, in a north-westerly direction, the remains of a Roman building have been recently discovered. Some curious specimens of plain and ornamented stucco, together with a few coins of some of the later emperors, and fragments of earthen vessels, were found within the square enclosed by the walls.' Later, in the Topography of Maidstone and its Environs (1839, 75) some more details were given: 'In a field not far from Thurnham Church, the remains of a Roman building have been recently discovered. Some curious specimens of plain and ornamental Stucco, together with a few coins of some of the later emperors and fragments of earthen vessels were found within the square enclosed by the walls.' A further passage in the same work (p. 122) under the heading of 'additions to Thornham, p. 75' is more informative though the direction of the site from the church is changed. 'ROMAN VILLA - the foundations discovered here were in a meadow, part of Thurnham Court, S. W. of the Church. They came to light, together with Coins, &c. which were found, 011 the ground being trenched for planting hops, and were removed to repair the roads.' A more comprehensive account was given by Thomas Charles, Doctor of Medicine, artist, antiquarian, collector and founder of the Museum at Maidstone (Roach Smith 1883, 141-6). In his abstract of 141 PAUL ASHBEE 'Roman Antiquities found at and near Maidstone in Kent' (Archaeologia, xxx (1844, 536) he observed that: 'At Thornham abouJ a mile from the foot of the hill and three miles to the east of Kits Coty House, the pavements and foundations of a considerable Roman Mansion were uncovered in 1833, It was erected in a damp clayey meadow, and was probably surrounded by woods. Of the form no account could be given, as the occupier of the farm, of which it formed a part, had broken up everything before Mr Charles obtained intelligence of the discovery. The field, however, was covered with fragments of pavement, which the labourers were breaking up when Mr Charles saw it. The principal apartments appeared to have been covered with a thin stucco, not more than two-eighths of an inch in thickness. The rooms were painted red, with borders of green and white. At one end of the building the workmen found a quantity of charcoal, some bricks like firebricks channeled, and a few flints partially vitrified. Many animal bones, fragments of the horns of deer and boar's tusks, were scaltered about; a quantity of pottery of the common kind, much broken, was also found, with a few coins. Amongst the latter there was one in large brass of the younger Faustina, three or four small brass of Constantine and a base denarius or Antoninus Pius, reverse BONVS EV ENTVS.' Thomas Charles may have collected fragments of the painted wall-plaster, some pottery and coins, and preserved them in his collection, whence they would have passed to the museum at Maidstone of which Chillington House, his home, was the nucleus. At the time of the discovery, or shortly after, bronze objects, fragments of samian ware, pieces of glass and wall-plaster were collected by Robert Rugg, a founder member of the Kent Archaeological Society. Presumably he heard about it from John Salmon Rugg (a brother?) of East Court, at Detling, Jess than a mile distant. These pieces, and some bronze brooches dug up upon the site of Detling Vicarage during August, 1831, were presented to the museum at Maidstone in about 1850. A record of this gift is in a manuscript catalogue and guide compiled by Edward Pretty (Roach Smith 1883, 146). Thus, not all the pieces from the Thumham building were necessarily gathered by Thomas Charles. R.F. Jessup (1930, 263) saw the material from Thurnham, at the museum at Maidstone, in about 1928 and considered it to date from about A.D. 100-150. R.E.M. Wheeler (later Sir Mortimer) was more explicit when he wrote his account of Romano-British Kent (Wheeler 1932) for the third volume of the Victoria County History. His account (p. 125) was as follows: '46. THORN HAM- Traces of a 'considerable mansion' were noticed here in 1833 on a 'damp clayey' site just north-west of the church, and three miles south-east of Kits Coty. The remains comprised foundations, pavements, wall-plaster painted red with borders of green and white, flue tiles, charcoal and 'vitrified' flints at one end of the building (i.e. the furnace), many potsherds, animals' bones, includi11g deer and boar, 142 A ROMAN BUILDING ATTHURNHAM and a few coins-a base silver of Pius, a 'first brass' of the younger Faustina, and some Constaminian coins. The foundations were removed and no plan made.' Sir Mortimer cited the S.C.L. (Lampreys 1834) account and that by Thomas Charles (1844) and said that a saucer of Drag. form 33, stamped SVOBNIM, probably an Antonine potter, some rude pottery and a quern were all in Maidstone Museum. Sadly, he repeated the error of direction from the church in the S.C. Lampreys paragraph in that work which was subsequently corrected (1839, 122). Prior to the 1833 unearthing and scattering of the remains of this Roman building at Thurnham, 'Roman urns and other remains of that nation' were often noticed near the hill upon which the wall-remnants of Thurnham castle stand (Harris 1719, 52, 317; Hasted 1797-1801 (V), 529). This passage and the Thomas Charles, the S.C. Lampreys and other early accounts of the Thurnham Roman building, were repeated in various forms, in topographical and guide-books throughout the nineteenth and into the first decades of this century (Coles Finch 1925, 174; Maxwell 1932, 28). From about 1931 onwards, deep road cuttings and abandoned chalk workings were scrutinised for traces of what was thought to have been a Roman cemetery. Although, Saxon graves had been encountered in 1913 (Beck 1940), close by a deep cutting, nothing was found. The use of secondary sources may well account for the incorrect positioning of the Thurnham Roman building upon a map which was with the various fragments preserved in the museum at Maidstone and the reiteration of the location as 'south of the church' certainly accounts for the site not being found until 1932. It was reasoned that, even after a century, traces of a substantial Roman building would have survived although it had been plundered to build the south wall of Thurnham church, in which tiles can be seen. Moreover, when during 1932 the site of a Roman building was located, in the northernmost of the small fields, almost entirely surrounded by woodland, on the eastern side of Honey Hills Wood (N.G.R. TQ 798572), it was thought of as a different Roman site. It was not until pieces of red wall-plaster with green and white banding were found during the 1933 excavations, these being identical with those preserved in the museum at Maidstone, that it was realised that the Roman building seen by Thomas Charles and others in 1833 had been re-discovered. The field was to the south-west of Thurnham church as had been indicated in the 1839 accounts. It was always assumed that, because the two small fields were almost entirely surrounded by woodland, the Roman building had been discovered and broken up when woodland had been cleared and 143 PAULASHBEE the fields created. Subsequently, it has emerged that this may not have been the case. A 1709 map of Thurnham Manor, by John Watts of Boxley (Hull 1973, 16) depicts Honey HilJs Wood and Field with lineaments which are substantially those of 1933. The two small fields were a single larger field, Honey Hills. Indeed, with other adjacent fields, Honey Hills was a part of the park of Corbies Hall (Appendix). The remains of this manor house, located in 1932, were destroyed by totalitarian agriculture during the early 1950s. It is known from one account that, in 1833, a period of agricultural innovation (Burnham and McRae 1978, 20), the remains of the Roman building were an impediment to hop cultivation. Living memory in 1933 claimed the Honey Hill field(s) as hop-gardens until just after the 1914-18 war. The point of a pole was found during the excavation. A note of the 1933 excavations was put into Archaeologia Cantiana (Cook 1934, 196), descriptive articles in the local press (South Eastern Gazette, 14 October, 1933; Kentish Express, 20 October, 1933) and popular features in the national press (Sunday Express, 22 October, 1933; Daily Sketch, 24 October, 1933). Apart from these, no other account of the work has appeared. Subsequent fieldwork and excavation in the vicinity has allowed something of the site and its setting to be seen and has extended the range of Roman buildings that stood in what became the Honey Hills field. In his consideration of Roman roads, I.D. Margary (1946, 3; 1948, 212) suggested that a branch road left the well-known Rochester-Maidstone-Wealden route (Jessup 1930, 175), in a southeasterly direction where it was marked by an alignment of lanes and tracks, past Street Farm, Boxley, and Boxley, Detling and Thumham churches. Overt traces, now largely obliterated, ceased at Thurnham church. However, fieldwork in 1949, by the present writer, eastwards of Aldington Court Farm (N.G.R. TQ 812754), disclosed banks, tracks and lanes which would suggest that, if some modern road be included, this branch road would extend as far as Tanyard Farm (N.G.R. TQ 905520), a short distance beyond Lenham. In 1958, excavations were undertaken, just to the south-east of the site investigated in 1933, for the Maidstone by-pass (A20(M)) was to traverse the Honey Hills field (Pirie 1960). The much reduced remains of two structures, not dissimilar to the foundations found in 1933, were uncovered. One, with apses, drains, opus signinum floors and white wall-plaster, about 80 ft. from the 1933 site might have been a detached kitchen, the other, some 200 ft. away, was thought to have been an outbuilding. At the time of writing the sites of the buildings found in 1958 are beneath the earthworks and carriageway of the trunk-road. Those of 144 A ROMAN BUILDING ATTHURNHAM the 1933 investigation are at its margin, but any remaining foundations may have been scattered by deep ploughing. TOPOGRAPHY AND GEOLOGY Thomas Charles termed the site a 'damp clayey meadow', which description was as apt in 1833 as in 1933, because it was upon the Gault, the narrow corridor of dark, bluish-grey clay between the Chalk and the Lower Greensand (Burnham and McRae 1978, 33, Fig. 22). Before deep boring took water from the chalk, its winter water-table rose and these claylands became saturated. Indeed, the meadow adjacent and to the south-east of Honey Hills (Hull 1973, 16), traversed by two small streams, was frequently flooded. A number of farms (e.g. Parsonage Farm, Thurnham, N.G.R. TQ 802572) and other buildings are directly upon the Gault, but no major settlements, for its considerable expansion and contraction, upon wetting and drying, disrupts foundations. Because of its characteristics a reason, not immediately clear, must exist for the siting of what may have been a substantial building upon such a subsoil. Gault clay produces pale yellow bricks: the colour of the mass of roofing-tiles encountered during the excavation (Fig. 1) was this pale yellow. Whereas nothing can be inferred from the building remains encountered in 1933 and 1958, an installation central to a series of clay workings cannot be entirely eliminated. Coppiced woodlands are a traditional feature of the Gault claylands and, appropriately managed, there would have been no shortage of fuel (Rackham 1976, 51). In 1933, Honey Hills, the field in which the patent traces of a Roman building had been found a year previously, was near derelict, rough, rabbit-ridden, pasture, victim of the agricultural depression. The Honey Hills of the 1709 map (Hull 1973, 16) had been divided into two smaller, almost equal, portions (O.S. Maps, Kent, Six inches to One Statute Mile, Sheets XLII, N.E., XLIII, N.W.), having been hop-gardens. In 1940, the fields were ploughed and throughout the war, and subsequently, they carried cereal crops. THE 1933 EXCAVATIONS Excavations on the site of the Roman building that had been located in the Honey Hills field began on 21st September and continued until 19th November. Permission to excavate was given by Sir George Hampson of Thurnham Court and Mr Charles Brown at Parsonage 145 PAULASHBEE Farm. For the most part the work was carried out by the present writer (then living in Bearsted), aided by F.W. Walkling. Incidental assistance was given by friends and neighbours. Much encouragement and good advice upon archaeological matters in general, and Romano-British studies in particular, besides excavation organisation, was given by Norman Cook, then Sub-Curator and Keeper of Archaeology at Maidstone Museum. He was also instrumental in arranging an exhibition of material from the building, to the Maidstone Natural History Society at their meeting held in the Museum on Monday, 16th October. Their President, Mr J.W. Bridge, at this meeting, encouraged the Council to vote five guineas for the excavation, which they kindly did. In the event, two workmen (Mr James Earl, from Ware Street, and Mr Ernest Ragget from Bearsted) were employed from 23rd October to 5th November. The initial cuttings had located a wall and an opus signinum floor, their work disclosed further foundations. Photographs were taken during the progress of the work by C.E. Fisher, then Keeper of Natural History at Maidstone Museum, and general assistance was given by Barbara Laidler, who later joined the Morven Institute of Archaeological Research at A vebury. Areas of optimum concentration of pieces of broken pottery and building materials in rabbit scrapes and chucks determined the siting of the initial cutting. This, about 7 ft. by 5 ft., exposed an indeterminate mass of clay, flint nodules and building materials. A second such cutting, to south-eastwards and separated from the first by a baulk about 1 ft. 6 in. in width, revealed the northernmost of the opus signinum floor remnants. Thereafter, it was extended to isolate the floors while the initial cutting was enlarged to expose walling. The workmen followed and exposed the rest of the walls. Because both floor and walls were nowhere more than about 1 ft. 6 in. below the grass-grown modern humus, sections were considered uninformative and thus unnec􀁢ssary. THE ROMAN BUILDING REMNANT The building remnants are best considered beneath two headings: (1) those disturbed, presumably in 1833, and (2) the undisturbed opus signinum floor remainders. (1) The disturbed Building Remains (Fig. 1) An extension southwards of the cutting which had disclosed the undisturbed opus signinum floor remnants uncovered, first of all, a 146 A ROMAN BUILDING AT THURNHAM PLATE I (Photo.: C.£. Fisher) Eastern Wall Junctions at the south End of the excavated Area. 147 PAUL ASHBEE roughly rectilinear mass of broken yellow roofing-tiles and, secondly, the bottom two or three courses of a wall. The broken yellow tiles, such as would have been made from the Gault clay of the immediate locality, were in a uniform layer only 5 in. in depth and were 1 ft. 4 in. below the modern surface. Almost all were pieces of roofing-tiles, with prominent side flanges, but there were smaller pieces of the half-round tiles, used for sealing the butt-joint. One or two pieces were large (an almost perfect example was deposited in the Museum at Maidstone) and almost a quarter of a ton was removed. Beneath this layer was compact clay, the undisturbed Gault. A westerly extension of the cutting which had located the wall lower courses, disclosed its juncture with another wall (Plate I), of which only about two or three courses remained. It bounded the undisturbed floor remnants but was separated from them by a mass of clay, flint nodules, pieces of ragstone, of broken red floor and of flue-tiles, lumps of yellow mortar and scraps of wall-plaster. This NE-SW wall, of which some 80 ft. were uncovered, was disclosed, as were the other walls, by following them. However undesirable this nμy have been, only jumbled, redeposited, material, a layer, which cloaked the wall remnants and extended downwards from beneath the modern humus to the undisturbed gault clay, was removed. Thus, a part of the plan of the building was recovered (Fig. 1). Hyphenated lines indicate walls which were located by probing with a substantial, especially forged, iron rod. The wall footings, which rested upon the Gault clay, were two or three, but rarely four, courses of medium or large nodules of flint and pieces of ragstone, set in a yellow, hard, sandy mortar. The nearest sources of suitable ragstone are the outcrops on the western side of the steep stream valley just to the south-east of Bearsted parish church (N.G.R. TQ 802554) or in similar circumstances near Weavering Street (N.G.R. TQ 791559). All these walls rested upon the Gault clay and, as far as could be seen from their meagre remains, each was keyed one to another, suggesting a single, uniform, building operation. At the south-western corner of the excavated area (Fig. 1) were the broken remains of an opus signinum floor, smooth white cement with, adhering to its underside, small fragments of red tile and brick similar to the undisturbed floor remnants. A piece of white plaster remained upon the wall which bounded this broken pavement on its northern side while a sherd of coarse pottery lay on the clay beneath the broken-up flooring. The uniform height of these remains of walls, or, perhaps, footings, and the jumbled material, clearly derived from the destruc- 148 A ROMAN BUILDING ATTHURNHAM PLATE II (Photo.: C. E. Fisher) The undisturbed Opus Signinum floors from the South-east. lion of such walls, is, presumably, direct evidence of the 'breaking up' witnessed by Thomas Charles, S.C. Lampreys, Roger! Rugg and, probably, others. Similar demolition, in an earlier age, the source of the tiles and other pieces incorporated into the southern wall of the nave of Thurnham parish church (N.G.R. TO 804576) may have been neither so thorough nor so extensive. (2) The undisturbed Floor Remainders (Pl. II; Fig. I) Presumably, before the building was broken up, these opus signinum floors extended to the wall which bounded them on their western side. This is suggested by the yellow mortar, upon which the smoothed, white, tile-fragment-strengthened cement was bedded, found on the surface of the clay between floors and wall. The two rectangular floor remnants, a northern and a southern. were separated by a square-sectioned, humus-filled, channel in which was found a samian ware base (Drag. 27), some small pieces of coarse pottery and an iron nail. Clearance, in a northerly direction, of the northernmost of the two floor remnants disclosed another channel which separated them from a third, similar, floor remnant, only a part 149 THVR,NHAM 1 933 NG R TQ ]9'155]2 PAUL ASHBEE .......... _ ........... .. -':' '' r ·-.....-.. ...... ,' , ,' ' ' ' ,: ': ': '' ' ' : / '.: ', ' ,' '' ,' '' '' '' '' '' ' : I , '' '' I , ', , l. IM IT .5 : ................... ....... .. .. .. .... .............. blSTU11􀁹ANCE 8ETWtEN rouN!>ATIO>H 4 rl 0 0 "-.S t/•fLINT'i o􀁵us 􀀉tGtJINUM ...... .. li:A G ST ON£ WAlL IMG- r L 00 I!. R.EMNANTS 􀁺 110 K. E >J R.00flWG T I l l S. 0 s> U \ 􀀃 I CdJ t tJ UM fl00111NG --􀀝---------------------4oF 􀀞􀀟--􀀠--------------- .zoM Fig. 1. Plan of the Remains revealed by the 1933 Excavations. of which was excavated. This was also square-sectioned, slightly wider than the first, its bottom being the close-packed flint nodules upon which the floors were bedded. At its eastern end this channel had been partially damaged and, because of the break-up of the north-east corner of the northernmost floor remnant, seemingly widened. It was also filled with humus and its excavation produced a samian ware rim (Drag. 27). Above these opus signinum floors, and below the modern humus, was a sandy, mortar-laden, soil which contained, besides broken flint nodules, quantities of broken pieces of wall-plaster, mostly lying in 150 A ROMAN BUILDING AT THURNHAM the soil painted face downwards. At the northernmost limit of the cutting that exposed the floor remnants there was a localised lens of wall-plaster pieces, some 5 in. in depth. Most of these pieces found in the soil were red, many bearing green and white lines about 3/s in. in breadth, and identical with the pieces preserved in the museum at Maidstone. On these opus signinum floor remnants there were areas of white wall- or ceiling-plaster lying face downwards. Although friable and soft, pieces of some size could be peeled off with a trowel. Indeed, an endeavour (on 10th October) was made to lift about a square foot by backing a cleaned area with fabric and pouring hot glue on to it. Beneath this skin of plaster there were found sherds of samian ware (Drag. 18/31; 27; 30; 36; 46), the head of a bronze statuette (or anthropomorphic handle ?) (Plate III), ? silver button, fragments of glass vessels and small pieces of coarse pottery. A bronze nail, found on the tip, also may well have come from these floor remnant surfaces. As far as could be seen from their edges, these opus signinum floor remnants rested upon the raft of flint nodules that became visible when the dividing channels were emptied of their humus infill. At the time of their excavation it was thought that these channels might have been drains, indeed, a sump was sought in the area at their eastern ends. Their sharp edges, noted even in 1933, and even infill suggest that they might have been the channels left when sleeper timbers for the support of partitions had rotted or were removed. THE ARTIFACfS AND ARCHITECTURAL FRAGMENTS Only a small number of objects were found in association with the opus signinum floors and thus stratified. The remainder were from the ubiquitous jumbled mass of broken-up building materials, mingled with lumps of clay, beneath the grass-clad humus which mantled floors and wall-footings. Thus, those found with the undisturbed floor remains are listed separately from those from disturbance. (1) FROM THE UNDISTURBED OPUS SIGNINUM FLOORS (PI.II; Fig. 1) Samian ware Rim sherd, Drag. 27, from the southern channel; sherd from a straight-sided dish, Drag. 18/31, from the surface of the southern remnant; base fragment, Drag. 27, from the northern channel; rim 151 PAUL ASHBEE PLATE 111 () (Photo.: J. 8/omfie/d) The small Bronze Human Head Representation. fragment Drag. 27, decorated sherd Drag. 30, rim fragment Drag. 36, and the side of a small bowl, Drag. 46, all from the surface of the northern remnant. Coin Small bronze, possibly of Gallienus, from the surface of the northern remnant. Metal objects From the northern floor remnant: small bronze representation of a human head (Plate Ill); a disc, ? silvered, with a shank at its centre 152 A ROMAN BUILDING AT THURNHAM on one side, 7/s in. diameter, ? a button; a bronze nail, with a disc-head, found on the tip but probably from the surface of one of the floors. Glass Fragments of a light blue cup with a rolled rim; fragments of a small bottle; fragments of a flat, olive-tinged, dish; two pieces of thick, blue-tinged, near-opaque, window-glass; a lump of fused glass; all these pieces from the surface of the northern remnant. Coarse pottery Handle of a ? flagon, reddish texture; body sherd of fine, darkgrey, texture; sixteen coarse, dark-faced, body sherds. The first two pieces are from the surface of the northern remnant, two more were from the infill of the southern channel and the remainder from the surface of the southern remnant. (2) FROM THE UBIQUITOUS SUB-HUMUS DISTURBED LA YER Samian ware Rim fragment, Drag. 18/31; base fragment, Drag. 27, with graffito T on the exterior; bowl fragment, Drag. 36. Coins Bronze CONSTANTIVS; bronze IMP C ALLECTVS. Metal objects Bronze strap-end; small iron knife-blade; small piece of iron, a ? handle; six forged iron nails. Glass Piece of the strap-handle of a large jar; a small (¼ in. long) cylindrical opaque blue glass bead. Coarse pottery Fifty-one sherds of coarse pottery included a sherd from a widemouthed jar with a grey slip, decorated with two parallel girth lines, a substantial mortarium rim-sherd of hammer-head type and five base sherds. The remainder were mostly rim-sherds of reddish and dark 153 PAULASHBEE textures. Body sherds were collected but not counted. There was a single Upchurch ware body sherd. Stone Fragment of a hone of inverted U-section. ? Ragstone. Wall-plaster The greatest quantity of coloured waU-plaster fragments was found in a consolidated lens at the northern end of the floors area. This included numerous dark-red pieces with grey and white banding, some green pieces, some yellow and some pink pieces. One particular piece showed a spray of leaves, green on a cream background. The red colour is best described as a deep scarlet upon which had been painted the white bands. These white bands had sometimes been splashed with green. Even green bands were scarce. The yellow used had a deep golden hue and some pieces displayed white bands which had been dusted with grit while others showed that black had been used with yellow. The pink plaster had a light white, limey, character. The few pieces of white wall-plaster were thick and heavy. Tiles In addition to the compact deposit of yellow roofing-tiles, of a kind that could have been manufactured from the Gault clay of the immediate surround, numerous small fragments of substantial red floor-tiles and combed box flue-tiles were found. All the pieces of red tiles were small and none of them, in terms of size, approached the magnitude of some of the pieces of yellow flanged roofing-tiles. ENVOI Archaeological retrospection apart, to write an account of one's adolescent attempt at excavation, undertaken more than a halfcentury ago, is a task that falls to few. In so doing there must be guard against the injection of qualities that the enterprise never had. As a technical performance it was probably no worse, and, perhaps, rather better, than many of that time. The turf was removed, and stacked, from approximately rectangular areas which were then dug into (Piggott 1983, 30). Notions of stratigraphy and relationship were from J.P. Bushe-Fox's Richborough reports (1926; 1928), available from the Kent County Library, while R.A. Smith's British Museum Roman Britain Guide (1922) and R.G. Collingwood's Roman Britain 154 A ROMAN BUILDING ATTHURNHAM (1930) contained comparative details. The rationale was a conviction that it was necessary to see what was beneath the surface traces together with the sudden availability of assistance. Little can be said regarding the foundations and floor remnants other than recognition of the possibility that they were a part of a building planned upon the same lines as, for example, the villas at Boxted, Farningham or Faversham (Blagg 1982, 57, Fig. 26). A further factor is the nearby branch road (Margary 1946, 33; 1948, 212). Such a by-way could have served various installations along its line. Further fieldwork may eventually place them. The locating of the Thurnham Roman building remnant resulted from what would, today, be termed field-walking. Thurnham, unlike Bearsted, had many fields which were regularly ploughed. Also, in the 1930s, rabbits were endemic and their multitudinous chucks often allowed a field to be explored as if it were tilled. The information regarding extant (Appendix) and destroyed sites was plotted upon the Six inches to One Statute Mile Ordnance Survey Maps maintained in the museum at Maidstone and is now in the National Monuments Record (Royal Commission on Historical Monuments for England) at Southampton. Today, it is thought that many parish boundaries may have been determined in early, even prehistoric, times (Taylor 1983, 104--5, 181). Thus the time spent, long ago, upon rural reconnaissance was not entirely time wasted. APPENDIX Corbies Hall (Old Thurnham Court) in 1933 Early in 1933 the existence of this ruined building was indicated, as worthy of investigation, by Mr Charles Brown of Parsonage Farm. Situated in a wood about 300 yds. SW of the farm (N.G.R. TQ 801571), it was shown upon the Ordnance Survey, Six Inches to One Statute Mile sheet (Kent, Sheet XLIII, NW) as 'Ancient Ruin'. Mr Brown had always known the building remains as 'Corbett's Hole', a debasement of the earlier name (Hull 1973, 16, the John Watts Thurnham Manor Map of 1709) which was preserved in the name of adjacent woodland, Corbier Hall Wood. The site was scrutinised on 7th April, 1933. Although forgotten by 1933, there had been an excavation of the building, some seventy years earlier and a brief account had been published by Charles Wykeham Martin (1862) of Leeds Castle (Roach Smith 1886, 3-10). This was as follows: 155 PAUL ASHBEE 'Sir George Hampson has lately been making some excavations on his estate in the parish of Thornham, Kent, and he has laid bare what appears to have been the lower story of the hall of a mansion called Corbie's Hall from the name of its founder. lt appears by Hasted to have been built in the reign of Richard JI. It must have been a building of some importance, as the dimensions of the cellar under the hall are 66 feet by 27 feet. A wall runs down the cell/re to carry the floor-joists. The room above, as the wall would have a set off at the level of thefioor of quite six inches, must have been 67 feet by 28 feet. ft must have been warmed by a firep)ace in the wall, with a chimney, as the floor was evidently of wood. The whole was surrounded by a moat, of which considerable traces are left on one side, facing about north-east. A small building was discovered and removed last year which was probably a dovecot. 1 did not see it, and it is now removed; but it was inspected by Mr Pretty, Under-Secretary of the Kentish Archaeological Society. I understand it was a circle of about 24 feet in diameter, wiJh a solid mass about 8 feet in diameter in the centre. Only the foundations remained, showing the position of two small doorways. There appear to be further traces of foundatwns, but they are covered over with underwood and cannot easily be uncovered. There can be no doubt as to the name of the building, as an old map of the estate, about 1640 (I understand from Sir George), shows on the spoJ in question the words "Here stood Corbie's Hall" In 1933, Corbie's Hall appeared to have been a rectangular building about 60 ft. by 30 ft., built upon a low rectangular mound, about which were traces of a moat. The walls and foundations had in them ragstone, chalk and flint nodules bonded by sandy mortar. In 1933, it was being quarried for stone. Pieces of square red tiles, with rounded corners, about V2 in. thick with square peg-holes were scattered over the site and its immediate area. Ash tree-stumps were upon the walls which, in one place, had been exposed to a depth of about 5 ft. The principal north wall was at least 3 ft. thick and it rested against a bank of rammed chalk, 1 ft. in thickness and 3 ft. in depth. Among the debris of the building there was an abundance of oyster and cardium shells and a sherd of gritted ware, square-rimmed with pink core and whitish slip. The heaps of soil, discarded by the search for flint and stone, yielded a quantity of dark-faced gritted, pottery, characterised by its square rims. By 1957 the site, and much of the surrounding woodland, had been razed, in the interests of cereal agriculture, and no more than a scatter of stones and flint nodules marked it. REFERENCES Anon. 1839 Beck 1940 Topography of Maidstone and its Environs. R.T. Beck, 'Unrecorded Saxon 156 A ROMAN BUILDING ATTHURNHAM Cemetery at Thumham, Kent,' Antiq. Journ., xx (1940), 380-2. Blagg 1982 T.F.C. Blagg, 'Roman Kent' in (Ed.) P.E. Leach, Archaeology in Kent to AD 1500, CBA Research Report no. 48, 1982, 51-60. Burnham and McRae 1978 P. Burnham and S. McRae, Kent the Garden of England, 1978. Bushe-Fox 1926 J.P. Bushe-Fox, First Report on the Excavation of the Roman Fort at Richborough, Kent, Oxford 1926. Bushe-Fox 1928 J.P. Bushe-Fox, Second Report on the Excavation of the Roman Fort at Richborough, Kent, Oxford 1928. Charles 1844 T. Charles, 'Roman Antiquities found at and near Maidstone in Kent', Archaeologia, xxx (1844), 535-8. Cook 1934 N.C. Cook, 'Archaeology in Kent, 1933', Arch. Cant., xlvi (1934), 195-7. Coles Finch 1925 W. Coles Finch, In Kentish Pilgrim Land, 1925. Collingwood 1930 R.G. Collingwood, The Archaeology of Roman Britain, London 1930. Harris 1719 J. Harris, The History of Kent, i, 1719. Hasted 1797-180i E. Hasted, The History and topographical Survey of the County of Kent, 1797-1801. Hull 1973 F. Hull, Kentish Maps and Mapmakers, 1973. Jessup 1930 R.F. Jessup, The Archaeology of Kent, 1930. Margary 1946 I.D. Margary, 'Roman Roads in west Kent', Arch. Cant., lix (1946), 29-63. Margary 1948 l.D. Margary, Roman Ways in the Weald, London 1948. Maxwell 1932 D. Maxwell, The Pilgrims' Way in Kent, 1932. Piggott 1983 S. Piggott, 'Archaeological Retrospect 5', Antiquity, !vii (1983), 28-37. Pirie 1960 E. Pirie, 'Thurnham Roman Villa', Arch. Cant., lxxiv (1960), 162-70. Rackham 1976 0. Rackham, Trees and Woodland in the British Landscape, 1976. 157 Roach Smith 1883 Roach Smith 1886 (S.C.L.) 1834 Taylor 1983 Wheeler 1932 Wykeham Martin 1862 PAUL ASHBEE C. Roach Smith, Retrospections Social and Archaeological, i, 1883. C. Roach Smith, Retrospections Social and Archaeological, ii, 1886. S.C. Lampreys, A Brief historical and descriptive Account of Maidstone and its Environs, 1834. C. Taylor, Village and Farmstead, 1983. (Sir) R.E.M. Wheeler, 'RomanoBritish Kent', VCH (Kent), iii, 1932. C. Wykeham Martin, 'On some Discoveries at Thornham (Kent)', PSAL, Second Series, ii (1862), 103-4. 158