The Medieval Hospital of St Mary The Blessed Virgin, Ospringe (Maison Dieu): Further Details of its original layout revealed by excavations at The Fairways

ArticlesVolume 131

25 Jun

Written By KAS

The Medieval Hospital of St Mary the Blessed Virgin, Ospringe (Maison Dieu): further details of its original layout revealed by excavations at THE Fairways

andrew margetts

with contributions by

Luke Barber, Sarah Porteus and Sheila Sweetinburgh

Historical Background by Sheila Sweetinburgh

The Hospital of St Mary, Ospringe, on the main pilgrim route to Canterbury, was probably founded in the late twelfth century but it was Henry III’s particular interest in this institution over four decades until his death in 1272 which underpinned its considerable expansion at this period. Henry’s patronage may have encouraged a profusion of local benefactors who granted land or property on or adjacent to the precincts while others donated holdings in Ospringe Street (Sweetinburgh 2006; forthcoming). However, it is also likely that family and neighbourly connections among the local population aided the flow of benefactions to St Mary’s. Thus the hospital grew into a considerable complex of buildings on the north side of Watling Street together with various individual buildings and holdings on the south (see, for example, the map in Smith 1979, 85). As well as its interests in the Faversham and Ospringe area, it had acquired considerable income from lands granted in Staple and Wingham during its early history.

As patron Henry was concerned about the hospital’s charitable role, for the poor and pilgrims, but also in terms of its intercessory activities. The provision of a chapel was an essential prerequisite. Notwithstanding its importance, the building of a chapel had potentially serious implications for the diocese because it was outside the parochial system, necessitating a negotiated agreement between the holder of the advowson of the local parish and the hospital authorities, or its patron. In 1235 the abbot of St Augustine’s Abbey, Canterbury, who held the parish church of Faversham, did agree to the building of a chapel at St Mary’s Hospital.

Despite its very promising beginnings St Mary’s suffered varying degrees of financial hardship during the fourteenth century, the main contributory factor being the king’s incessant demands on its hospitality for himself, his household and for retired retainers placed in its care (corrodians). The Black Death (1349-50), which struck Ospringe partic-ularly hard, no doubt caused much disorganisation. The hospital also suffered considerable mismanagement under various unsatisfactory Masters over the years. In 1516 Henry VIII deemed it to be a suitable asset to hand over to his new College of St John the Evangelist in Cambridge and thus the hospital lost both its status and the bulk of its revenues. Its deterioration was rapid and by the time A Survey of Kentish Estates was compiled, c.1571, the complex was in a ruinous condition apart from the still-standing chapel (Smith 1979, 84-5), the last resident of the ‘spetill house’ having been a hermit.

Results of the Archaeological Investigations at the Fairways

An archaeological watching brief was undertaken during works to install a pumping station and attendant pipe run on land at Fairways, Ospringe Street (Fig. 1). The work comprised the archaeological monitoring of an open cut trench to accommodate new services, as well as excavation of two large pits for the new pumping station. Monitoring took place in October 2007 during which time the investigation succeeded in identifying significant masonry remains associated with the medieval hospital.

The excavations removed approximately 0.5m of overburden and revealed a modern concrete culvert before the archaeological horizon was encountered. The latest archaeological remains comprised a disturbed demolition layer dating from the fifteenth to early sixteenth century with a substantial quantity of the material deriving from an episode of roof collapse.

Beneath this demolition layer, orientated roughly east-west and approx-imately 6m from Ospringe Street, a length of wall and associated footing was exposed (Wall A, Figs 2 and 3, sections 3, 4). The north facing elevation of this wall was neatly faced with complete, or nearly complete, un-weathered Upper Greensand ashlar blocks (Plate I), usually with oblique or horizontal tooling on their surfaces. These were laid in rough courses, bonded by an off-white crushed shell (cockle) lime mortar suggesting that all were from a single building phase. Evidence of an architectural feature was indicated in the north face by the presence of chamfered Greensand blocks. The possible remains of a stepped footing also existed at the base of this wall. The southern elevation, by comparison, was comprised of roughly faced flint and Kentish rag, although the original façade on this side may have been removed or robbed (Plate II). The upstanding wall was built upon a roughly faced flint spread footing and had been cut by a modern service that had served to detach a chamfered block towards the eastern baulk.

Against this baulk and at right angle to Wall A was a large piece of chalk ashlar and associated roughly faced flint Wall B. This masonry, orientated north-south, was interpreted as the remnant of a wall probably destroyed during the laying of a nearby BT service (Wall B, Plate II, Figs. 2 and 3, section 3) and was most likely a remnant of the masonry found by S.E. Rigold in 1957 (Rigold 1964, fig. 1). Rigold was unable, during the work that he undertook, to examine the western face of this masonry and it proved the same during this attempt due to truncation by the service.

The eastern face of Wall B recorded by Rigold bears a striking resemblance to the construction of the northern face of Wall A. Rigold interpreted his masonry as a bridge abutment as it was obviously intended for a ‘massive substructure’ and he assumed that the stream ran adjacent to it; however it was clear during this phase of excavation that the stream ran further to the west and was culverted through the hospital complex.

It now seems more likely that Walls A and B represent part of the base of a Bell Tower or steeple as mentioned in the c.1571 Survey of Kentish Estates (transcription within Smith 1979): ‘There was once a Bell Tower called The Belfrey, but no walls remain’.

The Kent Archaeological Rescue Unit’s work adjacent to the site confirmed the location of the hospital chapel (finding its vaulted cellar) and noted that Rigold’s wall was of a different construction to that of the chapel remains (Parfitt 1990). This helps to substantiate the interpretation of the Walls A and B being related to a bell tower for they are in the correct location west of the chapel.

The third structural remain exposed was probably the most substantial. This comprised a length of up to 7m of roughly faced flint footing with indications of Greensand ashlar walling above (Wall C). This was orientated approximately north-south and seems almost certainly to be the remains of the eastern wall of the hospital Common Hall (Plate III, Figs. 2 and 3, section 1). G.H Smith excavated the northernmost bay of this building during 1977 and suggested from his findings that St Mary’s hall was orientated north-south (Smith 1979). He found this impossible to prove but it seems very likely that this recent work in conjunction with recent geophysical survey results from Barkaway’s Paddock (FSARG 2008) confirms his theory that the hospital’s ground plan was similar to St John’s, Canterbury, and St Mary’s, Strood, and unlike the usual east-west orientation found at most other sites.

Built into the wall of the Common Hall was a feature that seemed to be largely constructed of tile. This feature was only partially revealed but it seems probable that this represents the location of a hearth that served, and was a later addition (probably sixteenth-century) to, the Common Hall. Without proper exposure it is difficult to say exactly what form this hearth took but it seems likely that it was similar to those found by Smith in the 1970s (Smith 1979).

Immediately adjacent to the Common Hall’s wall footing (Wall C) ran a considerable length of largely intact (though truncated to the south) medieval culvert. Well constructed in Greensand ashlar, this culvert was preserved in excellent condition and is a stretch of the largely fragmentary one found by Smith during his excavations to the north (Smith 1979).

In addition to the walls, culvert and hearth, the watching brief identified the location of a probable floor level in the area of the new pumping station (Figs. 2 and 3, section 5). This possible surface seems to be in the right location for the Common Hall’s interior. However Smith noted that the northern bay he excavated had a compacted clay floor as opposed to compacted flint. Comparable cobbling was encountered within the chapel remains by KARU in 1989 (Parfitt 1990) and it seems that the interior of the Common Hall was similarly covered with cobbling probably relating to a post-dissolution yard surface.

The Finds

The Pottery by Luke Barber

The only medieval pottery from the site was recovered from the medieval demolition layer. All of these sherds are unabraded suggesting they have not been reworked by later activity despite a little intrusive nineteenth century material being recovered from this deposit. The majority (22/170g) are from a single small barrel-shaped jug with ‘apron’ of green glaze in Cheam whiteware typical of the mid fifteenth to very early sixteenth century (Pearce and Vince 1988). A further sherd (1/36g) from a heavier whiteware jug with spots of green glaze and notable black iron pellet inclusions may be from a fourteenth- to early fifteenth-century Kingston-made vessel. The only other vessel represented (2/30g) consists of a further jug in an oxidized sandy ware with reduced surfaces likely to be from the late Tyler Hill industry from north of Canterbury. A later fourteenth- to mid/late fifteenth-century date is probable for this vessel.

All other pieces are of nineteenth-century date and include two blue transfer printed pearlware base fragments from a water jug retrieved from the overburden, fragments from a yellow ware bowl and a fragment of an iron-washed salt-glazed English stoneware bottle (i.e. ink) found in the disturbed fill of the medieval culvert, as well as intrusive glazed red earthenware sherds from the demolition layer.

The Stone by Luke Barber

The archaeological work recovered a large quantity of stone, most of which consisted of worked building material of medieval date. Full details are housed with the archive. A chip of Kentish Ragstone was recovered from Wall A and in unstratified deposits. The latter consisted of a massive roughly faced block (48.5kg) measuring 280 x 210 x 580+mm. The remainder of the medieval stone consists of complete or fragmentary building blocks in Upper Greensand. The largest group came from the facing of Wall A where a number of complete/nearly complete unweathered ashlar blocks are present, usually with oblique or horizontal tooling on their surfaces. Complete measurements include 350 x 225+ x 170mm, 240 x 140 x 90mm, 180 x 130 x 70mm and 205 x 135 x 60mm. Only one architectural fragment is present, a chamfered plinth of undiagnostic date, some 126mm+ high, with horizontal and oblique tooling on its worked faces. A few pieces of roughly faced Upper Greensand building blocks are also present in this context. A number of the blocks/block fragments have traces of an off-white crushed shell (cockle) lime mortar adhering suggesting all are from a single building phase. Further smaller fragments from similar blocks were recovered from the medieval demolition layer. The source of this material is almost certainly the Reigate area of Surrey. Reigate stone became increasingly common in this part of Kent from the mid/late twelfth century being used in the nearby churches of Luddenham and Davington as well as at Faversham Abbey (Tatton-Brown 2001; Philp 1968). The stone probably reached the site from stock-piled material at Battersea, via the Thames and Faversham Creek, and was certainly a common freestone used in the area during the thirteenth to fifteenth centuries, after which is was used mainly for internal work due to its poor weathering (Tatton-Brown 2001).

The Ceramic Building Material by Sarah Porteus

Ceramic building material (CBM) was recovered from two contexts: the demolition layer and the tile-built hearth and included glazed mono- and polychrome floor and roof tile of medieval date and post-medieval brick and tile.

A total of ten fragments of glazed Medieval floor tile were recovered from the demolition layer, at least one of which appears to have been reused. All the floor tiles are characterised by knife cut bevelled edges and un-keyed bases. The floor tiles are comparable with those from the 1977 excavations (Smith 1979). The floor tile can be divided into four fabric types:

Fabric 1: the first fabric type is Tyler Hill (group 1 from the 1977 excavations) and probably originates from Tyler Hill near Canterbury. Decorated tiles impressed with slip and plain glazed tiles in this fabric are probably of thirteenth- or fourteenth-century date. The decorated tiles have a central floral design within a double outer circle with a small, possibly trefoil, corner design (Fig. 4, nos. 1 and 2). The design directly corresponds to design ‘J’ from the 1977 excavations in both style and dimensions. Where size could be established the floral tiles were 113mm square by 22mm thickness. The plain tiles have a green glazed appearance and are badly abraded with complete dimensions of 112mm square by 21mm thickness.

Fabric 2: the second fabric type was coarse with abundant medium sized quartz and black grains of glauconite(?) with sparse calcareous inclusions up to 2.5mm. The design is badly abraded but appears to be a diamond lattice pattern (Fig. 4, no. 4). No parallel was found for the design and the tile was smaller in size than the other tiles with a complete width of 61mm and a thickness of 17mm, the tile appeared to have a rectangular shape with slight flaring of width. The tile is probably of thirteenth- to fifteenth-century date.

Fabric 3: the third fabric group is made of a fine sandy fabric with moderate to abundant fine quartz and fine cream silt banding. Of the three tiles identified in this material one is definitely of decorated polychrome style, the other two examples are abraded though have slight impressions and slip traces which indicate patterning. Where patterning was visible insufficient remained to reconstruct the design (Fig. 4, no. 3). Tiles in this fabric were of uniform thickness of 20mm and only one complete breadth of 122mm could be measured. The origin of the fabric is uncertain.

Fabric 4: the fourth fabric group, made of a pale pinkish cream silty fabric with abundant fine sand and orange iron rich clay inclusions. Group four was represented by a single reused tile with rich green glaze of 20mm thickness. The tile is unlikely to be from a local source and may be Flemish in origin and of fifteenth/sixteenth-century date.

A majority of the early roof tile recovered from the demolition layer is peg tile made of the same Tyler Hill fabric as the type 1 floor tiles. The earliest peg tile is in Tyler Hill fabric with some splash glazing suggesting a twelfth to fourteenth-century date. Later unglazed peg tile, probably dating to the fifteenth to seventeenth century is in two fabric types which may be local variants of the Tyler Hill fabric. One variant has a scatter of red iron rich silt inclusions in addition to the coarse angular quartz, the second variant has sparse inclusions of the coarse angular quartz with sparse coarse calcareous inclusions; tile made in this fabric are less coarse but over-fired and brittle suggesting they are less well made.

All other later roofing tile recovered dates from between the sixteenth and nineteenth centuries, and the fabrics see a move away from the Tyler Hill production to other local tile production centres. A single fragment of yellowish pale cream fabric with sparse pinksh red iron rich silt and very sparse coarse quartz (CAT37), and fragments of silty fabric with moderate calcareous speckling and fine cream silt banding (CAT32) were identified. Peg tile also appears to be late medieval or post-medieval in date, most likely sixteenth to eighteenth century, and are mortared together using a loose white lime mortar. The tiles are of a fine orange fabric with a scatter of fine mica and sparse red iron rich silt inclusions with fine black sand.

Other post-medieval ceramic building material recovered included a single example of a mathematical tile. Mathematical tiles are used as a form of cladding for the exterior of buildings which give the impression of the wall being made of solid brick. These tiles are common only to the south-east of England, particularly Sussex and Kent and provide excellent weatherproofing and give the impression of high quality brickwork. Often used for the façade of buildings, mathematical tiles were both more expensive to make and tax than brick. The earliest known examples date from 1725 and the fashion for mathematical tiles dies out around 1850 (Brunskill 1997).

Also recovered was a fragment of nibbed pantile in a fine orange fabric with fine quartz and a scatter of mica and fine black sand dating to the later seventeenth to nineteenth century, a nineteenth/twentieth-century drain fragment and a fragment of vitrified brick of uncertain date.

The medieval tiles correspond to those found in earlier excavations on the site and are representative of tiles used in monastic buildings and some parish churches in the thirteenth and fourteenth century. The decorated tiles almost certainly originated from the buildings of St Mary’s hospital. The probable imported plain glazed Flemish tile is of slightly later fifteenth/sixteenth-century date and may relate to later repair of the buildings.

Conclusion

The findings of this investigation make a small but important contribution to the body of knowledge about the Hospital of St Mary at Ospringe. The wall remnants located in this work clearly relate spatially to elements uncovered during previous excavations and by comparing evidence, such as construction style and materials used, it has been possible to determine or suggest which part of the hospital complex the encountered structures relate to. Perhaps most importantly, the work provides further insight into the hitherto un-investigated southern half of the Common Hall as well as suggesting the location of the Belfry. The work also further confirms that the hospital’s ground plan was similar to other early or contemporary Kentish hospitals. As such these may form a regional group, although it seems just as likely, especially for parts of the southern half of the complex, that the layout of St Mary’s was dictated by both the stream and the line of Watling Street.

acknowledgements

Archaeology South-East would like to thank 4Delivery Limited who commissioned the work on behalf of Southern Water Ltd. Thanks are also due to Pat Reid of the Faversham Society Archaeological Research Group (FSARG) and the Heritage Conservation Group, Kent County Council, for their guidance throughout the project. The project was managed by Neil Griffin (fieldwork) and Louise Rayner (post-excavation). The illustrations were produced by Hannah Faux, Fiona Griffin and Justin Russell.

The detailed site and research archive is currently located with Arch-aeology South-East prior to deposition with a suitable local repository.

bibliography

Brunskill, R.W., 1997, Brick Building in Britain, London: Cassell Group.

FSARG 2008, Resistance Survey Report for Barkaway’s Ospringe, Report no. 003 (unpubl.).

FSARG website www.community-archaeology.org.uk: Projects/Understanding Ospringe/Keyhole Pits/K62: a report on the small-scale 2009 excavation in the garden of the Ship Inn (next to the Fairways) which uncovered a wall stretch.

Parfitt, K., 1990, ‘Archaeological Excavation and Recording at Nos. 14-18, The Street, Ospringe, 1990’, Kent Arch. Review, 100, 228-229.

Pearce, J. and Vince, A., 1988, A Dated Type-Series of London Medieval Pottery: Part 4: Surrey Whitewares, LAMAS, Special Paper No. 10.

Philp, B., 1990, 14-18, The Street, Ospringe, near Faversham, Kent Minor Sites Series, No. 2.

Philp, B., 1968, Excavations at Faversham 1965, KARGC.

Rigold, S.E., 1964, ‘Two Kentish Hospitals re-examined, St Mary, Ospringe and SS. Stephen and Thomas, New Romney’, Archaeologia Cantiana, lxxix, 31-69.

Smith, G.H., 1979, ‘The Excavation of the Hospital of St. Mary of Ospringe commonly called Maison Dieu’, Archaeologia Cantiana, xcv, 81-184.

Sweetinburgh, S., 2006, ‘Royal Patrons and Local Benefactors: the Experience of the Hospitals at Ospringe and Dover in the Thirteenth Century’, in Religious and Laity in Western Europe, 1000-1400: Interaction, Negotiation and Power, eds E. Jamroziak and J. Burton, Turnhout, 111-129.

Sweetinburgh, S., 2010, ‘The Hospitals of Medieval Kent’, in Later Medieval Kent, Kent History Project, Boydell Press.

Tatton-Brown, T., 2001, ‘The quarrying and distribution of Reigate stone in the Middle Ages’, Med. Arch, 45, 189-201.