Page content begins after here
Founded 1955
Excavations at Elsyng Palace 2007 (site code ENF07) – an update by Martin J. Dearne wth a contribution by Neil Pinchbeck
In Bulletin 186 (September 2007) the author outlined the work the EAS has been doing since 2005 on the scheduled ancient monument of Elsyng Palace in the grounds of Forty Hall to help the borough get permission to plant new saplings in the gaps in the Lime tree avenue that runs across the site, and the archaeology that we as a result have recorded. However, as noted at the time, there were further excavations after that article was written and now that they are complete, as is the programme of sapling planting permitted by English Heritage, it is appropriate to provide a final update on the work (please see Bulletin 186 for the background to and location of the work and pit and trench code system used).
The Late Fifteenth Century Courtier's Palace, and Later Midden and Developments, in the Area of Pit 33
The significant extension of P33 within the outer courtyard of the (Tudor) palace by Trenches 3 and 4 reported in Bulletin 186 continued with the cutting of Trench 5 which allowed the full, three phase, story of this area to be reconstructed as follows.
Phase 1
Over the natural, with a west to south west sloping surface, a rammed pebble spread (? part of a surface leading from the gatehouse into the (in Tudor times outer) courtyard) with integral brick features had been lain, the mortar used indicating a late fifteenth century date. The pebble spread was known by the time of the last article but we had only glimpsed the brick features in Trench 4. Trench 5 showed them to include (Fig. 1) a north west to south east orientated c. 1.10 m square surface of handmade, hard fired, unfrogged orange bricks and part bricks, set on a bed of mortar, surrounded by a slightly upstanding border of similar bricks/part bricks on edge, and the localised use of chalk fragments for the pebble spread's surface. Clearly an integral construction with this was a brick built, rectangular sectioned, open topped drain, which ran for over 1.75 m south east from near the surface’s south eastern edge in Trench 5. The drain was overall 43 cm wide, had an internal width of 23.5 cm, an internal depth of 14 cm and fell extremely steeply at 1 in 4.7 to the south east with suggestions at the southern baulk that the fall increased. Its base was constructed of coursed, (mainly part) bricks (identical to those in the surface) bonded again with relatively soft, sandy yellow mortar. It would have been proud of the pebbled surface by c. 16 cm at the north end, but its rapid fall meant that it was at or below the surface of the natural before the southern end of Trench 5 and some form of later removed loose covering (such as roof tiles or wooden slats) must have been necessary to prevent its very rapid blockage.
It is tempting to see the surface as the base for a water cistern or tank with the drain to take away waste water, and one implication of the existence and nature of the drain is that it almost certainly fed into a larger drain at a deeper level. It must be strongly suspected that that was a continuation of the late fifteenth century arched brick drain excavated in 2005 and now gives us a better idea of its course. The chalk edging round the ?tank base might be entirely decorative but may also have facilitated the identification of the feature at night (the author is grateful to L. Pinchbeck for this suggestion).
Fig. 1: The Late Fifteenth Century ?Tank Base and Drain in Trench 5 with the Charcoal Layer at the Top of the Midden in Section Above the Base
Phase 2
Subsequently, as previously reported and now more fully defined, a midden (rubbish dump) around 3.6 – 3.8 m in diameter, came to occupy the roughly bowl shaped depression that had come to be worn into the pebbled surface, and to cover much of the brick ?tank base as well as a little of the already disused and clay silt filled drain. There is no reason not to assume the re-use of the pebbled surface in the Tudor palace arrangements and the consequent heavy wear given its period of use may well have created the depression in which the midden subsequently formed; whether the ?tank base and drain were still in use in Tudor times is unknown. The midden was bulk sampled for later environmental analysis (still ongoing) and had had building demolition material added to it, at least in one case certainly prior to the deposition of a charcoal capping (Fig. 2), which one suspects resulted from a bonfire on the site during palace demolition. The absence of wine bottle glass, clay pipe and Delft Ware tentatively suggested a (?later) sixteenth century date for the midden but the pottery recovered suggests that the deposit at least continued to accumulate into the first half of the seventeenth century. As the demolition material in it was from structures using late fifteenth century mortar an implication is that what structures were being demolished (? in the mid seventeenth century) in the area were as much late fifteenth century in origin as Tudor, and reinforces the impression that Henry VIII reused Lovell's courtier's palace wholesale as the basis for the outer courtyard of his own palace. At some point during the process of midden development though a new dump of rammed pebbles had been deposited over much of Trench 5 at least, probably to create a new surface/path skirting and perhaps c. 0.30 m above the level of the midden, though it had also sunken into the earlier drain and eventually the midden had encroached upon it as it continued to grow. At this time, probably by now out of royal hands, the palace is likely to have been in significant decline but the laying of this new surface suggests that its maintenance had not been abandoned entirely.
Phase 3
Overlaying the midden and the pebbled spreads was a demolition deposit, in some areas with concentrated crushed pinky white mortar and window glass, some of it painted (Bulletin 186, p 12 Nos. 5.1 – 5.11), belonging to the demolition of the palace c. 1657. Above it in some areas was a relatively sterile make up layer, then another layer of demolition material and it appeared likely that this final demolition dump represented, with the make up layer, the process of landscaping the area with a variety of demolition derived materials. Topsoil then overlay this.
Animal Bone (summarised from a specialist archive report by Neil Pinchbeck)
Although some summary of the general conclusions Neil Pinchbeck has been able to draw from the animal bone recovered during the project was given in the last article, attention has particularly been focused on the above midden’s collection and it is worth illustrating some of the information that it has provided. The two graphs below summarise the sort of analysis of the material undertaken (uniting the finds from the test pit and trenches):
Table 1: the faunal assemblage from the midden by species (fragment count)
As can be seen, though this is the number of fragments of bone recovered not the number of individual animals they represent, the midden included pig, chicken, rabbit and ovicaprid (sheep and goat) as well as much commoner cattle. There was evidence for marrow extraction from the latter two and a distinct preponderance of cattle rib bones even with allowance made for their greater fragmentation; but horse, rat, dog and a badger (Meles meles) were also present. The animal bone thus suggests that meat consumption, or at least processing, in the area of the site investigated was primarily of beef with lesser amounts of lamb/mutton/goat and pig, and some rabbit and chicken (possibly under represented because of the susceptibility of the small bones of these species to fragmentation and decay/complete consumption by dogs). The butchery evidence (the material including little indicative of 'table scraps') indicates the regular extraction of marrow from all larger species (especially cattle) and analysis of the ungulate bones indicates a preponderance amongst long bones of metacarpals and metatarsals which may relate to the removal of low meat yield elements of carcases and their retention in a kitchen context (?for stock production). The presence of proximal and intermediate phalanges but no distal phalanges was also noted and may indicate the arrival on site of carcases which had been skinned, leaving the hoofs on, for tanning.
Table 2: the skeletal element (fragment count except ribs by min. number)
The presence of non food domesticates (dog and horse) and of rat is unsurprising in a palace context, especially in the former cases as it was adjacent to a chase, though the presence of badger in a stratified and not apparently animal disturbed midden context is interesting and might relate to the utilisation of its bristles.
Also interesting is that the features of some finds made specific deductions about the size and type of animals represented possible. In the case of three equine (horse) scapulae from the midden which were sufficiently complete to reconstruct, giving an overall length of 43.2 cm (17"), it was possible to say that the animals had originally stood 13.81 hands (1.4 m) which would have made them slightly too large to be considered as members of most breeds of modern pony (max. approx. 13.5 hands), but too small to be considered as a lady’s hunter (min. 15 hands). Rather, they would fall within the size range considered as cobs (average 13.7 hands), typified by the modern Welsh cob.
P36/Trench 6 – Further Late Fifteenth Century Evidence
The expansion (as Trench 6) of a pit (P36) in which previously another rammed pebble courtyard surface had been seen some way north of that discussed above also emphasised how far late fifteenth century features had been left intact and reused in the Tudor palace. The pebbled surface was badly damaged by tree roots and a modern planting pit but was up to 70 cm thick. The significant finding though was that it abutted the tree root and demolition disturbed remains of the same mortared brick surface seen nearby previously in pits P38 and P38A (Bulletin 186, p 8 and Fig. 3). However, it was apparent that the traces of white Tudor mortar seen then should be reinterpreted as repair work because this new larger exposure clearly showed that the original mortar used between and under the bricks was typical of the late fifteenth century. The surface had presumably just been retained and where necessary repaired in the Tudor period. Moreover, that the brick surface was so extensive and abutted a probably contemporary, certainly external pebbled surface means that it too was external and not the floor of the front room of a ?tower’s basement as previously suggested from geophysical evidence. Indeed, geophysical evidence on the site is increasingly being shown to be more difficult to interpret than it initially appears and whilst the ?tower interpretation could still be correct the presumed size of this structure must now be reduced and a large (?9 m wide) brick surface be envisaged in front of it at the north side of the (in Tudor times outer) courtyard.
Overlying the brick surface (and filling slumped and missing areas of it) and the damaged areas of the pebbled surface where not removed by the modern intrusion was a layer of demolition rubble including bricks notably retaining late fifteenth century soft, sandy yellow mortar adhering to one face and a hard white (Tudor) mortar to the other; further evidence for Tudor repair or additions to existing structures rather than their replacement.
Fig. 2: The Damaged Brick Surface with Remnants of the Abutting Pebbled Surface Beyond
North of the Palace
The shortcomings of geophysics were shown again just north of the palace where Trench 7 expanded P44, in which we had seen a chalk lined bowl shaped feature previously, and then crossed what was showing up on resistivity surveys like a structure in advance of the palace perimeter. In fact it was a rammed pebble surface, probably a minor roadway leading away from the palace, perhaps towards its water gardens, running along the crest of a significant natural break of slope. We already knew from P45 and P46 (Bulletin 186 p 9) that the modern topography north of the palace had been modified terrifically by dumping when the palace was demolished and this was further evidence of that. In fact the bowl shaped feature resolved itself into a locally damaged white mortared pebble and chalk fragment facing to the slightly concave down slope, suggesting that the palace once sat near the edge of a steep slope down towards Maidens Brook which had been given a white surfacing to contrast with its red brickwork. A single sherd of pottery from the road plus the type of mortar used in the surfacing of the slope tentatively date them to the Tudor development of the site. The existence of such a steep slope would have made the late fifteenth century courtier’s and later Tudor palaces, and their views to the north, more impressive than modern topography would suggest and may help to explain the siting of predecessor Medieval structures here as it would have provided a defensive advantage.
Fig. 3: The Roadway (Left) and Mortared Chalk Slope (Square Hole is the Previously Excavated P44) with Demolition Rubble Seen in Section
The former slope had been entirely disguised during palace demolition by a series of dumps of clay and demolition material and these produced some interesting finds including a fragment of brick with part of a chequers board crudely scratched on it, the other half of the early sixteenth to late seventeenth century bronze belt buckle (Bulletin 186 p 12 No. 2.1) found in P44, a bronze belt strengthener still backed by leather (Fig. 4) (probably from the same belt) and a concentration of sherds from a number of vessels, mainly jugs. Yet again demolition material shows that the structures being demolished in c. 1657 (here one suspects the curtain wall or buildings on its line immediately south of Trench 7) were probably of late fifteenth century date with Tudor repairs/modifications.
Fig. 4: Copper Alloy Belt Strengthener Retaining Leather
Further Geophysical Problems
Trench 8 sought to establish the cause of a linear high resistivity anomaly known to run north from the vicinity of P42 and to establish why P42 had not encountered it when it had been excavated, the anomaly, on analogy with the nearby P43 (Bulletin 186 p 8 and Fig. 2), being presumed to be caused by a wall of the palace. In fact the test pit had not encountered it because it was not there! All that was was a limited area of ?lime mortar slurry, a ground makeup dump perhaps relating to landscaping after the demolition of the palace and yet another rammed pebble surface above it, this time on balance probably a ?cambered promenading surface of the seventeenth century, examples of which elsewhere on the site do not show up on resistivity surveys as this one seems to have.
The Future
This emphasises the increasingly obvious problems in trying to map the palace through geophysical survey data and now that our work in connection with replanting the Lime tree avenue is complete we will be seeking permission to continue research excavations with a view to establishing where it can be presumed to actually be reflecting palace remains and where not. Meanwhile the most recent work, except for the midden environmental analysis, has again been fully written up in an archive report (available at cost – see a committee member) distributed to the relevant authorities and archives and thanks are again offered to all the members of the EAS who participated in it.