Excavation of a Medieval Garderobe in Snargate Street, Dover, in 1945

EXCAVATION OF A MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER, IN 1945 By M. M. Rix and G. C. DUNNING With contributions by P. A. SABINE, J. E. MOREY, E. G. CROWFOOT and P. TUDOR CRAIG I. THE EXCAVATION By M. M. Rix, M.A. INTRODUCTION BETWEEN 1943 and 1945 I was stationed in Dover as an instructor in the Army Education Centre where I met Mr. Eric Taylor of the Civil Defence. On the morning of February 19th, 1945, the demolition group in which he was working to level a bhtzed site between Snargate Street and Adrian Street, Dover, broke into a vault dug into the hillside: on the spoil heap from previous levelling a piece of eleventhtwelfth century pottery was found which he brought to me. On my recommendation the levelling there was suspended and plans made to excavate the site in so far as war-time conditions allowed.1 Digging began on March 17th under the supervision of myself and Mr. Taylor with the assistance of three boys from the Dover County School. It soon became clear that the site was the basement floor of a garderobe or medieval lavatory. By Aprh 5th floor-level had been reached in part of the garderobe; during the following week the whole interior was cleared and what later proved to be a Roman building, into which the garderobe had been intruded, was discovered. Within another week, I had been posted away from Dover and was soon sent overseas. The finds were left with Dover Museum (then curatorless and in store) and the Forces Education Centre (which later returned to civilian use). The notes of the excavation were left with Mr. Taylor who later became Curator of the Dover Museum. During my absence abroad Mr. Taylor died suddenly, and as he had no relations and few friends, it was impossible to salvage his papers. The fohowing report has been tardily put together from fragmentary notes, sketches and the photographs of others, as a memorial to Eric Taylor, but for whose 1 A short note on the site is in Arch. Cant., LVTH, p. 74. 132 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER vighance the site and much of its contents might have been lost, but for whose untimely death Dover might have, what it has long needed, someone to record its numerous archaeological treasures. THE SITE The Castle Cliff on one side and Western Heights on the other embrace the mouth of the River Dour which gave its name to Roman Dubris, as to modern Dover. While the Castle Chff towers vertically above the vaUey and sea shore, Western Heights descends more gently to the harbour, and where its chalky tail finally disappears into the silt of the Dour estuary Snargate Street runs on its seaward side. Snargate Street, the fashionable quarter of an earlier Dover, the probable limit of the Roman coastline just here, now presents a slummy, dilapidated, blitzed appearance. In the chalk tail of Western Heights that rises inland of Snargate Street's eastern end the garderobe was discovered. It is the basement floor of a medieval building inserted into the footing of a Roman gateway (Fig. 1). The garderobe first came to hght early in 1945 when Civil Defence teams were at work clearing the remains of buildings destroyed by enemy action in this area: on February 19th they broke into what appeared to be a tunnel running into the hillside in a northerly direction, which ultimately proved to be the eastern half of the garderobe. Among the spoil was found a large fragment of the upper part of a Norman wine pitcher published by Mr. G. C. Dunning.1 It seems clear that this fragment came from the area surrounding the garderobe, rather than from within the garderobe itself: in the early period of the excavation a dozen smaU pieces of Samian ware were found, together with tufa and characteristic pink Roman mortar, all in the soil surrounding the garderobe, none in the garderobe itself. THE STRUCTURE The structure consisted of a pah of paraUel barrel-vaulted chambers connected by twin arches supported on a central column (Fig. 2). The waUs were of coursed stone (mostly Kentish rag) and the vaults and arches of chinch (i.e. quarried hard chalk) of truncated wedge shape so as to hold in position by theh own weight, without mortar. At the north end of each chamber was a hole in the ceiling above a shaUow chute built into the stone wall. The internal measurements of the whole structure were 8 ft. long by 8 ft. 6 in. wide by 7 ft. 6 in. deep from the floor to the keystone of the vaulting. The vaulting of the eastern chamber had fallen in previous 1 Antiq. Journ., XXV, 153. MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER 0 I *-\ A D e i >> u s h • K S 11 ^ 3 M k t t . G » > T 6 20 0 50 100 FEET ' • ' ' ' ; 1 1 FIG. 1. Plan of part of Dover, Kent, with medieval garderobe hi black. to the 1945 operations, but that of the western chamber was in good preservation except for the southern end and presented interesting features. The south wall had been shorn off, so that whether it was a blank wall or had outlets along the natural line of drainage towards the sea there was no means of knowing. Into the lower courses of the junction between the south and west wall a large rectangular block of clunch had been incorporated which excavations subsequent to 1945 proved to be a Roman sill stone in situ (Fig. 2)'. In the west wall there were Wo rectangular cavities, 4 ft. 6 in. from the ground, 10 in. from the end of the wall, one at either end: a simhar cavity occurred in the west wall in the same relative position at its northern end. These cavities varied between 4£ and 1\ in. wide, between 6 and 1\ in. high and between 9 and 15 in. deep into the thick- 134 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER 10 FEET FIG. 2. Medieval garderobe formerly in Snargate Street, Dover, Kent. 135 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER ness of the waU. They were no doubt used to hold timber supports for the scaffolding upon which the vaulting was constructed. The column supporting the communicating arches consisted of four square blocks of freestone measuring 9£ in. broad and long and 5£ in. high, surmounted by a rectangular capital, measuring 9f in. square at the bottom and 12£ in. square at the top and 6£ in. thick: the upper 3-f in. had a vertical edge from which the stone was chamfered to the lower edge. The clunch arches that sprang from this capital to the northern and southern walls were shghtly thicker than the top of the capital, the keystone of each being 14 in. long. Timber supports were placed beneath these arches by the kindness of the Dover Borough Surveyor's Department, but the pubhc nature of the site resulted in the filching of the timbers, the collapse of the arches and of the western vault. When I visited Dover in 1948 the structure was ruinous and derelict. The vaulting was constructed of truncated wedge-shaped clunch blocks of varying size, presumably quarried and shaped in the vicinity. They were each marked with strokes on the face that showed—I, II, III in order of ascending size. Measurements of characteristic blocks were as follows. They were all 6 in. thick at the top and 9J in. high: they varied in length from 15| in. to 8 in., and in thickness at the narrow end from 5 to 3 | in. The chutes were not central to either section of the north waU, nor were they exactly simhar in width or depth, thus giving the impression that theh design and construction was largely a matter of rule of thumb. The floor was flat and weU cemented, as if it were intended to clean the chamber out from time to time: this theory is further supported by the chipped nature of the clunch blocks that form the arches between the chambers: they had clearly been subjected to rough usage at some time between the building and abandonment of the garderobe. The N.E. corner of the floor had been tiled with roofing-slates imported from Cornwall (p. 152). THE FINDS The finds included pottery, bone, shell, fabric, metal, glass and decorated stone work. Mr. Dunning has kindly examined the pottery and provided the report which follows (p. 138). As whl be seen it included glazed and unglazed medieval sherds, also a number of fragments of the painted ware from western France brought over by the Gascon wine trade. Although many fragments could be fitted together, no remains of a complete pot were found, which is not surprising in view of the nature of the site. The pottery was well distributed throughout both chambers. 136 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER Much domestic animal bone, oyster shell and whelk sheU was discovered: at one level a remarkable number of small bhd bones were found: it was impossible to preserve any of these. In one particularly damp part of the west chamber a mass of fabric was found which was sent without being allowed to dry to Dr. H. J. Plenderleith at the British Museum Laboratory for treatment and identification; thanks to him, sufficient of it was recovered to be identified as silk, and for the diamond pattern of its weave to be clear. Miss E. G. Crowfoot had kindly reported on the silk (p. 155). A certain amount of heavily corroded iron was found: also a plain bronze tag or " point " was found in association with the silken material. Seven small fragments of glass were found. The majority were apparently flat window glass, but two seemed to belong to blown glass vessels, of which one was identified by Dr. D. B. Harden as probably coming from the Near East. A number of fragments of carved and worked building stone were found in the fining. Two of these are late twelfth century, and the other two are early thirteenth century (p. 156). One of the latter, carved with stiff leaf ornament, is of considerable interest for the development of this style in Kent, and Dr. P. Tudor-Craig has kindly discussed its context (p. 157). There is now no means of telling from what buhdings or structures these fragments were derived, before being incorporated in the debris that accumulated on the floor of the garderobe (Plate I). CONCLUSION In the present fragmentary state of our knowledge of medieval Dover, few conclusions can be drawn from the study of the garderobe and its position. The name Snargate Street certainly indicates that the waU of Dover in the Middle Ages ran in this area, but whether the garderobe was the base of a tower that formed part of the defences, it is impossible to say. AU that is certain is that judging by the pottery, the building was in use during the thirteenth century. The marked blocks in the vaulting suggest that there must have been some system for mass producing the various sizes of clunch blocks, probably in the quarry itself. The position of the garderobe in relation to the modern house line in Dover (see Fig. 1) makes it clear that the backs of the houses in Adrian Street rested on the south wall of the structure and that theh back gardens descended steeply from that point as is proved by a flight of steps. Only further excavations on either side of the site can establish further details about the medieval context of the garderobe. 137 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER II. POTTERY AND OTHER FINDS1 By G. C. DUNNING, F.S.A. POTTERY The pottery from the filling of the garderobe is remarkable for the range of wares represented in a deposit which appears to have accumulated in a comparatively short space of time, and for the diversity of its origins. Detailed descriptions of the pottery foUow, but first may be given a summary of the various groups and the trading connections which these demonstrate. 1. Fragments of three jugs of painted ware imported from western France, which were brought over by the wine trade of Gascony. These jugs provide the best evidence for dating the deposit and the other material contained in it. 2. A group of jugs of red ware with plastic decoration, or plain and green glazed. These belong to a large group widely distributed on both sides of the North Sea, which owes its diffusion to the wool trade. 3. Glazed jugs of a plainer sort, and domestic cooking-pots and bowls, which were made at the kiln site at Tyler Hill, near Canterbury. 4. A large dish which originated in east Cornwall, and no doubt reached Dover in the course of the coastwise trade in Devonian roofingslates, of which examples had been re-used to patch the floor of the Snargate garderobe. The assemblage of pottery is therefore of interest not only for its own sake as iUustrating the diverse kinds of wares which reached Dover and were used there, but also as showing how archaeological material can be used to demonstrate the trade connections of this medieval port by land and sea in the late thirteenth and early fourteenth centuries. 1. Imported French Pottery (Fig. 3) The deposit contained fragments of three jugs imported from Western France in the late thhteenth and early fourteenth centuries. Two of the jugs are polychrome ware, finely painted with green and yeUow designs outUned in dark brown, and the thhd jug is monochrome with green glaze covering the body, spout and handle. 1. Part of side and complete base of fine white ware with thin colourless glaze on body. It is decorated above the contracted foot with a dark brown line above a green band. There is no sign of decoration on the smaU part preserved above the brown line. The jug was probably pear-shaped, about 13 in. high, and it would have a large bridge spout opposite the handle. The shape and size, as restored in 1 In all the hazards that the finds have suffered it is not surprising that some of the objects listed by Mr. Rix have been lost. This report is on all the finds that can now be identified in Dover Corporation Museum. 138 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER the drawing, are based on simhar jugs from London, Ipswich, and Glastonbury Abbey. The decoration on the upper part of the body would be either a floral scroll with leaves and bud, or large heatershaped shields. 2. Small fragment from the upper part of a jug of fine pink ware with lustrous glaze. The painted decoration shows two lines in dark brown enclosing bright green, identified as the neck and part of the body of a bhd. The complete design was a bhd with large beak and long tail , r i ? C\ v\ f *ra ff s1- A k- if v*> 1 4 i i / ' i I %. C3 m '••••—"J w "V>> < ^ II \ I I \ I \ ni li II \ i // n / y \ smmSfM \\ / ; FIG. 3. Imported French Pottery (J). feathers, like a peacock, of which examples are known from London,1 Felixstowe, Stonar,2 and Cardiff. The bhds are in pahs flanking a small yeUow shield, aheady represented at Dover by a sherd found on the site of Fox's Bakery in Queen Street.3 3. Three fragments from the same, jug, of fine white ware with lustrous glaze. One piece is a bridge-spout, and the others are the rim 1 Archceologia, LXXXHI, p. 128, Fig. 13, e-f. 2 Arch. Cant., LTV, p. 60, PI. H, 5-6. 3 Arch. Cant., LXTV, p. 147, Fig. 12, 35. 139 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER opposite the spout and the lower half of a wide strap handle; all are glazed in hght green. The glaze is streaky and mottled, and poorer in quality than on the polychrome ware. The fragments belong to a jug about 9 in. high, glazed green over the enthe surface. A complete jug of this type, probably found in London, is in the collection of the Society of Antiquaries1 and fragments have been found at Stonar, Pevensey Castle, and elsewhere. These jugs were made in the same region as the polychromes, but at a slightly later date in the first half of the fourteenth century. 2. Imported Decorated Jugs (Fig. 4) 1. Upper part of smaU jug of fine brick-red ware. The neck and upper part of the body are covered with a thin glaze of good quality, lustrous yellow with a few green spots. The decoration is in white shp in high relief. A narrow strip forms three arcades, with the lower ends joined by peUets impressed by a ring and dot stamp. Inside each arcade is a rosette of four petals, marked at the centre by a stamped pellet, and in each upper space is a simhar floral motif of three petals. On the front of the neck is a human mask built up of applied clay. The ears are of white clay, weU modelled, with the lobes indicated. The left eye remains and is a peUet with ring and dot stamp, green glazed. The nose is missing, but part of the chin remains. The neck of the jug is cylindrical, with thin-edged rim, and the handle is plain and chcular in section. The shape of the jug has been restored after No. 2. 2. Lower part of simhar jug of light red ware with thin white surface shp outside. Light yellow glaze covers the shoulder and thins out towards the base. The base has a moulded edge and is raised at the centre. On the shoulder is a green-glazed pellet with ring and dot stamp, and parts of two petals of a rosette in hght red clay. The decoration was therefore of the same pattern as on jug No. 1, and has been restored accordingly on the drawing. 3. Body of simhar jug of hght red ware with white surface shp on upper part. Over-all thick dark green glaze on neck and body down to the shoulder, also on the handle and in a broad strip below it to the base. The jug is plain, with base moulded as on No. 2. Handle plain, circular in section. Not illustrated. Part of side of jug of hght red ware with white surface shp. Dark green mottled glaze above the shoulder. In shape the jug was similar to No. 3, but taller above the bulge. 4. Part of neck and side of large jug of brick-red ware with white surface slip. Over-aU lustrous glaze of good quality, clear yellow with green streaks and flecks. The decoration is the same in design as on 1 Archceologia, LXXXIII, p. 129, Fig. 14, 2. 140 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER ' FIG. 4. Imported Decorated Jugs (£). 141 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER the smaUer jugs Nos. 1 and 2, and is carried out in dark brown slip. The arcade is narrow and ridged. In the upper space is a disc about 11 in. across, and at its centre a white pellet with ring and dot stamp. Below this is one petal of a rosette marking the lower end of the arcade. This group of jugs is easily distinguished from the pottery known to have been made in Kent by its finer quality and the use of plastic decoration. It has, moreover, analogues amongst finds from a number of sites in East Anglia and the eastern counties. For these reasons the group is regarded as having reached Dover in the course of coastal trade from East Anglia. The red ware is distinctive of pottery found in the Cambridge region, where decorated jugs of this fabric appear to range from the thhteenth century to the fifteenth. The earlier jugs of the series are decorated in a variety of techniques, either with a zone of chcular stamps on the neck and vertical ribbing on the body,1 or painted in white slip with strip and peUet design,2 or other simple linear patterns. Away from Cambridge, a jug of the first type has been found in Colchester. 3 Jugs of the second type have also been found at Colchester4 and at Ipswich.5 The last two jugs, and those of this sort at Cambridge, have sagging bases with thumb markings, simple linear decoration in thin white shp, and the technique is generally primitive. The trade in these jugs reached Kent, for Mr. Stebbing has closely similar fragments in his collection from Stonar. From this site he also has face jugs and green-glazed decorated sherds with significant paraUels at Cambridge and in the Midlands, and even as far north as Yorkshhe. The finds at Stonar are evidence that in the thhteenth century the ports of East Kent were linked by coastal trade with East Anglia and possibly with ports further north up the east coast. The pottery discussed above is rather earlier in the thirteenth century than the jugs from Dover, which are technically more advanced, have flat moulded bases, a white surface shp to conceal the red body colour and so enhance the colour of the glazes, and are decorated with plastic designs in high rehef. In the University Museum of Archaeology at Cambridge there are comparable sherds from the University Press in Mill Lane and from St. Catherine's College. These are rims and necks of jugs in brick-red ware with white surface shp and speckled green glaze, very simhar in character to the glaze on Fig. 4, 4. Although no parallels are forthcoming at Cambridge for the pattern of arcades and rosettes, these are by no means frequent anywhere in the south. The style has antecedents in the first half of the thhteenth 1 Rackham, English Medieval Pottery, PI. 33. 2 Rackham, English Medieval Pottery, PI. 52. 3 Colchester Museum Report, 1914, p. 14, PI. V, 2. * Ibid., 1928, p. 69, PL XXIII, 1. 13 Ipswich Museum, No. 1929.91. 142 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER century,1 and in the form of scroUs and other curvilinear patterns is familiar in the Midlands.2 The later medieval red wares at Cambridge are distinguished by elaborate and varied patterns incised through a buff surface shp (graffito technique).3 Jugs of this class have been found at a number of sites in Cambridge but, as in the case of the earher red wares, the place of manufacture (assumed to be in this region) is not known. Away from Cambridge the only example known is the jug with heraldic decoration in the Canterbury Museum, and presumed to be a local find.* This brief discussion whl suffice to show that a specific relationship exists between certain jugs of red ware found in East Kent and finds made in East Anglia, centering on Cambridge. The jugs appear to be the products of the same industry, and in date range from the thhteenth century onwards for about two centuries. The finds from Kent show that this connection was maintained throughout the period; the earhest phase is represented at Stonar, the next at Dover, and the last phase at Canterbury. Study of aU the material has enabled the Dover jugs to be placed in a wide context of coastal trade, but has not defined theh date any closer than as required by the associations at the site, and on typological grounds. Probably they belong to the latter part of the period covered by the deposit, that is, the early fourteenth century. The long-standing trade along the east coast which is demonstrated by the pottery is not capable of explanation on the Enghsh evidence alone. Mention must also be made, however briefly, of the finds of comparable pottery on the Continent. The material is very diverse in character, but may be divided into five groups: 1. „ Jugs with large stamps of scallop-sheUs. Several examples are from London (GuhdhaU and London Museums), and single examples or fragments from Stonar, Pevensey Castle, and Knaresborough Castle, Yorks. In Belgium there is a jug found at Bruges. In HoUand there are three finds: in the Wieringermeer5; at Welsrijp,6 between Franeker and Leeuwarden; and at Kloosterterp, south of Ferwerd. The Welsrijp jug is a remarkable vessel, and has rows of scahop-sheUs above and below a series of large stamps of the figure of a fiddler standing inside an elaborate arch or doorway. 2. Jugs with scroll patterns in shp, as found at Leicester and 1 At White Castle, Monmouthshire. Aniiq. Journ., XV, p. 333, Fig. 2, 6-7. 2 K. M. Kenyon, Excavations at the Jewry Wall Site, Leicester (Society of Antiquaries Research Report, 1948), pp. 236-43. 3 Proc. Cambridge Antiq. Soc, XLVI, p. 21-6. A Proc. Cambridge Antiq. Soc, XLVT, p. 26, Fig. 11. Rackham, op. cit., PI. 59. 6 Oudheidkundige Mededeelingen, N.R. XIII, p. 40, PI. VI, 1-2. • Bonner Jahrbucher, CXLII, p. 171, PI. 45, Fig. 1, middle. 143 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER Coventry,1 Nottingham, Lincoln, and Cambridge. In HoUand a fine example is from the manor house at Starrenburg,2 near Rotterdam, built about 1240. 3. Jugs with shp patterns, some forming chevrons or arcading, and petals or floral motifs in the spaces. A fine range of these motifs is on jugs found at White Castle, Monmouthshire,3 dated to the first half of the thirteenth century. Probably there was a trade in pottery to south-east Wales from the Midlands.' The jugs from Dover (Fig. 4) are fine examples of this class. A comparable jug is from Nijeklooster, a monastic site north-east of Sneek, on the east side of the Zuider Zee, in the Friesch Museum at Leeuwarden. This jug is of hght red ware with hght yeUowish-green glaze on the body and neck. The chevron is black, and the triple petals in the spaces are white. The base has a moulded edge and is raised at the centre. Other jugs of this class have large rosettes on the body, as on a pot from Malines. Another typicaUy Enghsh motif, the strip and peUets, is weU represented on a jug found at Hamburg.4 4. Jugs with a long tubular spout springing from the upper part of the body, and held to the rim by a strut. These are grouped in Yorkshhe; several at York, also at Rievaulx Abbey, Malton, Whitby Abbey, and Scarborough. They passed by trade down the east coast to East Anglia, and have been found at Tydd St. Giles, near King's Lynn, Cambridge, Chesterton, and Dunwich. Fragments of this type (though the spout is missing) are from Stonar. The type has not yet been found abroad. The finest examples of this class have elaborate plastic ornament on the sides of the neck and body, covered by a rich green glaze. On each side are two armed horsemen with long kiteshaped shields, and below them are stags attacked by hounds. Evidently the composition represents the manly pursuits of war and the chase. A large fragment is from Cambridge,5 and recently the Peverel Archaeological Group has found a superb jug in a late thhteenth century context at Nottingham.6 5. Face jugs. Usually these are of grey ware with green glaze. The masks are placed laterally against the rim, and may have a long beard being held or stroked by a hand. The main occurrence is in the Midlands and East Anglia, at Nottingham, Lincoln, Peterborough, 1 Excavations at the Jewry Wall Site, Leicester (1948), p. 236-43, Figs. 69-75. 2 Berichten van de Rijksdienst voor het Oudheidhundig Bodemonderzoek in Nederland (1951), I, Plate after p. 39. 3 Antiq. Journ., XV, p. 333, Fig. 1, 3, and Pig. 2, 6-8. 1 B. Rackham, Medieval English Pottery, PI. 89 (wrongly stated as found in Denmark). 6 Rackham, op. cit., PI. 12A. 0 G. C. Dunning in Peverel ArcMeological Group, Annual Report for 1964. 144 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER King's Lynn, near Cambridge, Norwich, Great Yarmouth, and Dunwich. Typical examples have been found at Stonar. Abroad the type is known only once, from the Wieringermeer.1 At Dover the face jug is decorated with slip pattern (Fig. 4, 1). The distribution of these five classes of pottery is shown on the map, Fig. 5. In England the scallop-shell ornament occurs mainly in London, but the main incidence of the other motifs plotted is further north, in the Midlands and Yorkshire, and the places of making of these jugs are doubtless in these regions. There is a marked concentration of the pottery at or near the eastern ports, such as King's Lynn, Great Yarmouth, Ipswich, Sandwich, and Dover. The pottery found on the Continent is considered to be Enghsh in origin. It is certainly English in inspiration, and the bulk of it may be accepted as exported from this country. It is possible, of course, that some of the material was made abroad in imitation of the Enghsh styles, but this does not greatly affect the main issue. The pottery is seen to be grouped in two areas; first in Belgium and south Holland, between Bruges and Rotterdam, and secondly in north Holland in the provinces of Friesland and Groningen. In addition there is a single outlier at the great medieval port of Hamburg. The economic background of this carrying trade in pottery is clearly the trade in wool, the leading export of medieval England. The Netherlands was one of the most important markets for Enghsh wool, and the distribution of the pottery abroad shows concentrations in the region of the staple wool ports of the Low Countries, and in the hinterland of the north Dutch ports engaged in the trade at the close of the thhteenth century.2 The available evidence for dating the pottery places the bulk of it in the second half of the thhteenth century, that is, during the period of greatest activity of the wool trade. 3. Tyler Hill Ware (Fig. 6) J. Body and base of jug of grey ware, orange-red surface inside, brown outside. Dark green glaze on body down to bulge. Body of slender baluster form, with zone of four ghth-grooves above the bulge. The base is sagging, and the edge has closely set thumb-marks, which press the angle down to the lowest level. The restoration is based on similar jugs from the well in Canterbury Lane and from the site of the County Hotel, Canterbury, also of Tyler Hill ware, giving a jug about 14 in. high. 2. Upper part of jug of grey ware, brown surface inside, brown with red patches outside. Thick dark green lustrous glaze on neck and i Oudheidk. Mededeel., N.R. XIII, p. 40, PI. VI, 8. 2 Cf. H. C. Darby, Historical Geography of England before 1800, p. 304 ff., with Figs. 61 and 56. 145 • SCALLOP-SHELL STAMPS SCROLL DECORATION A SLIP PATTERNS A TUBULAR SPOUTS + FACE JUGS lOO MILES FIG. O. Distribution-map of Decorated Jugs (thirteenth and fourteenth centuries) in England and abroad. MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER 6W FIG. 6. Tyler Hill ware 147 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER body down to bulge. Decorated with zones of two girth-grooves on neck and body. Six wide bands of white shp extend down the side from the neck to the bulge, and are covered by the glaze. The shape is restored after simhar jugs from the Canterbury Lane weU, giving a jug about 15 in. high, with thumb-pressed base. 3. Lower part of large jug of grey ware, orange-red surface inside and outside. Patch of lustrous green glaze on bulge. Decorated with girth-grooves on bulge. Wide sagging base, with the edge closely thumb-pressed; the marks do not, however, reach to the lowest level of the base. 4. SmaU dish of grey ware with brown surface. The inside of the base and side are covered with mottled green glaze. The dish has a curved side and sagging base, and the outside surface has been smoothed by trimming with a knife. A precisely similar dish was associated with thhteenth century pottery at a site in Stour Street, Canterbury.1 5. Upper part of cooking-pot of grey ware, brown surface inside, dark grey-brown outside. The rim is sharply moulded outside, and has a wide internal bevel with sharp inner edge. On the bevel is a line of smaU pin-holes, which pass through to the outside. 6. Upper part of large cooking-pot of grey ware with orange-red surfaces, grey toned outside. The rim is heavily moulded outside, and has a wide internal bevel with pin-holes passing through to the outside. Decorated with apphed thumb-pressed strips. Below the neck is a broad strip pinched up into peaks, and from it vertical strips pass down the body. The pot was about 16 in. maximum diameter and 13 in. high. The pottery in this group has exact parallels amongst the material from the kiln site at Tyler Hill, 2 miles north of Canterbury.2 This apphes to the types of the jugs, theh decoration and glazes, and minor but characteristic features such as the bands of white shp on the body. The same identity apphes to the cooking-pots, which in ware and rimsections, pin-pricking through the thickest part of the rim, and the finger-printed decoration, are also exactly matched at Tyler Hill. The identification of these pots found at Dover as products of the Tyler Hill kiln raises the question of the extent of the trade in pottery from this centre in the late thirteenth century, and fortunately there is a certain amount of material. The situation of the kiln close to the large and important medieval city leaves little doubt that the bulk of the pottery was destined for Canterbury, either for use in the city or to be sold in the markets there. The most remarkable evidence at Canterbury is from the well found Bj, 1 Arch. Cant., LXIV, p. 68. 2 Arch. Cant., LV, p. 67. 148 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER between St. George's Street and Burgate in 1952. In the filling were over fifty jugs and pitchers, as weh as vessels of other types. The majority of these jugs are identified as made at Tyler HiU.1 Apart from Canterbury and Dover, five other sites have produced this ware. At Richborough, a certain amount of medieval pottery was found near the fallen east waU of the Saxon Shore Fort. Much of this comprises pieces of jugs, bowls, and cooking-pots of Tyler Hill ware. At Stonar, the medieval port situated on the Wantsum Channel, Mr. W. P. D. Stebbing has recovered several fragments of the same types and recognized theh origin.2 At Sandwich, a very fine jug, now in the Canterbury Museum, was found in digging in the graveyard of St. Clement's Church in 1953; the jug is 14J in. high and in shape, ware and decoration is a close parallel to the large jug from Tyler Hill.3 From Marshside, Chislet, on the west side of the Wantsum Channel, there are thumbed bases of jugs and part of a skillet in the Heme Bay Museum. The last site is Wingham, where fragments of jugs, some with incised decoration, were found in digging trenches in 1955. The distribution of Tyler Hill ware, as far as known at present, is shown in Fig. 7. Three of the sites are grouped at the eastern end of the Wantsum, and clearly the pottery reached this area by water transport from Canterbury (or Fordwich) along the Great Stour. The concentration of sites near Stonar suggests that this was a distributing centre, from which Dover could readily have been reached by ship along the coast from Sandwich Bay. Alternatively, the pottery could easily have travelled to Dover dhect by road from Canterbury, though finds from intermediate sites are lacking to show that this land route was used to distribute the pottery from Tyler Hill. The recent find at Wingham does indeed suggest that Watling Street was used to reach the villages nearer Canterbury. Dover is the site furthest from Tyler Hill and 17 miles to the southeast; this distance is weU within the range known to have been reached by the products of a medieval pottery kiln.4 It is uncertain if Tyler HiU supplied pottery to the west of the Blean area. The larger settlements in mid Kent had their own local kilns, such as those recorded at Week Street, Maidstone,5 and at Potters Corner, Ashford,6 both working at about the same time as Tyler Hill. All the evidence goes to show 1 Information from Mr. S. S. Frere, F.S.A. For other Tyler Hill pottery from Rose Lane, Canterbury, see Arch. Cant., LXVIII, p. 132 ff. 2 Briefly noted in Arch. Cant., LV, p. 47. 3 Arch. Cant., LV, p. 58, Fig. 1, 1. 4 Mr. E. M. Jope has mapped the pottery made at Brill, Bucks, in the thirteenth century. It is densely distributed over the Oxford area for 20 miles from the kirns, but the finer wares were traded farther afield for about twice this distance. See Trans. Bristol and Glos. Arch. Soc, 71, p. 71 and Fig. 11. 5 Arch. Cant., LV, p. 64. « Arch. Cant., LXV, p. 183. 149 13 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER that Tyler HiU supphed pottery to the coastal part of Kent east of Canterbury. The pottery from this local source was no doubt cheaper and obtained more easily and regularly than the finer quality wares brought by sea from East Anglia, which have aheady been described. ' .Ww-O rH* * t^ * .V'-,-.- »x. I a >'•%\W K* At! i IV^.'.'N MA 4? J »> *msK. « m#? A / ; ' : ^ ' A i ASS «»w »w ^ 1 ( V J C J \ ~\ wmmm I r.r J 1 t^ i ^ ~> z1 1 s I \ y C J r-. / I Ml ^ ; v A /- /•n *-^ \ V / c I *-**,n i FIG. 9. Roofing-Slate (£) The hone was submitted to the Geological Survey and Museum for petrographical examination, and Mrs. J. E. Morey has kindly made the following report (Enquiry 1412): " The hone is made from an irregularly banded dark reddish- 153 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER brown ferruginous shtstone. The rock is composed of grains up to 0-2 mm. in diameter of quartz, feldspar (plagioclase and perthitic orthoclase), muscovite, biotite and calcite. Fragments of mudstone and chert also occur. There is a considerable quantity of limonitized micaceous material arranged in streaks and lenses. Iron ore grains consist of leucoxene, ilmenite, or magnetite altering to hmonite. The matrix is a fine-grained micaceous and quartzose cement heavily stained with limonite. Calcite occurs in patches. " The rock is a fairly common type. It is possible that it comes from the Palaeozoic rocks of the Midlands or Welsh Borderland." .' i * '. SB • H f ' ( ! >l Hi FIG. 10. Whetstone (£) The Dover whetstone is a well documented addition to a number of sharpening stones made of local rocks in the Midlands, which were exploited for this purpose in medieval times. The rocks differ in character, probably indicating several sources of supply over a wide area. Some half a dozen hones have now been recorded, and the details may be briefly summarized as follows. The earliest in date are hones of the twelfth century, of pink micaceous sandstone from Alstoe Mount, Rutland,1 and of buff sandstone from Deddington Castle, Oxon.2 Slightly later are examples of the twelfth or thhteenth century, of 1 Antiq. Journ., XVI, p. 401. 8 Berks. Arch. Journ., L, p. 68. 154 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER brown quartz-siltstone from Sunningwell, Berks,1 and of yellow sandstone from Castle HiU, Shaftesbury, Dorset.2 FinaUy, the hone of brown shtstone from Dover is dated late thhteenth or early fourteenth century. These few examples suffice to indicate the extent of the trade in hones from the Midlands to southern England, beginning in the twelfth century and continuing during the Middle Ages. No doubt many more hones of these and simhar rocks exist in museums, but caution is necessary in coUecting the evidence, because already in the Roman period the rocks of the Midlands and the Welsh border were exploited for the same purpose, and the hones traded far afield.3 It may be added that important whetstones of schist were extensively used in England during this period.4 Examples have been found at several sites in Kent, including Canterbury and Stonar, and though hones of this rock may be expected at Dover, none is yet known. FIG. 11. Pattern of a piece of silk fabric SILK FABEIO Fig. 11. The drawing of the pattern of the piece of silk was submitted to Miss E. G. Crowfoot, who kindly writes as follows: " There is not much to be said without being able to see the actual textile. From the drawing the weave can be described as having the appearance of an extended birdeye twill, slightly hregular, repeating on 39 threads in one system, and 33 in the other. Unless there was selvedge present on the fragment one cannot say from the drawing which is warp and which weft. This irregularity may, of course, be a mistake 1 Ibid., p. 68. 0 Information from Mr. E. Jervoise, F.S.A. 3 K. M. Kenyon, Excavations at the Jewry Wall Site, Leicester, p. 273. See also Trans. Shropshire Arch. Soc, LIV, p. 142. ••> G. C. Dunning in Proc. Isle of Wight Nat. Hist, and Arch. Soc. (1938), II, p. 682ff. and Excavations at the Jewry Wall Site, Leicester, p. 230 and Fig. 65. 155 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER in the original drawing, though it is quite a likely pattern. I have made a couple of diagrams (Fig. 12) which I hope whl make clear what I mean. In both the weft repeats on 21 threads; on the left (a) the warp repeats on 27 threads, while on the right (6) it repeats on 21 threads. I should think that the Dover shk must have been a fine weave, as coarse bhdeye twills often have a cross instead of the dot in the centre to avoid long floats." a FIG. 12. Diagrams of irregular birdeye twill extended AROHITECTTIEAI FRAGMENTS PI. I. Four fragments of carved or worked stone. Mr. R. Gilyard- Beer, F.S.A., has kindly examined these and his remarks are included in the descriptions. The specimens were submitted to the Geological Survey, and Mrs. J. E. Morey has kindly reported as follows: " Nos. 1 and 4 are a buff-coloured, fine grained ' powdery ' limestone composed of rhombs and fragments of crystalline calcite in a matrix of micro-crystalline granular calcite. No. 2 is similar, but finer grained. " No match has been found for this buhding stone amongst specimens in the Survey's coUections. It does not appear to be Portland Stone or Aire of the usual Bath building stones, but we whl keep the rock type in mind for future reference." 1. Pyramidal top with beading down the corners. Possibly a decorative nail head of smaU dog tooth. Late twelfth century. 2. Part of roU moulding with double chevron or two straight, round-sectioned mouldings. End of twelfth century. 3. Part of shaft of Purbeck marble, 2f in. diameter. Probably early thirteenth century. This material was used extensively for detached shafts in the thhteenth century.1 4. Fragment with stiff leaf ornament, keyed for setting in a socket. Second quarter of thirteenth century. This fragment is of particular interest for the development of this style in Kent, and Dr. Pamela Tudor-Craig has kindly written the following report. 1 Cf. G. Dru Drury, " The Use of Purbeck Marble in Medieval Times," Proc. Dorset Arch. Soc, (1948), 70, p. 74 ff. 156 PLATE I INCHES\ Snargate Street, Dover, Garderobe. Architectural fragments. [faee p. 156 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER THE DOVEE FRAGMENT OF STIEE LEAE By PAMELA TUDOR-CKAIG, Ph.D. The history of stiff leaf in Kent during the first half of the thhteenth century could be written for the most part in terms of a gradual assimilation of continental elements introduced at Canterbury into a flat late twelfth century convention. The choh of Rochester Cathedral is the chief of few exceptions to this rule. Hence the importance of the discovery in Dover of a fragment that belongs to a national rather than regional context. Two styhstic features of this smaU portion of stiff leaf (Plate I, 4) are informative: the shape of the trefoil in which it terminates, and the presence of two further pahs of lobes springing from the stalk or midrib. The dexter lobe of the small trefoil has faint indications of a protrusion in its central area, and a double-arc curve to its outer rim. SmaU leaves with embossed lobes and broken rims are characteristic of the period immediately preceding the commencement in 1245 of the rebuhding of Westminster Abbey. They occur at York, Lincoln, and SouthweU, and most prominently in the Extension added by Bishop Hugh de Northwold to the east end of Ely Cathedral between 1234 and 1252. The earher High Vaulting Shafts of Northwold's Extension also use the moth of long straight mid-ribs with lobes proceeding laterally from them, but the popularity of that element was waning. In the second phase of the building they were abandoned in favour of the richer effect of layers of superimposed crotchets. The treatment of corbel cones by covering them with long vertical stalks bearing horizontal lobes had been current in a group of south coast churches1 as early as 1215-20. So the Dover fragment combines the small embossed leaf of c. 1235-45 with the subsidiary lobes of c. 1215-40, giving it an approximate styhstic date of 1235-40. However, a thhd field for foliage sculpture was being explored during the same decade, 1235-45. From tentative beginnings in the niche heads of Wells West Front, the idea of allowing stiff leaves to grow directly out of the mouldings developed, a symptom of the blurring of distinctions between architecture and sculpture which led ultimately to the Decorated Style. At Ely, tufts of stiff leaf adorn cusp-points in the arcade spandrels, but examples of the apphcation of the principal to arch springers, in the manner probably used at Dover, can be seen in the Gloucester " Reliquary," and in a fragment now embedded in the wall of the slype of St. Albans Abbey. There the treatment of the leaves suggests a date of c. 1260, but the general shape 1 Boxgrove Priory (corbels in the angles of the central tower), New Shoreham ohoir, and the Holy Sepulchre Chapel, Winchester Cathedral. 157 MEDIEVAL GARDEROBE IN SNARGATE STREET, DOVER of the moth, and the strong mid-rib continuing the architectural line, bespeak a common purpose with the Dover fragment. This strong mid-rib in the Dover piece passes through three stages. At the trefoil end of the stone it is carefully worked to a slightly angular section. The rehef is then abruptly broken off, and only the outline continues across the body of the stone, at the further end of which the angle alters through a few degrees, and the surface of the sthl flat rib is smoother. Corresponding differences in degrees of finish exist between the trefoil and the first pah of lobes, and between them and the second pah. It is possible that the stone has been damaged, but more hkely that its carving was never completed. As the morsel of stone stands now, it would be incapable of containing the motif outlined upon it without alteration of the angle of the trefoil. A simhar incompatibility between stone and design occurs on another unfinished fragment of stiff leaf, part of a smah Purbeck marble capital found at Corfe. If the sculpture is unfinished the unsightly prominence of the drhl holes on the Dover stiff leaf are excused1; the scratched-in rib on one of the side lobes and the faintness of the protrusion on the trefoil are thereby explained, and the presence of the fragment in a pit since about 1300 understandable. 1 Drill holes appear in Winchester Cathedral capitals, early thirteenth century, and Lincoln nave, c. 1240. 158