Canterbury: The post-war Excavations

CANTERBURY: THE POST-WAR EXCAVATIONS S.S. FRERE The bombing of London and Canterbury together with other cities in the early years of World War II opened a new era in British archaeology. In those distant days it is hard to remember constraints such as lack of funding, need to develop new techniques, absence of large-scale organisation (especially for post-excavation work), and" the generally amateur character of all involved. Medieval dirtarchaeology itself hardly yet existed. Roman archaeology, of course, was established as a respectable discipline; but even so very little was known of the lay-out and still less of the history of Romano-British towns. It was therefore primarily the hope of learning something of Roman Canterbury which inspired the formation of the Canterbury Excavation Committee early in 1944. The Roman archaeologist was more fortunate than his medievalist colleague in that, although some Roman cities lay beneath modern ones where they were hardly accessible, there existed a number of others which the chances of later history had left as empty open sites. It had been on these open sites that attention had naturally focussed when at the end of the nineteenth century a need was felt to learn more about Romano-British town life, and the fashion for large-scale excavations developed. Long before the advent of aerial photography the parch-marks of the Silchester streets had been observed from the convenient grandstand of the city's defences and the skeleton of its town-plan appreciated.' Appropriately enough, therefore, it was at Silchester that the first large-scale campaign of urban archaeology began with the Society of Antiquaries' nineteen-year excavation (1890-1908), followed soon afterwards by a ten-year campaign (1901-10) at Caerwent and then by a three-year excavation at Wroxeter, cut short by the Great War. These excavations taught us much about Romano-British town- 1 G.C. Boon, Silchester, The Roman Town of Calleva (Newton Abbot, 1974), 22-4. 29 S.S. FRERE planning, types of urban housing, temples and drains; but it would hardly be unfair to describe them all as mono-period excavations;2 for the discipline of stratification only slowly took root, and little could be learnt of Romano-British urban history. The excavations served to illustrate a single phenomenon known simply as 'Roman Britain' despite its life of 350 years. In F.J. Haverfield's chapters on towns in the 1915 edition of his Romanization of Roman Britain the emphasis was perforce on the application of Roman institutions to a barbarian province, and on a timeless entity, the character of the Romano-British town. By 1924, when G. Macdonald revised and published Haverfield's Roman Occupation of Britain, there are a few tentative dates suggested for the first growth of some of the cities, although these were based on the totality of the finds rather than on stratified evidence. Even as late as 1930, when the turn of the last 'greenfield' site, Verulamium, came up for large-scale excavation, the generalisation is still broadly true that Roman Britain possessed a military history which could be reconstructed in some detail, but its urban history had hardly yet reached infancy. R.G. Collingwood, in the edition of his Roman Britain issued in 1932, was able to tell us that the earliest Roman London, before A.D. 60, was largely a city built of timber and clay; while the recent discovery of the Wroxeter forum inscription enabled him to make some short generalisations on the later development of Viroconium. This first phase of urban work culminated in the Verulamium excavations of 1930-34, directed by an imaginative genius, Mortimer Wheeler, who was an archaeologist of an entirely new kind. Instead of assembling 100 workmen under a foreman and himself visiting once a week to collect the finds, he collected an experienced group of archaeologists - names such as P.K. Baillie Reynolds and K.M. Kenyon among them - to supervise particular parts of the work and himself ranged the countryside from Wheathampstead to Prae Wood mounted on a white horse and framing the larger strategy, while his distinguished wife, Tessa, saw to the details of the stratification. In 1936 the result was the publication, for the first time in Roman Britain, of an analysis of the genesis and history of a city in a report which gave chapter and verse for the vertical stratification as well as for the horizontal plan. It was perhaps unfortunate, although inevitable, that the historical picture which emerged was so greatly influenced by Rostovtzeff's recently published and highly individual Social and Economic History of the Roman Empire with its exagger- 2 Cf. M.G. Fulford in Britannia, xv (1984), 357-61. 30 CANTERBURY POST-WAR EXCAVATIONS ated view of urban ups and downs; and by today's standards the work is short on detail. Nevertheless, the Verulamium Report was a landmark in our studies; it gave later excavators a wider goal of vision, greatly-improved technical methods and a theory of urban history against which to test their own results. But then came the bombings of World War II. Archaeologists had to shift direction from the leisured exploration of open sites to the urgent problems of what is now called Rescue Archaeology. However, modern Rescue Archaeology is a very different thing from what we knew in the early post-war years. It grew slowly. Canterbury led the way from April 1944, followed soon after, in 1945, by Exeter, Southwark and Dover, and by London from 1947.1 narrowly avoided arrest as a suspected enemy agent in Southwark one Saturday in 1944 when I was mapping bombed sites in preparation for a scheme of excavation: spectators thought I was noting down targets for further V-2 rocket attacks. That any steps were taken at all is surprising: 1944 was well before the end of the war, money was short, and there was a widespread feeling that it was somehow disloyal to contemplate active archaeology when there was a war to be won. That steps were indeed taken was through the initiative of two great men. One was Bryan O'Neil, Chief Inspector of Ancient Monuments, a leader of much vision, who extended to Canterbury the recently felt responsibility of the Ministry of Works for sites under threat of destruction, a responsibility which had already resulted in rescue excavations on the sites of new wartime airfields. The other great figure was Frank Tomlinson, a retired Major of the Buffs, who had an intimate knowledge of Canterbury and was influential with its councillors and municipal officers; he took the post of Hon. Secretary to the Committee and saw to all the local arrangements. Audrey Williams, later to become Mrs. W.F. Grimes, was seconded by the Ministry of Works from rural rescue excavations to take charge; and when she became Curator of the Verulamium Museum in 1945 and was unable to continue, I was invited by O'Neil to succeed her. It is hard to picture the conditions of those days. The only injection of government funds was the payment of £2 per week as subsistence to the Director while on the site. All the labour was provided by volunteers who were paid nothing, and who, indeed, were glad to finance their own lodging or camping arrangements in return for experience gained. A camp-site was made available at St. Augustine's College. We were fortunate to be the first in the field, for there existed a large untapped reserve of amateur interest, eager to take part in excavation. The Canterbury dig was both larger and more continuous than comparable digs at Exeter or Southwark, and was 31 S.S. FRERE H H AREA EXCAVATED # AREA NOT AVAILABLE tf ^ % (D SCALE 500 FEET Fig. 1. Plan of the South-east Part of Canterbury, showing the stage of excavations reached in 1949. infinitely more salubrious than the latter. I remember a site in Southwark where each morning someone dropped in a packet of ordure, neatly wrapped in a paper bag. The more attractive environment at Canterbury soon built up a workforce of mostly very skilled volunteers who returned again and again. Many of the well-known names in later British archaeology began their excavating careers at Canterbury. With this financial and organisational background, digging was bound to be small-scale. I was then a schoolmaster, and digging was tailored to the school holidays, a programme which also suited the volunteers. We dug for 6-7 weeks in the summer, 2-3 weeks each Easter, and - in two successive years - for two weeks even at Christmas. But this last experiment was not a success; the first year the ground was frozen so hard as to be difficult to dig, and in the second rain converted the dumps into dangerous avalanches of mud. To begin with, the policy of the Committee was confined by two controls. One was the knowledge that reconstruction would take place as soon as possible along existing frontages: it was to these frontages that attention had first to be directed. As it happened, most 32 GENERAL PLAN OF ROMAN BUILDINGS (MODERN STREETS / ' / OUTLINED) y/ w \ \ * \ + POTTERY KILN Key to Roman Canterbury as excavated. 1. Late 2nd Century Buildings (part of). 2. 1st Century Road Metalling and Wall Founda- \ tion. 3. Late 1st Century Courtyard Building in Butchery Lane. 4. Concrete Foundations, Wall Footings and Well, 2nd and 3rd Centuries. 5. Conduit of tile and stone running eastwards. 6. Bath-house, 2nd to 4th Centuries. 7. Wall Footings, Gravel Road and Timber Buildings. 8. 2nd Century Buildings in Rose Hotel Yard. 9. Massive remains which may have been Town Baths. > z H m 50 cd c 50 *: TJ o on H k > tfl X O > > H O z Fig. 2. General Plan of Roman Buildings in the Area of Fig. 1. S.S. FRERE PLATE I St. George's Street Baths. of them had cellars, into which the buildings had collapsed. Fortunately, the possibility of unexploded bombs lurking in the rubble caused the city authorities to empty the cellars, and it was thus to the cellars that we turned. They had the advantage that they had been dug through the upper levels of deposit, and might lead us straight to Roman levels underneath. This situation also suited the volunteers, who could not easily be expected to dig trenches 3-4 m. deep from the surface or to complete such a task in the time available. The second constraint was that in any case lack of resources prevented the hire of men or equipment to tackle sites from the surface. The cellars, mainly of eighteenth- or nineteenth-century date, had already removed the possibility of learning much about the medieval frontages. By 1949 the situation shown in Fig. 1 had been reached. The areas excavated were virtually all in basements along the main streets. The results (Fig. 2) were perhaps more fruitful than such a policy deserved: the Roman buildings discovered included the St. Margaret's Street Baths, the Rose Lane House, the Butchery Lane House, the St. George's Street Baths and the enigmatic building in Burgate Street. Of these, only the St. George's Street Baths was still standing above cellar level (Plate I). Its good state of preservation 34 CANTERBURY POST-WAR EXCAVATIONS PLATE II • > Butchery Lane: View of Cellar with Mosaic. yielded interesting evidence on the control of differential heat in the various rooms.1 However, virtually nothing had yet been learnt of the Roman street-grid, and above all there were huge areas between the modern streets were no digging had been done. One problem with the newly-exposed cellars was that there existed no plans of them; and they were difficult to plan accurately because of their irregular shapes; walls were rarely straight, and some had collapsed or were standing to varying heights. It was hard to find fixed t points from which to measure. I had had no experience of large-scale surveying, but fortunately we acquired the devoted services of a retired architect, A. Martyr Smith, who enjoyed making outline plans for us unassisted; this was a great help in the early days, even if the plans were not always quite accurate at the level of the cellar floor. Later the City itself undertook planning in preparation for redevelopment, and its officers, Mr. L.H. (later Sir Hugh) Wilson, the City Architect, and Mr. H.M. Enderby, the City Engineer, both gave generous assistance to the work of the Committee, whose links with the City were further strengthened when, after the retirement of Major Tomlinson, John Boyle, the Town Clerk, became its Hon. 3 The Archaeology of Canterbury, vii (Maidstone, 1983), 38-9. 35 S.S. FRERE PLATE III ?f:*r i s • \ ~*» Fountain Hotel: View of Cellar. Secretary. Another bonus of the time was that the 1874 Ordnance Survey 1:500 plans of Canterbury were still almost all in print. Thus the new turn which British excavators were forced to take in bombed cities was largely towards a vertical rather than a horizontal archaeology: sections gave valuable clues to development when integrated plans could not be obtained - a complete contrast to the old 'greenfield' type of exploration. The exigencies of the sites available usually prevented recovery of complete structures, but even a small isolated site might yield information of importance. At no. 3, Palace Street, for instance, one such cellar revealed a short length of substantial though robbed Roman wall and some street-metalling, while the street overlay pre-Roman Belgic features of interest.4 In St. Margaret's Street near its junction with Watling Street a single isolated bombed site explored in 1950-51 revealed a bewildering series of massive Roman walls beneath some shallow eighteenthcentury disturbances; here it was the discovery that the diggers of cellars beneath undestroyed properties on either side had found the Roman foundations too massive to remove which led to the iden- 4 The Archaeology of Canterbury, viii (forthcoming). 36 CANTERBURY POST-WAR EXCAVATIONS tification of the Roman theatre. Almost all the owners of these properties showed an interest in the work and gladly gave premission for examination of their cellars. One lady refused, but here patient waiting produced its reward; some years later, in 1956, the house was taken over by the Abbey National Building Society which began to lower the cellar floor; a massive wall of the theatre, 3 m. wide, was revealed as a major impediment to the work.5 In Butchery Lane three small mosaics in the northern wing of a private house survived only just below a cellar floor (Plate II), although the main range of this building remains inaccessible below the Lane and surviving houses on its west side. Similarly at the former Fountain Hotel in St. Margaret's Street part of the Roman Public Baths was found below the wine-vault (Plate III); the area adjacent was one of the few large car-parks in Canterbury, and it was only with the greatest difficulty that in 1949 permission was obtained to excavate a small part of it in an attempt to learn more of the plan. Unfortunately, this excavation turned out to be in the wrong place for the purpose; only when the car-park was totally excavated by the Trust in recent years could the structures encountered be fully explained.6 However, in the early years all knowledge was new, and all valuable where nothing was known before. If the results were often vertical rather than horizontal, these could nevertheless throw light on questions of continuity and on the character of settlement in successive periods. This point can be illustrated by a section at Canterbury Lane (Fig. 3) which shows over 3.5 m. of stratigraphy reaching from a third-century clay-walled Roman building at the bottom to the bombed deposits of 1942 at the top, in which a semi-melted half-crown was stratified. This coin does not survive amongst the finds, for Frank Tomlinson at once took it to the Bank, where it was credited to the excavation fund. The section gives a vivid sense of general continuity through the ages, although on this particular site the sub-Roman black soil was not sealed until a gravel floor was laid down in the ninth century, and there is a gap of four centuries in the record. The upper levels of the 1949 trench in the St. Margaret's Street car-park had been dug by machine. The possibility of such an undertaking was a new development. In 1948 the Committee issued a public appeal for funds. A letter to The Times signed by Archbishop Geoffrey Fisher, the Mayor and other prominent figures invited 5 See Britannia, i (1970), 91. 6 For this site, see The Archaeology of Canterbury, v (forthcoming). 37 S.S. FRERE CANTER.BUR.Y LANE SITE C C.XXII C. IV H BLITZ RUBBLE WALL 0 ODD 0 cntro C 3 CELLAR. n g N >\ J FR.ACMENT5 akV\» 311 £E> CHALK RUBBLE ® m WA L L OLlArsc S) X « T- N +559 Oj .D^ 1i x c n D K CELLAR. \ FLOOR. 8 CR.AVEL 10 FOOTINGS I BUILDIKC A 4- sssw SAXON FLOOR. m •:•!' mm lisi psftiflv&S ;'!«i K l i S GRAVEL >I P I T * ® ^ " ^ wm« W®!Vm-? m 2*»0 A S H \ |S$%'?M PIT « 6 DARK SiS 5ANDV SILT OCCUPATION m LOAM BURNT DENSE EARTH FLOOR CHALK. >*•D AUB !d BLACK SOIL 6 [VctX Steiret t = i RSBBBS Fig. 3. Canterbury Lane Sections. contributions to raise £5000. '. . . The time has arrived when, if this great opportunity is not to be missed, a more extensive effort must be made, with paid labour and on a more ambitious scale . . . It is also essential to have substantial funds available so that when necessary more digging can take place from surface level as distinct from the exploration of basements, and so that paid assistance can be obtained if a particular site is threatened by development and there is insufficient time to carry out excavation with unpaid labour . . .'. There were huge areas between the streets (Plate IV) waiting to be tackled, which were all the more important since it was now clear that the Roman streets were on quite different lines from the modern ones and that consequently Roman buildings would not necessarily be found in the cellars beside the latter. Conventional trenching from the surface in these spaces, on the other hand, would be of little profit, since the volunteer labour available could hardly be expected to dig more than one area measuring say 15-20 m. long by 2-3 m. 38 CANTERBURY POST-WAR EXCAVATIONS PLATE IV •:~ Aerial View of Canterbury, showing bombed areas. wide to the necessary depth of 3-4 m. within the limits of a single excavation-season. The appeal had the advantage of advance publicity not only from the Butchery Lane mosaics - mosaics always appeal powerfully to the public - but also from a Royal visit. The King and Queen had found time during a visit to the city in 1946 to look at the Butchery Lane pavement and to inspect a small exhibition of finds (Plate V). The largest area which was opened by machine as a result of the new resulting affluence lay between St. George's Street and Burgate Street, west of Canterbury Lane,7 but was in many ways disappointing. The whole area was badly mauled by medieval pits and by a sixteenth-century cellar, and time was lacking for total excavation. The main results were, first, the tracing of a considerable length of 7 The Archaeology of Canterbury, vii (Maidstone, 1983). Area R, 69-76. 39 S.S. FRERE PLATE V « ~ w - : The Royal Visit in 1946. north—south Roman street and its junction with one running east, on the intersection of which lay a very interesting Dark Age timber building; and, secondly, realization for the first time of the true character of a Roman urban street. It consisted of 1.5 m. of solid gravel containing ten successive surfaces; two tough hired labourers took three weeks to cut a section through it. The street had clearly been laid down at a very early date, for it lay directly upon an ancient plough-soil yielding only a few sherds of pre-Roman pottery; the layer contained small chalk nodules which had been spread to marl the soil in the manner described by the elder Pliny." The conclusion could be drawn that here was a Belgic field, and that the nucleus of the Roman street-grid went back to the Claudian period, suggesting urban development considerably earlier than elsewhere in Britain. Paid labour under a foreman was hired from Messrs. Wiltshier, the builders. Although good at moving earth and at erecting shoring, these men were of course completely unskilled in archaeological work, and their proper control raised problems. They started work at 7.30 a.m., much earlier than the volunteers, and had a different lunch-break; even the simplest directions had to be transmitted " NH xvii. 42; cf. 45. 40 CANTERBURY POST-WAR EXCAVATIONS CANTERBURY MARKS & SPENCER MED* WELL Hul l ©- BELCIC OCCUPATION 946 AREA l l / 'L.J! M j 1 -H -t r STREET WEENED rn K 1952 TD i l I J l.n r \ C3?r ^ \ © R 4 s f MS M ^ ^ L»C » i I i^ W J D m i i i Ma \ MI2 ii MH2 IMS \ H M7 H I i - ' H 4 MEDIAEVAL F00T1NCS -&. 11 I I SCALE •. FEET LYONS kjLMJCHJ 1 O 20 30 40 SO 60 70 METRE 1 Fig. 4. Rose Lane Plan. through the foreman, who himself did no digging. On the whole the system, later used at Verulamium, of directly employing a corps of student labourers, proved much more successful. What did the 1944—60 excavations achieve? It must be remembered that virtually nothing was previously known. In 1861 James Pilbrow, the City Engineer of the time, had published a paper describing his observations when the streets were trenched for main drainage,9 but the results were far from illuminating in any detail. Uncertainty even 9 Archaeologia, xliii (1861), 151-64. 41 S.S. FRERE PLATE VI Area R: Plastered Clay Wall. remained whether the Roman city ended at the branch of the Stour which bisects the modern city near the County Hotel, or whether it extended westwards to the medieval West Gate. Still less was known about the medieval archaeology of Canterbury, as distinct from its architectural history. The chronology of medieval pottery hardly existed. Almost all medieval sherds were normally assigned to the thirteenth century. One of the first discoveries was that the city had been first settled in the pre-Roman Belgic period. Near Marks and Spencer's store in the Parade a Belgic drainage-ditch was found, full of pottery which belonged to the first half of the first century A.D.1" The plan (Fig. 4) well illustrates the small scale of operations; cellars soon became filled with excavated spoil and trenches had consequently to be kept small. Later on, Belgic levels were traced elsewhere in the city, as at no. 3 Palace Street; but the most important site of this period lay west of the main course of the Stour, in Whitehall Road, beneath what is now the main approach-road from London." Here the original '"Arch. Cant., Ixviii (1954), 101-43. " The Archaeology of Canterbury, viii (forthcoming). 42 CANTERBURY POST-WAR EXCAVATIONS Roman Watling Street was identified overlying part of a rectilinear Belgic building associated with deep delvings in the natural brickearth; nearby a complete sunken hut containing stratified floor-levels was uncovered. At the time this was an entirely novel form of building for this period although well-known in Saxon contexts.'2 Another achievement was the recognition for the first time of unfired walls of solid clay or brickearth. Such walls if timber-framed and especially if burnt are normally not hard to see. At Farningham Col. G.W. Meates in 1948 had identified a solid clay wall standing on a stone foundation,13 and later he found others of this type at LuUingstone.14 At Canterbury there were no helpful stone footings or sleeper-trenches. In 1949 in Area R two parallel lines of vertical white plaster were observed (Plate VI); they were 2 mm. wide and 0.3 m. apart; the painted surfaces faced opposite ways, and it was clear that this was a wall of brickearth.15 Later, in Canterbury Lane we found ourselves able to trace unplastered clay walls and to identify thick layers of horizontal clay as fallen wall-debris (Fig. 5); this had never previously been done, but it did provide an explanation for otherwise puzzling layers of brickearth which had been encountered on other sites in early levels and somewhat unconvincingly explained as dumps of make-up placed to combat damp conditions. The Roman theatre itself was a significant discovery.16 At the time the only other such structure known in a Romano-British city was the theatre of Romano-Celtic type at Verulamium, a building with seating carried on earthen ramps. The original late first-century theatre at Canterbury was a building of this type; but in the early third century it was converted to a theatre of classical type with seating carried at least partly on massive substructures of masonry. It is a curiosity that parts of the original gravel seating-ramps were retained in use instead of being swept away during the reconstruction. The new building had a diameter of 80.5 m., four-fifths that of the theatre at Aries and three-quarters that at Lyon; but both the latter were able to economise on substructure by being partly built into the sides of hills. This building is evidence of impressive wealth in Roman Kent. Another achievement of the early excavations was the light thrown 12 For pre-Roman Grubenhauser, see W. Rodwell in (Eds.) B. Cunliffe and T. Rowley, Lowland Iron Age Communities in Europe (BAR International Series No. 48, Oxford, 1978), 37. "JRS, xxxix (1949), 110; Arch. Cant., Ixxxviii (1973), 2, with PI. II. 14 G.W. Meates, The LuUingstone Roman Villa, i (Maidstone, 1979), 61. 15 The Archaeology of Canterbury, vii (Maidstone, 1983), 71-2. 16 Britannia, i (1970), 83-113. 43 S.S. FRERE CANTEKbUhXY LANE C.XXIII C V A TRENCH MECHANICALLY EXCAVATED TO THIS LEVEL 4.4-76 O.D fl U R.N SAXON PAT I mmm illillil SX/A. ///W4 CLAY WALL 3-eei Jteir&s I T+-36 O.D M. 12 C. XXIII C V D Fig. 5. Canterbury Lane Sections. on the defences. Not only was it established that the Roman Wall closely followed the line taken by the later medieval defences, enclosing the whole oval city of 52.6 ha. and not merely the eastern segment; but the date was found to be unexpectedly late, between 270 and 290, even though no trace existed of an earlier earthwork 44 CANTERBURY POST-WAR EXCAVATIONS defence.17 The absence of a late second-century earthwork is noteworthy; the explanation may be that until the late third century danger of Saxon raids was considered adequately countered by the early shore-fort at Reculver and by the various fleet-bases on the Kentish coast. Of Dark-Age Canterbury history is silent until the arrival of St. Augustine in 597, who found the place to be, as Bede describes it, a royal 'metropolis', containing at least one recognisable Roman church still standing within.18 It was an important event, therefore, when we began to find early Saxon sunken huts behind the Marlowe Theatre and in the Simon Langton school yard. At least one of these yielded Jutish pottery of the early or mid fifth century, and other sherds of similar date have been found on other sites. Archaeologists tend to build upon each others' conclusions, and today it is fashionable to follow the model established at Winchester, where except for a possible royal establishment there is thought to have been a hiatus in occupation within the walls.191 believe, however, that the evidence from Canterbury presents a different picture, one of continuous occupation throughout this phase. Formal proof of this contention is hard to establish in the absence of accurate dating for artefacts of the period, and continuity of occupation naturally does not carry with it any implication of continuity of urban organisation. However, the widespread scatter of early Saxon occupation within the walls should place the onus of proof on those who favour a break. We have noted that, when the excavations began, the chronology of medieval pottery was in its infancy. The problems of dating were slowly clarified, partly by observing sequences of intersecting pits or of stratigraphy, and partly by two fortunate discoveries. In Canterbury Lane the thick occupation-deposit above a Saxon gravel floor (Fig. 3, layers 11-14; Fig. 5, Section J-K, layer 4) contained - in addition to loom-weights, dog-dung and food-refuse - a large quantity of local pottery; and this was fortunately associated with sherds of imported Rhineland pottery known as Badorf-type ware and datable to the period 850-950. This material provided a late Saxon horizon, while in the Cathedral Precincts a group of sherds was found which could be shown to be earlier, but only just earlier, than the erection of Lanfranc's Dormitory c. 1080. Thus slowly a medieval pottery sequence was constructed despite the notorious lack of associated 17 The Archaeology of Canterbury, ii (Maidstone, 1982), for the defences. 18 Bede, HE i, 25-6. In i, 33, it is described as regia civitas. 19 M. Biddle, 'Winchester: The Development of an earfy Capital' in (Eds.) H. Jankuhn et al., Vor- und FriXhformen der europaischen Stadt im Mittelalter (Gottingen, 1973), 236-41. 45 S.S. FRERE j datable material such as coins, which in the medieval period were rarely lost in useful contexts. The archaeology of a city is never fully learnt. Future work will always adds information and will often change perspectives. Moreover the kinds of question to which an answer is sought suffer alteration over the years with the development of research. Despite the necessarily very restricted scope of operations, the early excavations at Canterbury succeeded in creating some kind of framework; but it is always desirable to test and extend such frameworks with renewed excavation and improved techniques. We are fortunate that the Canterbury Archaeological Trust has had the opportunity to do just that, and by operating on a much larger scale has been able both to correct false deductions and to extend our information so dramatically. In the early years the Kent Archaeological Society was able to publish results in Archaeologia Cantiana; but as time went on less could be accepted as other excavations competed for space, while at the same time the Canterbury excavations themselves amassed ever more massive collections of data. We owe the Society a great debt of gratitude now that new conditions have enabled it to sponsor the full publication of both the old and recent programmes of work in a new Monograph series, and for ensuring that the reports are complete in themselves without either separating descriptions from evidence in a frustrating series of fascicules or encapsulating large portions of either in unreadable microfiche. Once again Canterbury is leading British archaeology in a new direction. 46