Keyhole Report for K63

Faversham Society Archaeological Research Group
Understanding Ospringe OSP09 Report for Keyhole K63 & K63T
9 Ospringe Street, Ospringe, Faversham, Kent
Grid Reference: TR 00431 60829
1. Introduction
Number 9 Ospringe Street offered an unusual opportunity to the FSARG team. Jason, the householder had just finished installing a rear extension, which involved cutting into the steeply upward sloping rear garden. Now he had to create a soakaway and linking trench to drain water overflow from the extension. This gave us the chance to drive a 0.5 metre wide trench into the garden for around 4 metres, at right angles to Ospringe Street. At the end furthest from the house (the south end) the trench was to be expanded to create a one metre square keyhole for the soakaway pit. Number 9 is an 18th century cottage¹, facing directly onto Ospringe Street, and opposite the point where until the mid 19th century the road to Faversham joined Ospringe Street². Ospringe Street is the designation of the A2 trunk road as it passes through Ospringe. The A2 is assumed to follow the course of Roman Watling Street as it runs between London and Dover. Number 9 is around 400 metres east of the surviving buildings of the Hospital of St Mary of Ospringe, a staying-over place for pilgrims on the Becket pilgrimages³. Although a great many archaeological investigations have taken place to the west of Ospringe (mainly on Romano British sites⁴) and some intensive excavations on the site of the Hospital of St Mary⁵, this south eastern part of the village had never before been investigated.
2. Location of excavations
The location of the trench and pit were determined by the needs of the soakaway, but were ideal for the investigation. The trench started at the level of the extension and drove at right angles into the bank. The keyhole was at the southern end of the trench, with its western edge in line with the trench, but extending a further 0.5 metres to the east. The top of the keyhole pit K63 was at a height of 11.5 metres OD but after about 2 metres northward the land sloped down sharply and had been severely truncated at the south end next to the new extension, which was the northern end of the trench. The total length of trench and keyhole was 5 metres. The OD at the base of the trench, equal to the level of the floor of the extension, was 9.4 metres.

Fig 1a) K63/T in 1865⁶	Fig 1b) K63/T in 1907⁷

Fig 1c) K63/T in 2008⁸	Fig 1d) End of excavation showing relationship of site to house. The soakaway pipe is in place Jason and Pat
3. The procedures
The outline of the trench and keyhole were carefully measured out and marked with string. The level was set up to give heights OD, using a temporary bench mark, and a grid set up for the excavation. The trench and keyhole pit were excavated in parallel, with the trench working southwards from the lower northern end. A narrow baulk was left between the two, not removed until the last stage of excavation. Single context method was used for both sites using different sequences of numbers at this stage to avoid confusion. These numbers were matched up and merged at the post excavation stage. (See Appendix 1 for details) The keyhole was excavated to the maximum depth of 1.2 metres and the trench dug to match this depth. All spoil was sieved meticulously and the spoil heap scanned using a metal detector. Finds were set aside for each context and special finds given three dimensional coordinates to pinpoint the exact find spot. Features revealed were carefully recorded, with scale section and plans to show the relationship between contexts. Finally, once the maximum Health and Safety limit for FSARG excavation had been reached, the householder swiftly lowered the levels of both trench and pit further, with archaeologists watching.
4. The findings
The team had been told of a concrete path running southwards from the house through the garden, now covered by topsoil [01]. This path [02] was uncovered along with a same-age flowerbed [103]. On removal of the path, a well preserved brick path [04] was revealed. The shape and texture of the bricks (small, hard, red, often loaf shaped and without frogs) suggests an 18th/ early 19th century date for the brick path.
Fig 2: The brick path [04] and a typical brick

The path was set into a garden soil [03] which covered the whole of the area under investigation except in the truncated northern section. [03] was an ashy grey brown soil with well distributed fragments of pottery, coal and clinker, small brick and tiles fragments, bone, shell and clay pipe stem fragments. Although the pottery was dominated by 19th- early 20th century items, the amounts were not large and small sherds of early medieval and late medieval pottery were mixed in. The layer [03] shaded into a yellow brown soil with far less ash content [05], and this also extended across the area of interest, except for where the following interventions had taken place: • Into [03] in the area of the baulk left between the trench and the pit had been sunk a large pit containing demolition materials such as brick, tile and slate [06] [07], contemporary with [03] • There was also a possible post hole [114] [113] here, also contemporary with [03]. • A larger and deeper pit [14] [20] appeared to subsume the first two interventions, therefore preceding them in time.

Fig 3: Left: rubbish pits into [03] in K63 Right: bottom of the large rubbish pit [14] [20] with near complete peg tile, in K63T
The yellow-brown layer [05] had a high content of medieval pottery, the proportion of which increased as excavation progressed downwards, without there being any clear context differentiation. As with [03] the upper part was a well mixed soil, implying regular digging over. At the northern end, the [05] layer interfaced with a feature made of large flints [10] of which more later. In the northern part of the trench, a thin layer of chalky fragments [11] separated [05] from the deposits beneath, although it did not extend over the flint feature. In the pit, a rubbish pit complex going into the next layer down as well as the pit and post hole interventions coming down from [03] had disrupted any possible chalk layer. Beneath and around the rubbish pits in K63 was an orangey brown clay deposit, brick earth with soil mixed in. [115] This contained only a few worked flints, a small abraded sherd of medieval pottery, an iron nail and some oyster shell. The deposit [115] shaded down into artefact free orange brick earth, assumed to be the natural soil [16] at a depth of 1 metre. Further rapid excavation down a metre by the householder confirmed that this was a completely artefact free zone and therefore confirmed as the natural soil. The underlying Upper Chalk was not reached even at this stage. Moving northwards along the trench, the dominant features were a large rubbish pit [14] [20] at the southern end of the trench, overlapping slightly into the keyhole pit, and a 'ditch' [12] [15] across the middle of the trench running east to west. The 'ditch' section was overlain by the thin chalk layer [11]. The base of the rubbish pit lay within the range of the excavation but the cut for the 'ditch' [15] ran beyond the base of the trench. It appeared that the rubbish pit cut the 'ditch', making the 'ditch' earlier in date. Both pit and the 'ditch' contained small amounts of medieval pottery but the rubbish pit had a few sherds of 19th century pot, clay pipe stem and a near complete peg tile at its base. Both fills, however, were free of coal and cinder, unlike the upper layer [03]. The flint feature [10] at the northern end of the trench is best understood with reference to Appendix 2 (section, plan) and figs 4 and 5.

Fig 4: Flint feature [10] with gully [13] to the left. [13] contained a lot of medieval pottery.
It had obviously been severely truncated by work in the garden. The flints were roughly shaped by having the ends knocked off. In section, the flint assemblage forms a distinctive lens shape, tapering out southwards to a single layer of flints which are underlain by a layer with lenses of chalk [17]. This layer curves upwards to the surface just to the south of the flints, forming a hump across the trench. Between this hump and the main flint area to the north was a dip which was full of medieval pottery, bones and oyster shell. This dip was given the context label of [13] but is best seen as the lowest element of [05]. The hump was cut to the south by the cut [15] for the 'ditch', thus is earlier than the 'ditch'. The flint feature was 44 cm thick at its deepest but was almost certainly widening northwards before truncation.

Fig 5: Flint feature in section after the trench had been deepened to accommodate the soakaway pipe.
When the flint feature and the orange-brown layer with chalk lenses were removed, they were shown to be bedded on orange-brown clay deposits which were artefact free and seen as the natural brick earth deposit [16]. This trench K63T and pit K63 produced more medieval pottery than the rest of our Understanding Ospringe pits put together, see Fig 7⁹. This medieval pottery assemblage comprised mainly of large fresh-edged sherds and is a significant assemblage which will be catalogued in detail¹⁰. There were also early medieval shelly ware sherds, dating from around AD1100-1200 and quite a lot of late medieval (AD1400-1550) but pottery from the main medieval period was dominant. Most of it came from context [05] and associated contexts such as [13]. Examples of thumbed bases, glazed decoration and incised decoration were all found. Most of the assemblage was made up of pottery made a the Tyler Hill kilns up in Blean Wood¹¹, between Faversham and Canterbury, but a number of non-local wares, e.g. London Highly Decorated Ware, are represented.

Fig. 6: Pottery assemblage from context [13]. Mostly medieval Tyler Hill wares

Total weight = 3076 gms. Most of the K63/63T medieval pottery came from the trench. Because of the narrowness of the trench, the volume excavated is comparable to the other excavations.
Fig 7: Medieval pottery amounts by weight found by FSARG in the Understanding Ospringe project, OSP08 and OSP09.
5. Interpretation
Because of the large scale pits and ditches, this has been a difficult pit to sequence and phase, with the dating of the flint feature uncertain, as is explained below. A proposed Harris Matrix for this excavation is shown in Appendix 1, but is arguable. Beneath the top post medieval and modern layer (40cm deep) and the rubbish pits associated with this stage of garden use, the contents of the deposits were almost entirely medieval in date. [05] had abundant medieval pottery which because of its unabraded nature and relatively large size must come from a rubbish pit context rather than from a midden scatter across fields, such as we have found in other local small scale excavations¹². A substantial amount of this pottery (Fig 6) was found in a context in contact with the flint feature [10] - in effect, lying on the surface of [10], along with animal bone and shell. So what is [10]? The curious profile of this feature can be seen in Fig 5 and Appendix 2 (section). It seems far too thick to be a courtyard surface, yet too lens shaped to be a wall foundation. The tidiness of the flint arrangements and the underlying chalk/clay belt negated ideas of a collapsed wall. In fact, the flint feature looks very like the extreme edge of a Roman road, such as has been found at nearby Syndale, also running 20m south of the present A2¹³. This interpretation was enhanced by the presence of what seems to be a gully running east to west beside it to the south (occupied by our context [13]), then a chalk bank and a steep sided ditch beyond. The surface of the flint feature is, however, in direct contact with medieval deposits [in contexts 05 and 13]. Furthermore, only small residual amounts of Roman pottery were found in this excavation, as have been found in most of our Ospringe keyholes, and none were found in the 'ditch'. What does seem to be definite is that the flint feature was exposed during the medieval period and that a bank sloped gently upwards on which occupational rubbish was thrown. If the flint feature curved upwards, then the medieval pottery can be seen as accumulating in a gully alongside it to the south. The robust structure and materials of the 'road' imply that this was Watling Street renovated and reused, either as a road during the medieval period or perhaps as a courtyard surface behind a medieval property facing onto the main medieval road in the same location as the current A2. This may be over interpretation of a small and heavily truncated feature, but does explain the characteristics observed: only examination of neighbouring plots would take this further. One thing does seem more certain, though, and that is that here we have a newly discovered site of medieval activity in Ospringe.
6. Final comments
This was a challenging excavation for FSARG members, given the narrowness of the trench, which limited access and visibility, and the truncation of the northern end. Not only did we encounter important archaeology, we also needed to develop new skills to do it justice, not least how to work effectively at the bottom of a 0.5 metre wide trench. In short, a very worthwhile project from a number of viewpoints.
	Fig 8: Keith and Caroline digging a 0.5 metre wide trench.
7. Acknowledgements
Great thanks to Jason for giving us the opportunity to carry out this excavation, a rare chance to do something beyond the usual 1.5 x 1 metre keyhole.
¹Swale Borough Council c1990 Townscape Survey: Ospringe Village ²See Mudge map (1801) for Ospringe, CKS Local Studies KCC: Centre for Kentish Studies. Maidstone. ³KCC Historic Environment Record No TR06 SW15 ⁴e.g. Whiting, W 1925, 'A Roman Cemetery discovered at Ospringe in 1920' Arch. Cant. XXXV p 1-16 XCV p81-184 ⁵e.g. Smith, G.H. 1980 'Excavation of the Hospital of St Mary, Ospringe, commonly known as the Maison Dieu' Arch. Cant. Vol. XCV p81-184 ⁶OS 1865 Sheet XXXIV Scale 1: 2500 ⁷OS 1907 Sheet XXXIV Scale 1: 2500 ⁸OS 2008 Serial number 00332800 Scale 1: 1250 ⁹Reid, P 2009 Lecture given at the Maison Dieu, as the Annual Lecture, September 2009. Unpub. ¹⁰Reid, P & M.Spouse. Medieval Pottery Catalogue for K63 and K63T. Forthcoming ¹¹Cotter, John 2002 'A medieval kiln site at Daw's Wood, Tyler Hill, near Canterbury' Annual Report for 2000-2002, Canterbury Archaeology Trust, Part Three, Section 2 ¹²See, for example, the Report for K54, p 4 on the FSARG website: www.community archaeology.org.uk / projects / understanding ospringe/ keyhole reports/k54 ¹³Whiting, W, Hawley, T & T May 1931 'Report on the excavation of the Roman Cemetery at Ospringe' Antiq. viii, Oxford
Appendix 1: Harris Matrix for Keyhole 63 and Trench 63T
This keyhole and trench were actually the same excavation but excavated in parallel and only joined up towards the end of the excavation. Thus they had different context number sequences. On post excavation examination, the following adjustments were made. Also, some contexts were merged internally. Merge: 100/101 with 01 ; 102 with 02; 109 with 04; 105/ 106/108 with 03; 107 with 06; 118 with 07; 110 with 05/09; 116 with 21 and 16 The number shown in bold is the one used in the Harris Matrix. Records and archives have been left with the original numbers.

Appendix 2:
a) Scale section along the east facing side of the whole trench, including the keyhole section. original scale 1: 20
→N
For key, see (b)

b) Close up section and plan of the northern end of K63T, showing the flint feature [10]. Original scale 1: 10 In the plan, the context [13] shown on the section has been removed, leaving a gully. [13] contained a lot of medieval pottery.