SECTION 2.2.2:CLAYDON PIKE TRENCH 13INTRODUCTION PHASE 2: THE LATE IRON AGE/EARLY ROMAN
SETTLEMENT (c AD 25-125) PHASE 3: THE ROMAN COMPLEX (EARLY 2nd
TO EARLY 4th CENTURY AD) PHASE 4: THE LATE ROMAN VILLA COMPLEX PHASE 5: SAXON AND MEDIEVAL ACTIVITY IntroductionTrench 13 covered an area of approximately 8100 m² and represented
the main area of excavation at Claydon Pike. A number of features had
been observed as cropmarks during aerial survey, including enclosure ditches
and linear boundaries (Fig.
2.2.1:Trench plan of Longdoles Field ). The trench was located in
an area that rose slightly above surrounding marshland. Although disturbed
by ploughing, Trench 13 was not as badly affected as the other areas,
but the stratigraphy was, nonetheless, rather shallow. Phasing and ChronologyActivity in Trench 13 has been divided into three main phases (Phases 2, 3 and 4), and each of these has each been sub-divided into a further four phases. The phasing was established on the basis of stratigraphic relationships, while the chronology of these phases has been attributed by examination of the pottery and coins from key context groups. Analysis of the pottery indicated that occupation in Trench 13 was continuous from the early 1st to the later 4th century AD. Phase 2: The Late Iron Age/Early Roman Settlement (c AD 25 - 125; Fig. 2.2.4: Trench 13 composite phase 2 plan )SummaryThe main focus of activity during Phase 2 was exposed in Trench 13. By means of trial trenching and salvage work it was possible to identify additional linear boundaries and enclosures from the cropmark survey, and to relate them to features within the excavated trench. A dense concentration of features covering an area c 2400 m² lay at the heart of the settlement and was defined towards the end of Phase 2 by a substantial boundary ditch, 2502. The settlement nucleus included a complex of gullies, pits, small enclosures and boundary ditches, but there was no clear evidence of structures. Small quantities of daub were recovered, but these were not of sufficient quantity or density to demonstrate the presence of buildings. Pottery and ironworking debris provided evidence of occupation and industrial activity. The stratigraphic and spatial relationships of Phase 2 features in this area indicated four sub-phases, dated on the basis of pottery (Fig. 2.2.5 a, -b, -c, -d: Trench 13 phase 2 sub-phasing ). Occupation appeared to have lasted for approximately one century from c AD 25 to 125. Because chronological divisions within this period were problematic and pottery from many features was sparse, the dating is, in some cases, tentative. Phase 2 feature classes (Fig. 2.2.4: Trench 13 composite phase 2 plan )Features belonging to Phase 2 were classified by type in order to simplify
description and interpretation. Phase 2a features (Figs 2.2.4 and 2.2.5: Trench 13 phase 2 sub-phase 2a)The earliest features in Trench 13 were generally less substantial than later features. They included SE 1, curvilinear gullies 1551 / 537 / 528, circular gully 1757 and gullies 568, 567 and 592. Activity in this phase appears to have been concentrated in the north and east of the trench, and finds of Savernake pottery suggests a date from c AD 25 - 100. Sub-enclosuresSub-enclosure 1, located on the north side of Trench 13, was a three-sided enclosure open to the east. The long axis measured 15 m north-south, and the two east-west arms were 8 m long. The northern terminal had been truncated by Phase 4 enclosure ditch E22. The ditch was c 1.6 m wide with gently sloping sides and c 0.6 m deep and there was some sign that it had been recut (section 136). It contained fills 1805, 2624 and 2611. Finds were relatively sparse, comprising pottery and a few iron nails but in excess of 130 animal bones were also recovered, horse being the most common of the identifiable species. The ditch was cut by north-south linear ditch 2602, and gully 1961 and it post-dated east-west aligned gully 2627. Linear ditch boundariesThe earliest feature in the southern part of Trench 13 was ditch 644.
The ditch formed an arc enclosing an area 22.5 m in diameter, open to
the west and south. The north-western terminal was truncated by boundary
500 and it was not clear whether feature 1900 or 1710 (not shown on plan)
represented its terminal, although the former was most similar in profile.
The ditch had been recut at least once. The shallower original version
was 0.8 m wide and 0.3 to 0.4 m deep. The later version ditch was a deep
flat bottomed ditch, 1 m wide and 0.6 m deep. The later cut may not have
been continuous as it was not traced on the northern side. The small finds
assemblage recovered from the ditch included pottery, daub and window
glass. The features within the enclosed area did not appear to be associated
with the ditch. Linear gully boundariesGullies 568, 567 and 592 appeared to enclose a rectangular plot at the eastern edge of Trench 13. Gully 592 ran in a straight line for c 37 m on a NE-SW alignment. It was 0.4 m wide and 0.2 m deep. Its NW-SE return, gully 567, was c 13 m long and slightly more substantial at 0.6 m wide and 0.4 m deep. Gully 568 formed the western side, 10 m long, 0.4 m wide and 0.4 m deep. Gully 568 truncated a small circular gully, 1757, which enclosed an area c 3 m in diameter. A few sherds of pottery were recovered from the gullies. Circular gulliesCircular gully, 532, lay to the south of boundary ditch 566. It enclosed an area 3.8 m in diameter, and measured 0.5 to 0.9 m wide and 0.5 m deep (section 168). There ware no associated internal features. Small quantities of pottery and animal bone came from the gully fill. The animal bones included a complete donkey metacarpal. Short gullies and gully arcsA group of features, 1551, 537 and 528 ,together formed a penannular gully complex lying to the west of rectangular enclosure 568/567/592. Where excavated, the gully was 0.6 to 0.8 m wide and c 0.25 m deep (Section 172) and defined an area c 8 m in diameter. It appeared to be truncated by circular gully 532, but this relationship was not certain. A ring of five postholes lay within the enclosed area, concentric with the gullies and defining an area 5 m in diameter, but their relationship to the gully complex was not established and no dating evidence was recovered. The southern end of 1551 was truncated by enclosures 11 and 14 and overlain by Aisled Building 3. The eastern stretch of the gully was truncated by E 13. Phase 2b features (Figs 2.2.4 and 2.2.5: Trench 13 phase 2 sub-phase 2b)Phase 2b was characterised by sub-enclosures, but a move to more substantial enclosures was indicated by the construction of E 13 and E 14 during this period. Few major linear boundaries were in use during this time. Enclosures
|
Context no: |
Diameter |
Depth |
Comments |
1580 | 1.1 m | 0.75 m | no stone packing, probably robbed with B 8 wall 1581 |
1893 | 1.1m | - | stone packing survived |
2131 | 0.9m | 0.85m | northern edge overlain by B 8 wall 2132, top half was disturbed, packing material was only preserved in the lower layers |
2135 | 1m | 0.6m | a central fill change was clearly present, disturbed nature of the stone suggests it was due to robbing rather than a post-pipe |
2138 | 0.8m | 0.8m | sealed by B 8 wall 1591, small random blocks of limestone and large flags found collapsed into the posthole. Post may have been truncated at ground level, and subsequently rotted out. Void present on removal of the flags suggested post diameter of c 0.45m |
2153 | 1.1m | 0.5m | robbed and disturbed, sealed by the collapsed hypocaust floor (2116) of B 8. Top third of fill consisted of mortar and rubble consolidation |
2154 | - | - | laid stone rubble present at base; truncated by Phase 5 well 696 |
2194 | 1m | 0.75m | cut by B 8. Intact limestone rubble packing material formed a D-shaped void, measuring 0.8 x 0.3 m |
2195 | 1.1m | 0.75m | limestone rubble packing survived intact giving a cross-section post measurement of c 0.4 m. Cut by robber trench 2309 of B 8 |
2197 | 1.1m | 0.7m | limestone rubble packing survived from c 0.25 m depth giving a cross-sectional measurement of the post of 0.4 m - 0.5 m. Overlain by SE corner of B 7, cut by robber trenches 2112, 2106. No consolidation was made over the top |
2267 | 1.4m | 1.4m | larger post pit due to the presence of the Phase 2 SE 3. Overlain by Phase 5 wall 1999; disturbed by B 8 robbing. Collapsed packing located in lower fill, surviving in situ packing gave a cross-sectional measurement of c 0.35 m - 0.4 m for the post. The top 0.8 m had been infilled with sandy gravel, clay and some rubble |
2269 | 1m | 0.75m | sealed by a thick masonry slab and B 7 wall 2107. Beneath the slab a void c 0.4 m across was recorded, the majority of the packing surviving undisturbed |
2318 | 1.2m | 0.8m | cut by B 8 robber trench 2106. Top half robbed, packing stone collapsed to lie horizontally towards the bottom. D-shape post void c 0.50 m diameter |
Most of the pottery (4.2 kg) recovered from B 1 dated to after the mid
3rd century, particularly that from postholes 2135, 2153, 2195, 2267 and
2269, and that associated with foundation 2518. The remainder was 2nd
century, although not confined to the early part of this period. The presence
of later pottery suggested that the material was incorporated in the postholes
at the time of the destruction of the building, and therefore, indicates
a demolition date of mid 3rd century. Fired clay, iron nails and mortar
/ plaster fragments were most commonly recovered materials from the postholes.
A copper alloy coin dated to 259-268 AD came from posthole 2267. Ceramic
building material was found in most postholes, and ranged from small amounts
(under 70g) in postholes 2131, 2195, 2153 and 1580, to an average of 300-500
g in postholes 2197, 2318, 2267, 2138, 2135 and a large dump of 1600g
in posthole.
Mortar and plaster recovered from nearby Phase 3 b/c well 766 probably
derived from the demolition of B 1. It included type A, a fine white mortar
mix, almost certainly a wall plaster, which produced a very fine surface
for painting. Colours visible on this type of mortar / plaster included
rusty-red, green, yellow and white and some patterns. The presence of
painted plaster indicates more than a utilitarian use for the building.
Type E was also recovered from the well, a hard, fine pink mortar. The
animal bone assemblage from B 1 suggested domestic activity, especially
the bird bones (see Sykes, section 4.3). All of the bird bones from the
building came from the northern half of the structure, suggesting that
different activities took place in the two parts of the building.
B 3, an aisled barn, was situated on the east side of Trench 13, just
1 m inside enclosure boundary 559. The walls, 557, 556 and 688, were stone
founded with a mixed and unfaced small random rubble limestone footing
of up to 0.75 m wide. The overall dimensions of the building were 17 m
by 11 m.
Survival of the footings was good apart from the north side where they
overlay natural gravel. Elsewhere they were dug across the Phase 2 enclosure
ditches. Where the building overlay the ditches of E 11, the ditch tops
had been consolidated with rubble within the interior of the barn. A gap
of 1.75 m centrally placed along the southern wall marked a c 2 m wide
entrance. The walls were partially robbed on the south and east sides.
The robber trench, 736, was between 0.5 m to 0.7 m wide and 0.3 m deep.
A mid 4th century coin (SF 610) was recovered from the top of the fill.
Four sets of stone-packed post pits, 677, 676, 538, 678, 691, 685, 690
and 684, marked the position of the aisles. The building had three bays
but there was an additional central posthole, 683, at the eastern end.
It is unlikely that this was an aisle post, and its dimensions were not
recorded, but examples of aisled buildings with end posts do exist (eg
Somerford Keynes, see section 5), and given its position between two aisle
posts, this arrangement must remain a possibility.
The bays were 4 m wide, with 1.75 m wide aisles and the nave was 6.25
m across. The aisle posts were all of similar dimensions, ranging from
0.6 m to 0.8 m in diameter and 0.6 m to 0.7 m in depth. Undisturbed limestone
rubble packing was present in all but one, 690, where the packing had
been disturbed. Post-pipes ranged from 0.15 m to 0.20 m in diameter. The
small quantities of pottery recovered was mostly of mid 3rd century date,
and a late 3rd century coin (SF 408) was recovered from the top of the
packing of one posthole. These finds had probably been incorporated during
demolition.
An internal division, 689, was represented by a 1.6 m length of dense
limestone rubble c 0.8 m wide. It ran north from posthole 690 into the
interior of the barn. It sealed the packing of posthole 690 but could
have abutted the post. Three other postholes may also have been associated
with the structure. Posthole 683, mentioned above, lay midway between
aisle posts 678 and 684 at the eastern end of the building. It was slightly
smaller than the main aisle post pits but was similarly well packed and
the post-pipe was 0.2 m in diameter. Postholes 734 and 735 were much slighter.
Both lay less than 1 m from the inside edge of the southern wall, either
side of the entrance and may have been doorway features.
Pit 686 lay 1 m to the east of the south-western aisle post, 691. It contained
a near complete Savernake storage jar and had clearly been dug to take
the vessel. The rim had been broken off and was found within the jar.
A semi-articulated sheep skeleton (699) lay centrally within the building
(not shown on plan), but it was not clear whether it was contemporary
with the structure.
A patch of cobbling, 616, abutted the outer face of wall 556 on the east
side of B3, running at a slight angle to the building axis. It overlay
the eastern boundary ditch, 559, but was cut by Phase 3c fence line 2.
The cobbling was not contemporary with the construction of the building
and probably belonged to Phase 3 b.
It was difficult to date the destruction of B 3 but it had certainly gone
by the end of Phase 4 as enclosure ditch 501 (E 22) cut through the western
end. It seems likely that the building could have continued in use into
the early 4th century (Phase 3d or 4 a), probably in a slightly dilapidated
state. The two coins were recovered from 678 and 736, which were not very
reliable contexts, although the coin from robber trench 736 (dated AD
330-346) indicated the building was certainly demolished by the mid 4th
century. The Savernake jar in pit 686 indicated use of the building in
the 2nd century. The 2.2 kg of pottery recovered from floor make up layer
522 was a mixed assemblage, mostly dating to after 250 AD, suggesting
that the building was in use at least until the end of the 3rd century.
Layer 522 also contained a relatively large number of small finds. Most
were iron objects, including more than 28 nails, a rod, awl, knife, horseshoe
and tool collar, but also smithing slag, a worked bone object, vessel
glass, mudstone spindle whorl, a copper alloy vine leaf (probably from
a lamp), a pin and fired clay, including daub. The rubble of wall 688
produced a silver finger ring and an iron hobnail plate. Few small finds
were recovered from the postholes. but 538 contained a copper alloy bell,
677 contained fired clay, 678 a coin (see above), 690 fired clay and an
iron nail and 691 vessel glass split which joined fragment SF 483 from
context 693,a Phase 4 rubble layer. Only a small quantity of animal bone
was recovered from this building, and none from the general floor make
up layer, 522. This is in sharp contrast to B1 and suggests the possibility
of a non-domestic function for this building.
During Phase 3b the western boundary continued to shift progressively westwards, with a fence line (F 6) constructed immediately east of the southern part of the boundary, and a stone wall (2190) adjacent to the northern part. A second fence line, F 4, was positioned 21 m east of the first. The eastern boundary ditch was recut as a continuous ditch, 559. The trackway ditches were infilled and consolidated with cobbling, suggesting that the area may still have functioned as a thoroughfare. The gateway structure continued in use, as did both of the aisled buildings. A two room timber building, B 2, was erected on the western side of B 1, lying perpendicular to it. On the eastern side of the B 1 three ditches, 1595, 1594 and 781, created an internal boundary, possibly to control access to the building. A well was in use at the south-eastern corner of B 1 during this phase.
During this phase the western boundary ditch was recut as 2301. This
ditch was situated c 2 to 3 m west of Phase 2a ditch 2198 (sections 113
and 114), and was represented by at least three cuts ranging from 0.4
m to 0.7 m deep and c 1 m wide. Ditch 2301 was cut by Phase 2 ditch 2930,
which ran east-west in Trench 29, turning north at the gateway and then
followed the same alignment as 2301.
Ditch 2301 ran through the centre of the hollow separating Trench 17 and
Trench 13, and its fill was siltier than that of 2198 but contained no
waterlogged material. The pottery assemblage (c 2.2 kg) was mixed, dating
from the 2nd to mid-late 3rd century, indicating that it continued as
a boundary in use into Phase 3c. Small finds were few and consisted of
a bone pin and window glass fragments. Almost 300 animal bones were recovered
from a number of the ditches.
A length of wall, 2190, to the east of 2301, was probably built during
this phase. It was c 5 m long and overlay ditch 2191. A large deposit
of stone rubble, 2454, lying c 20 m to the south of the wall over ditch
2198 just north of the gateway wall 2331, may have been part of the southern
section of the same wall.
South of the gateway, ditch 2301 continued south as ditch 2161. It was
a recut of Phase 2a ditch 2162, and of similar dimensions. Two cuts were
visible, 1 m wide and 0.5 m deep. They respected the southern gateway
wall at its northern end and continued southwards as a linear boundary
south of the main settlement. The pottery (3.4 kg) indicated a date in
the second half of the 2nd century. The ditch also contained 2 kg of ceramic
building material and several iron nails, slag fragments, a piece of lead
sheet, a copper alloy pin and locally derived building stone. A total
of 244 animal bones was also recovered.
During this phase the gateway was likely to have continued in use. Although the trackway ditches had been filled in, they were sealed by a cobble layer, 2007, suggesting that this stretch still functioned as a thoroughfare.
Ditch 559 was recut during Phase 2b to form a continuous solid boundary ditch (section 175). The recutting may have occurred after the infilling of the trackway ditches. At some point towards the end of this phase or beginning of the next, the eastern boundary ditch in Trench 13 was consolidated with a layer of limestone rubble, 616 (Fig. 2.2.10: Phase 3 aisled building 3 ), which was crossed by fence line 2 in Phase 3c.
Ditches 1594, 1595 and 781 bounded two sides of an area 18 m by 10 m,
open to the north, with B 1 forming part of the west side. A 3 m wide
entrance lay on the east side. Ditches 1594 and 1595 were of different
dimensions, albeit contemporary. Ditch 1594 was 1.2 m wide and 0.4 m deep
whilst 1595 was 0.8 m wide and 0.5 m deep. Ditch 1595 terminated 6 m from
B 1. Ditch 1594 turned north from 1595 and continued for 4 m.
The northern terminal of ditch 781 was not well defined. Phase 4 ditch
780 cut the northern stretch of this ditch and it was difficult to trace
it through the intersection of the Phase 2 ditches of SE 3. A slight swelling
in the northern side of Phase 4 ditch 870 may represent the surviving
terminal. A deposit of rubble spread north-eastwards between the terminals
of 781 and 1594 (not shown on plan) and a concentration of this material
0.6 m² in size adjacent to the terminal of 781 may have represented
the base of a gateway.
The three ditches produced a significant pottery assemblage (2.73 kg)
of mixed material dating to the 2nd century and the second half of the
3rd century. A fragment of wall plaster suggested that 1595 was open during
the destruction of B 1. Fired clay oven fragments were also recovered.
Small finds from 1594 included a fired clay fragment, iron nails and a
lead weight. Ditch 781 also contained fired clay and iron nails. Almost
300 animal bones, including domestic fowl, were also recovered from these
features.
F 4 was situated on the southern side of the trench running approximately
NE-SW, parallel to F 6 (see below), but slightly obliquely to Phase 3c
fence lines 2 and 3. The northern part of the fence line, 834, appeared
to cut the southern edge of trackway ditch 620, and from this three further
posts 817, 1529 and 2226, ran southwards at intervals of c 2.4 m. Although
it cut the trackway, the fence line appeared to respect its line, implying
that the area remained a thoroughfare. The postholes had distinctive oval
shapes in plan and preserved evidence of stone packing and the post-pipe.
On the basis of its relationship to ditch 620 and F 2 and F 3, this fence
was constructed during Phase 3b. Dating evidence was inconclusive. A few
sherds of 1st century pottery (20g) were recovered from posthole 817,
and a late 4th century coin from the top of 834, but was not securely
stratified within the posthole.
CTX |
Maximum diameter |
Depth |
817 | 0.7 m | 0.4 m |
834 | 0.8 m | 0.36 m |
1529 | c 0.7 m | 0.2 m |
2226 | 0.7 m | 0.22 m |
F 6 was aligned NE-SW and bordered the rectangular enclosure in Trench
19. It included at least three aligned postholes, 2166, 2165 and 2164
and probably 2163, 2353 and 2354 after a break of 6 m. The fence line
respected ditch 2161, running parallel to it, cutting the eastern edge
of Phase 3a ditch 2162. The spacing between the first four posts varied
from 1.6 to 2 m, but 2353 and 2354 were separated by only 0.1 m. All the
posts were packed with limestone.
Phasing was problematic as no finds were recovered. F6 it post-dated Phase
3a boundary 2162, and was aligned with F 4, and was, therefore, assigned
to Phase 3b.
CTX |
Diameter |
Depth |
2163 | 0.48 m | 0.26 m |
2164 | 0.58 m | 0.21 m |
2165 | 0.4 m | 0.23 m |
2166 | 0.58 m | 0.21 m |
2353 | 0.5 m | 0.35 m |
2354 | 0.4 m | 0.3 m |
B2 lay in the western part of Trench 13, set at right angles to B 1.
It was represented by four shallow slots, 2509, 2510, 2513 and 2514, c
0.2 m deep and 0.6 m wide. The width of the slots suggested that they
held masonry foundations, subsequently entirely robbed, but the nature
of the superstructure remains uncertain. The structure was a rectangular
extension to B 1, divided into two rooms by slot 2510. It measured 9.5
m by 7.5 m overall and the individual rooms were c 6 m by 6.5 m and 6
m by 2 m. No slot existed on the east side against B 1. Slot 2509 ran
to within 0.5 m of the conjectured wall of the B 1 and slot 2513 ended
c 2 m from it.
B 2 overlay Phase 3a features 2526 and 2175, and was cut by Phase 4 ditch
700 (E 21). The pottery assemblage recovered from the slots was substantial
(11.8 kg) and mixed but generally relatively late and included a much
Black Burnished Ware, including 2nd and mid to late 3rd century forms.
The sherds were larger than those from B 1 and B 3, indicating that they
had been in use at the time of demolition. Small finds from 2509 included
iron nails, a looped iron pin, iron boot plate, an iron Nauheim derivative
brooch, vessel glass, an Old Red Sandstone whetstone and a bone pin. Slot
2513 contained a copper alloy coin dated to AD 364-378 and two iron nails,
and 2514 contained iron nails and vessel glass fragments. A group of over
300 animal bones belonged, for the most part, to the main domesticates
(Sykes section 4.3) but included red deer antler. The demolition of B
2 and B1 may have been contemporary.
Pit 663, in the south-east part of Trench 13, was an irregularlar shaped feature approximately 6 m in diameter and 0.5 m deep. It cut Phase 3a trackway ditch 620 (section 158) and Phase 2 ditches 645 and 644, and was sealed by a layer of hard standing, 647. This material may have been contemporary with, or equivalent to, cobbled layer 616 that sealed ditch 559 further to the east. Pit 663 was assigned to this phase on the basis that it contained mid to late 2nd century pottery. Small finds were rare and included a few iron nails, fired clay and a copper alloy pin. The size and composition of the pottery (2.1 kg) and animal bone (103 fragments) assemblages suggested that this material was dumped in the feature.
Well 766 was situated adjacent to the south-eastern corner of B 1. It
was not clearly visible during the first stages of the excavation as it
had been sealed during Phase 4, and part of the stone lining had been
robbed. It was 1.8 m wide at the top, narrowing to 1.4 m at the base (section
194) and was 1.26 m deep with a flat stone slab on the base. The dry-stone
wall lining continued to the base. The infill included layers of dark
grey sandy loam, mixed clay and gravel, along with a layer of charcoal,
7, and a small amount of rubble, primarily in layers 1 and 2. The large
pottery assemblage (c 3 kg) dated mostly from the mid to late 3rd century,
and all of the later Oxford colour coated ware came from the top two layers.
Small finds included over 16 iron nails, fired clay fragments, copper
alloy wire, two bone pins, an unidentified leather object and wall plaster
fragments distributed throughout the fills (object numbers 3206-3209,
3194). Most of the plaster was type A and probably derived from the destruction
of B 1 (see above). Most of the plaster came from the lower fill, 5, of
the well, and almost all of the nails came from layers above. It is possible
that it represented successive stages of building demolition and that
the well was infilled with the debris over a relatively short period of
time. The shallow depth of the well prevented the survival of organic
deposits. Over 650 animal bones were recovered, representing a large range
of species. The assemblage may have represented primary domestic refuse,
perhaps the remains of a single high-status meal (Sykes, section 4.3).
The well appeared to cut ditch 2156 and the pottery and plaster recovered
from it indicated that it was filled during Phase 3c and/or Phase 3d and
was, therefore, most likely to have been in use during Phase 3b and 3c.
During Phase 3c the western boundary appeared to continue in use, although the southern extent was not defined. A short stretch of wall, 2193, appeared to succeed the Phase 3b wall. The gateway seems to have remained in use. The eastern boundary ditch, 559, was infilled and cobbled over, and replaced by a 60 m fence line (F 2), running obliquely to the line of ditch 559. To the north of the fence line was ditch 562, associated with a large paddock to the east of Trench 13, visible as a cropmark. A shorter fence line, F 3, ran parallel c 27 m to the west of F 2. Two lines of posts ran at near right angles 13 m further to the west to form fence line 5. All three buildings continued in use, as did the associated internal boundary ditches.
During Phase 3c the western boundary, consisting of ditch 2301 and possibly 2161, continued in use. A small stretch of wall, 2193, and its robber trench, 2192, lay just north of the second phase gateway (section 113). This wall was the successor to wall 2454, which lay at the southern end of 2198. Wall 2193 was on a 1 m wide footing of small un-mortared rubble. It extended north c 10 m from the gateway but there was no evidence that it continued further. The relationship between the wall and the gateway structure was not certain, but on spatial evidence they were likely to be contemporary. The pottery from ditch 2301 was mostly mid to late 3d century in date, suggesting that the boundary was little used after this phase. Whether the boundary continued to the south was uncertain as there was no dating evidence to confirm this. Spatially the plan of the area remained little changed from Phase 3b to 3c and it was therefore possibile that ditch 2161 remained open during this phase.
Ditch 559 was replaced by a fence line, F2, situated in the western part
of Trench 13. Ditch 562, in the north-eastern corner of Trench 13, bordered
a large paddock c 60 m by 50 m in size, defined by ditches 562, 1539 and
1748. Ditch 562 (section 174) was c 2 to 2.5 m wide and c 0.9 m deep,
and ran perpendicular to boundary ditch 559. Although it post-dated 559
it turned north, continuing its alignment.
The cropmark survey indicated that 562 continued south-eastwards beyond
Trench 13, turning southwards after c 50 m and continuing for c 60 m.
It then returned in the direction of the trench, emerging in the south-eastern
corner as1539 (recut as 1748). A 7 m stretch of 1539 was exposed in Trench
13, and the top had been filled by limestone rubble in this area. Ditch
1539 was just over 1 m wide, but was later recut twice as 1748, c 2 m
wide and deeper than the former ditch. Little pottery (0.3 kg) was recovered
from the southern stretch of the eastern boundary ditch. Small finds included
a piece of melted lead and an iron nail. Just to the south of Trench 13,
cropmark evidence revealed a ditch running south from this enclosure,
cutting the line of ditch 559, and probably turning westwards as ditch
1801 in trench 26 (see southern and south-eastern trenches below).
Most of the pottery from 562 (2.3 kg) was late 3rd century in date and
possibly slightly later. Small finds included a few several iron nails,
an iron spearhead and vessel and window glass fragments. Over 300 animal
bone fragments were recovered from the ditches, most from 562.
F 2 was represented by a line of postholes spaced c 2 m apart which lay
on a NNE-SSW alignment 60 m long within the excavated area. They ran from
just south of the intersection of ditches 559 and 562 approximately parallel
to the western edge of B 3 and were probably associated with it. The postholes
ranged from c 0.3 m to c 0.8 m in diameter and from c 0.15 m to 0.35 m
deep. Most had pitched limestone packing, much of which was undisturbed.
Surviving post voids ranged from c 0.12 m to c 0.2 m in cross-section
and were square in plan.
This fence line stopped 5 m short of the 559 / 562 intersection. Three
additional postholes, 600, 604 and 596, ran at an angle for 7 m from the
north end of the fence line. They may have been contemporary with it,
forming the southern side of a 5-6 m wide entrance to the paddock defined
by ditch 562. A further two postholes (un-numbered) ran at an angle from
the southern stretch of F2. They may have formed part of an internal division
in the paddock. F 2 continued south beyond the point where ditch 1539
met it but was not located in the trenches south of Trench 13.
Small finds from the post line included a Millstone Grit quern from 609,
and iron nails from 621 and 623. Most of the pottery recovered was undiagnostic
but a late 2nd to 3rd century date seemed likely. F 2 cut the trackway
and its bordering ditches, and probably cut across cobbled area 616 which
overlay ditch 559.
Context |
Maximum diameter |
Maximum depth |
Size of post void |
Comments |
605 | 0.8 m | stone packed | ||
606 | 0.5 m | west of 559, east of B3 | ||
607 | 0.58 m | between 606 and 608 | ||
608 | 0.57 m | south of 607 | ||
609 | 0.5 m | south of 608 | ||
610 | 0.5 m | south of 609 | ||
611 | 0.5 m | 0.3 m | stone packed, post void visible, south of 610 | |
612 | 0.6 m | 0.26 m | stone packed, post void visible, south of 611 | |
613 | 0.53 m | 0.35 m | stone packed, post void visible, south of 612 | |
623 | 0.5 m | 0.26 m | 0.18 x 0.26 m | stone packed, post void visible, south of 613 |
624 | 0.45-0.6 m | 0.15 m | 0.13 x 0.14 m | stone packed, post void visible, south of 623 |
625 | 0.6 m | 0.3 m | 0.2 x 0.30 m | stone packed, post void visible, south of 624 |
626 | 0.6 m | 0.35 m | 0.12 x 0.36 m | stone packed, post void visible, south of 625 |
627 | 0.7 m | 0.33 m | 0.18 x 0.34 m | stone packed, post void visible, south of 626 |
268 | 0.3 m | 0.25 m | stone packed posthole, south of 627 | |
269 | 0.4 m | 0.26 m | stone packed posthole, north of 621 | |
621 | 0.6 m | cuts 547 | ||
636 | 0.7 m | 0.22 m | 0.17 x 0.22 m | stone packed, post void visible, south of 547,621 |
637 | 0.7 m | 0.21 m | 0.12 m | stone packed, post void visible, south of 636 |
638 | 0.7 m | cuts 620, south of 638 | ||
639 | 0.6 m | south of 638 | ||
640 | 0.5 m | south of 639 | ||
641 | 0.6 m | south of 640 | ||
642 | 0.6 m | south of 640 |
F 3 ran north-eastwards from the southern end of Trench 13 for a distance of 26 m. It was c 27 - 28 m west of and parallel to F 2 and was represented by at least eight postholes. Two, 1541 and 1540, appeared to be connected by a shallow gully, 1516, 2 m long. Intervals between the posts varied from 2 to 4 m. The fence ran northwards to the line of ditch 547, cutting across the trackway. It continued beyond the southern trench boundary. The sparse dating evidence available suggested a 3rd century date, and , considering its alignment, it was probably contemporary with F 2
Context | Diameter | Depth | Comments |
2084 | 0.5 m | 0.22 m | stone packed |
812 | 0.4-0.5 m | 12 | stone packed |
849 | 0.6 m | 0.3 m | stone packed |
1517 | 0.35 m | 0.07 m | |
1540 | 0.4 m | 0.3 m | stone packed |
1541 | 0.42 m | 0.2 m | stone packed |
1545 | 0.45 m | - | stone packed, 5 stones at edge formed pentagon shape, post pipe visible |
838 | 0.36 m | 0.18 m | stone packed |
F 5 was located c 13 m to the west of F 3 and was represented by four
postholes, 2217, 2206, 2205 and 2214, forming a c 10 m long alignment.
Three postholes, 2211, 2200 and 2225, appeared to represent a 6 m long
east-west aligned return to the fence line, but no further postholes were
exposed in the salvage area to the west.
The east-west line of posts was irregularly spaced, varying from 1.5 m
to 3 m from post centre to post centre. A gap of just over 2 m lay between
posts 2214 and 2205 on the NE-SW line. There was then a break of 5 m before
the fence line continued as two more posts 2 m apart. The two sets of
postholes were not, however, precisely aligned with each other. Only a
few sherds of pottery were recovered, including 150g from 2214, dated
to the 2nd century. F 5 could truncated Phase 2 features but no relationships
were observed with later features. F 5 was, therefore, assigned to Phase
3c on the basis of its spatial relationship with F 2 and F 3.
Context |
Diameter |
Depth |
Comments |
2200 | 0.55 m | 0.35 m | stone packed, post void 0.13 x 0.19 m |
2211 | 0.25 x 0.4 m | 0.12 m | contains one burnt stone fragment |
2225 | 0.6 m | ||
2214 | 0.5 m | 0.26 m | stone packed |
2205 | 0.42 m | 0.4 m | stone packed |
2206 | 0.64 m | 0.52 m | large stone packed posthole, post void seen flush with noeth west side |
2212 | 0.59 m | 0.42 m | large stone packed posthole |
During Phase 3c B 1, B 2 and B 3 continued in use. Ditches 1594, 1595 and 781 were also still open and well 766 was probably operational during part of this phase. At some point at the end of this phase or beginning of the next, B 1 and B 2 were demolished, the well filled with demolition rubble, and ditches 1594, 1595 and 781 infilled.
The end of Phase 3 saw a re-ordering of the enclosed area. The western boundary was infilled, and wall 2193 was robbed (2300). Either during Phase 3d or Phase 4, the southern part of western boundary ditch 2301 was covered by a layer of cobbling that extended into Trench 19, north of the rectangular enclosure. B 1 and B 2 were demolished, and internal boundaries 781, 1594 and 1595 went out of use. These features were replaced by a single small square building, B 7. Three oven / hearth features were exposed in the interior of this building, suggesting domestic activity. Aisled building B 3 continued in use at some level. Lying just to the west of B 3 was well 502, which may have belonged to both Phase 3d and 4a/b. It has been described with Phase 4 features. F 2 may have continued into this phase, given the spatial relationship with B 3, and F 3, therefore, may also have still been standing. The majority of the pottery from ditch 562 was late 3rd century in date, but some may have been later, indicating that the paddock to the east of Trench 13 may have also continued in use up to this point.
B 7 was a small square structure, measuring 8 by 8 m internally, erected
over the northern end of B 1. Its east wall sealed four of the aisle posts
of B 1. Only two small fragments of wall or footings survived, a 2 m long
fragment of the east wall, 2107, and and 0.8 m stretch of the north wall,
2139. Both wall fragments were coursed and regularly faced in contrast
to B 1 and the later B 8 which succeeded B 7. A total of 70 g of 3rd and
4th century pottery was recovered from 2107.
Wall 2139 was sealed by a layer of levelling material, 1949, beneath the
floors of B 8, while wall 2107 had been reused in B 8. Wall 2107 sealed
B 1 aisle post 2269. A large masonry block had been placed over the truncated
aisled post, which had subsequently rotted, leaving a void. The walls
elsewhere were not deeply founded. Robber Trench 1977, robbing wall 2139,
was barely 0.1 m deep, and 2106, robbing wall 2107, was 0.14 m at most.
Robber trench 2106 continued south from wall 2107 for c 4 m. It was overlain
by Phase 4 layer 1929 (B 8) and Phase 5 wall 1999. Robber Trench 1977
ran for 5 m from the north-east corner of the building, before it was
cut by robber Trench 1947 of B 8. It was not traced west of 1947, possibly
due to the poor survival of the stratigraphy in this area. A small hearth,
1948, overlay robber trench 1977 but was sealed by 1949, a floor make-up
layer in B 8. This places the hearth stratigraphically between B 7 and
B 8, perhaps representing activity between the demolition of B 7 and construction
of B 8.
No pottery or small finds were recovered from the hearth, and robber Trench
1977 did not produce diagnostic pottery but 4th century pottery (0.6 kg)
was recovered from 2106, along with an iron nail and wall mortar / plaster
fragments.
Definition of the south and west sides of the building were more problematic.
The line of the southern wall showed clearly as a layer change between
internal layer 2152 (see below) and external layer 2196, and possible
robber trench 2112 was visible running east-west in this area. It contained
type F mortar / plaster fragments (a soft and porous mix), abutted layers
2152 and 2196, and was overlain by 1929 (a layer associated with B 8).
The western section had been removed by medieval well, 696. Given the
truncation of robber trenches 1977 and 2112 and the poor preservation
of features, the western part of the structure is conjectural. It appeared
likely that it was on the line of the western wall of B 8, 1556. The robber
trench of this wall, 2102, continued south beyond the south-west corner
of B 8 and ended on line with 2112, the southern robber trench of B 7.
Robbing would have taken place during the demolition of B 8. Internal
layer 2152, a black sandy loam, appeared to be a contemporary horizon,
although not showing signs of compaction expected of a floor surface.
Only 78 g of pottery was recovered from it along with an iron nail and
copper alloy coin dated to AD 250- 400. A series of hearths, ovens, postholes
and pits were defined at this level and appeared to be associated with
B 7. Layer 2152 petered out to the west, exposing B 1 aisle post. An area
of disturbance, layer 2145, associated with robbing of aisle post 2267,
spread north to also seal aisle post 2269. Phase 2 layers 2143 and 2144
were visible from below. Layer 2111, to the north of 2152, was also a
black sandy loam, overlain by 1920, demolition layer of B 7, and abutting
robber Trench 1977. It is likely that this layer was contemporary with
2152. It contained 4th century pottery (0.82 kg), a fired clay fragment,
several iron nails, an Old Red Sandstone whetstone and iron arrowhead.
Most of the 81 animal bones associated with B 7 came from this layer.
Three ovens or hearths lying within the interior were probably contemporary
with the use of the building (Fig.
2.2.12). Feature 2136 was poorly preserved but some stone facing survived
on the south and east walls. The remains of this feature measured 1.4
m by 0.8 m. The north-western side was overlain by wall 1591 but the hearth
impinged on the cut of aisle post 2267. During the construction of B 8
this post had been sealed just below ground level by limestone slabs and
there was no evidence for similar sealing during the use of B 7. The pottery
from the hearth indicated a late 3rd century date.
Hearth 1965 was situated in the north-eastern part of B 7. It was defined
by two perpendicular lines of burnt and pitched stone forming an area
1 m by 0.8 m from which a patch of burnt clay c 0.7 m in length extended.
Burnt stone lay within it and to the north, against the outer edge of
the pitched stone, was an oven of flat burnt stone c 0.4 m by 0.4 m. Hearth
1965 overlay layer 2111, and was sealed by destruction layer 1920, which
sealed interior layer 2152. No pottery or small finds were recovered from
the hearth, but 4th century pottery was recorded in the demolition layer.
Oven 2113 lay to the south of 2136, orientated NW-SE. It was circular
in plan, and constructed of up to four courses of limestone, with pitched
slabs on the eastern side representing a fireplace, leading to a stokehole
(Fig.
2.2.12 section). The fill was a black sandy loam containing charcoal,
and a small quantity of charred grain (wheat, barley and flax). The secondary
fill covered both the oven and the stokehole. The oven was well preserved
and had been set deeper than the other features within B 7. It was sealed
by layer 1929. It may have been a simple type of bowl-shaped corn drier,
as defined by Morris (1979, 182), although the quantity of grain was very
slight. Small finds consisted of fired clay fragments including daub,
and a mortar / plaster fragment. Only 0.76 kg of undiagnostic pottery
was recovered.
A concentration of postholes, 2147-2151, 2155, 2265 and 2266, was located
to the east of hearth 2113. They were all of a similar size, c 0.4 m diameter
and 0.3 m deep with limestone packing. Postholes 2147, 2148, 2149 and
2151 formed a 1 m² four post structure around the stokehole to 2113.
A few pottery sherds were recovered from the postholes and no small finds
were recorded. An oval pit, 2140, c 1 m x 0.50 m and 0.22 m deep, lay
adjacent to these postholes. It contained a greyish black fill of sandy
loam with charcoal which produced 300 g of ceramic building material,
and 0.40 kg of pottery.
B 7 may have been erected soon after the demolition of B 1, and it did
not appear to have a long lifespan. Certainly the severed aisle post in
2138 had not begun to rot as, wall 1591 was deliberately laid over it
during construction of B 8. An arc of gully, 2114, cut the south-east
corner of B 7, but was cut by the robber trench 1578, of the north wall
of B 9. There were few finds to indicate the function of the building,
but it was well constructed, and the presence of wall plaster (if it did
indeed relate to this building) suggests a domestic dwelling of some kind.
The presence of hearths and a possible corn-drying oven associated with
grain and chaff suggests that at some point crop processing took place
within the building.
Gully 1988 post-dated B 1 but ran on the same axis as the eastern aisle
post line. Its northern terminal lay just north of the south-east corner
of B 7. It extended south for c 7 m where its southern end was truncated
by ditch 700 (E 21). The gully was narrow but comparatively deep, measuring
0.55 m wide and 0.5 m deep. It contained a rubble fill with late 3rd to
4th century pottery (4.5 kg). A causeway of coursed limestone slabs and
gravelly mortar, 2126, had been constructed across the gully, approximately
1 m in length and at least as deep as the gully.
Gully 1988 post-dated B 1 but would logically predate B 8 boundary wall
1587. It is likely to be contemporary with B 7, perhaps reinforcing a
boundary originally defined by B 1. Gully 1988 and 2108 were overlain
by Phase 4 layer 1929. Apart from pottery, finds from the gully comprised
iron nails, a whetstone, mortar and plaster (probably from B 1), fired
clay and a bone bobbin. An assemblage of 188 bones was also recovered,
and included domestic fowl and hare.
Oven 2103 lay to the south of ditch 1988 and was cut by ditch 700. Three stone courses survived and it had clearly been robbed. It was filled with charcoal, burnt stone and 0.48 kg of pottery of 2nd to 4th century date. A coin dating to the late 4th century was obviously intrusive. Large quantities of glume bases of spelt wheat were recovered from the feature. It was clearly later than B 1, and spatially more likely to belong to Phase 3d or 4a as it lay within the entranceway to the Phase 4b villa.
Aerial photographs indicated that at the north-eastern corner of Trench 13, ditches 559 and 562, extended in a north-westerly direction to join ditches 2301 and 2198. This northern boundary, designated 548 / 2642, was only observed during salvage operations. A parallel but slightly smaller ditch, 2641, ran 2 to 3 m to the north, and at the western end turned south into 548, c 5 m short of the western boundary. It was cut by the later phases of 548 / 2642 and was sealed by hearth 2634. Further to the west, ditch angles 2198 and 2301 were indistinguishable and their fills merged with ditch 548. Very few finds were recovered from the ditches, implying that they were at a remove from the main areas of domestic activity.
Three fence lines, F 1, F 9 and F 7, were thought to belong to Phase 3, but could not be securely assigned to a sub-phase. F 1 and F 9 may have formed two sides of an enclosed area, 11 m NE-SW (F 9), and 13 m NW-SE (F 1). F 9 included postholes 2621, 2613, 2635, F 1 of 2623, 2609, 2622, 2601 and 2606. Posthole 2607 joined both lines.
Six postholes running NE-SW represented a fence line 13 m long. The postholes
were on average 2.25 m from post centre to post centre, the exception
being 2607 and 2623 at 4.5 m apart. This gap suggests the fence line predated
Phase 4 ditch 2608. The fence line may have extended one post further
in either direction, the evidence having been destroyed by Phase 4 ditches
1553 and 700.
All were stone packed, with ceramic building material showing in the upper
fill of 2609 and 2623. Diameters ranged from 0.3 to 0.44 m and depth from
0.18 to 0.38 m. The posts were sealed by layer 2596 (unphased). F1 post-dated
Phase 2, cutting across the tops of ditch 2602. Few finds were recovered.
Context |
Diameter |
Depth |
Comments |
2601 | 0.38 m | 0.34 m | stone packed, cuts 2602, overlaid by2596 |
2606 | 0.3 m | 0.2 m | stone packed, overlaid by 2596 |
2607 | 0.3 m | 0.25 m | stone packed |
2609 | 0.38 m | 0.18 m | stone packed, CBM in upper fill, overlain by 2596 |
2622 | - | - | stone packed, overlain by 2596 |
2623 | 0.44 m | 0.3 m | stone packed, CBM in upper fill |
F 9 ran perpendicular to F 1, from its northern extent at posthole 2607, southwards for 11 m. Another posthole, 1932, would extend this to 17 m, but this was less certainly associated with the alignment. The postholes were irregularly spaced, between 2.5 m and 5 m apart, and it is possible that other postholes on this line may not have survived due to the presence of Phase 4 ditch 2608. All were 0.3 m diameter with the exception of 2613, which was 0.5 m, and all were stone packed. F 9 was cut by Phase 4 ditch 765, but few finds were recovered.
Context |
Diameter |
Depth |
Comments |
2607 | 0.3 m | 0.25 m | stone packed |
2613 | 0.5 m | 0.35 m | stone packed, cut by765 |
2621 | 0.38 m | 0.2 m | stone packed |
2635 | 0.3 m | 0.2 m | stone packed |
1932 | 0.3 m | 0.2 m | cuts layer1910 |
F 7 ran NW-SE for 7 m from the southern terminus of 2175. It lay parallel to gully 2710 (see below) and c 26 m south of it. It comprised four posts, 2177-2180, set between 2 m and 2.5 m apart, all stone packed. They varied in diameter from 0.4 m to 0.6 m and in depth from 0.14 to 0.2 m. F 7 was aligned on the same axis as the Phase 3 enclosure ditches, and appeared to truncate Phase 3a ditch 2175.
Context | Diameter | Depth | Comments |
2177 | 0.6 m | 0.14 m | stone packed |
2178 | 0.4 m | 0.2 m | stone packed |
2179 | 0.5 m | 0.18 m | stone packed |
2180 | 0.6 m | 0.2 m | stone packed |
Pit 2160 lay to the east of the western boundary, and appeared to truncate
Phase 3a internal boundary 2175. It was c 2.4 m wide and 1.3 m deep and
was truncated by Phase 4d ditch 501 (E 22). The pit contained just over
1 kg of pottery tentatively dated to the 2nd century and 54 animal bones.
No small finds were recovered. The pit sides were fairly steep and the
base flat. The lowest layers were waterlogged (section 113), and the feature
may well have been a waterhole. Environmental samples were taken from
the feature and are discussed in section 4.4.
Pit 2172 was located to the south of B 1, cutting into Phase 2a ditch
2156. his feature was rectangular in plan, 1.3 m by 0.9 m and 0.3 m deep.
It contained 0.25 kg of undiagnostic pottery and an iron nail, along with
a small amount of animal bone. On the basis of its relationship with 2156
it was assigned to Phase 3b or later.
Pit 2218 lay partly below the limit of excavation in the south-west part
of Trench 13. It cut Phase 2 gully 2202, but relationships with Phase
3 gullies 2208 and 2210 were not clear. The pit contained few finds.
A NW-SE row of roughly circular pits lay between F 4 and F 3, just south
of trackway ditch 620, although it is not known if they were associated
with one another. Pit 587 lay just west of F 3 postholes 1545 and 1540
and was approximately 1.3 m in diameter. It contained small amounts of
2nd century pottery and vessel glass fragments. Pit 826 lay 6 m to the
north-west and also produced very small amounts of 2nd century pottery.
Pit 1505 lay south-east of F 4, cutting Phase 2 gully 804. It contained
a small amount of 3rd century pottery and an iron nail.
A sub-rectangular pit, 513, c 3 m across lay to the north-east of Trench
13. It contained 2nd to 3rd century pottery, an iron nail and fragments
of daub. It was cut by Phase 4 ditch 501.
Immediately north of 547 and west of F 2 two intercutting pits were exposed.
Pit 703 was c 2 m in diameter and c 0.4 m deep, and contained pottery
(1.8 kg) broadly dating from the mid 1st to 3rd centuries, an iron nail,
and a large quantity (90 fragments) of animal bone. It was cut by adjacent
pit 704, which was oval in shape, c 3.5 by 4 m, and 0.4 to 0.5 m deep.
Finds included fired clay fragments, coal, a piece of melted lead, an
iron nail, animal bone (73 fragments) and pottery (0.58 kg) dating from
the mid 2nd to 3rd centuries. This feature cut pit 703 and a Phase 2 ditch
from E 15.
Feature 1730 lay at the southern limit of excavation and was badly truncated
and difficult to classify. The cropmark plan did not indicate whether
it extended south of the trench. It was 3 m wide and c 0.8 m deep and
contained a number of small finds including fired clay fragments, several
iron nails, smithing slag, vessel glass, a spindle whorl and lead weight,
as well as 2 kg of pottery of mid 1st to mid 3rd century date. An assemblage
of 147 animal bones was also recovered.
Short gully 2710 ran perpendicular to Phase 3a gully 2175. The western
terminus was unclear, but was thought to terminate just short of 2175.
It extended for 8 m before being truncated by Phase 4 ditch 700 (E 21).
Gully 2710 was 0.8 m wide on average. The pottery (0.86 kg) indicated
a date in the second half of the 3rd century, evidence that this gully
was unlikely to be contemporary with 2175. Small finds were limited to
an iron nail, a miscellaneous iron object and vessel glass fragments.
In the south-western corner of Trench 13 gully 1737 ran NE-SW, overlying
a series of Phase 2 gullies. It contained 1.76 kg of pottery and almost
200 animal bones in addition to fired clay fragments, several iron nails,
a melted lead fragment, copper alloy ligula and copper alloy coin dated
AD 335-341.
A possible corn-drying oven, 1537, lay in the south-east part of the
trench, over-lying Phase 2 ditch 643. It was constructed of two parallel
lengths of limestone 'wall' 1.5 m long and 0.55 m apart. The western wall
turned at right angles at the southern end for a further 0.4 m, but the
corresponding side had been robbed. A single course of stonework survived
and so the 'flue' was only 0.1 m deep at most. It appeared to be a corn-drier
of traditional T-shaped design (Morris 1979, 10) with the stoke hole at
the northern end showing traces of burning. Further signs of burning occurred
at the southern end of the flue. There was no dating evidence for the
structure, but T-shaped corn-driers generally appear from the 2nd to 4th
century AD, the vast majority belonging to the later period. It could,
therefore, belong to Phase 3 or 4.
A second L-shaped corn-drier,1364, was recorded in the salvage area about
20 m north of Trench 13 (not shown on plan) but was not fully excavated.
It would have lain outside of the main enclosure, and could have belonged
to either Phase 3 or 4.
During the early 4th century AD, Building 7 appears to have been partially demolished and replaced by a building with masonry foundations, B 8, which can be described as a modest 'cottage style' villa. Another building to the south, B 9, comprised two rooms, the smaller of which seemed to contain a hypocaust. Dating evidence suggests that these two structures were contemporary, and formed part of a small residential complex. After a short period they appear to have been architecturally unified with the addition of a 'corridor' wall along the eastern side, which probably created a courtyard between the two main structural elements. To the north and east of the villa were a series of ditches and pits associated with the occupation of the building. In the mid to later 4th century, the southern part of the building was dismantled and a substantial ditched enclosure constructed around the remaining elements. Two further rooms were added in the north-east, blocking what had presumably been one of the main entrances. A series of enclosures appeared to have branched off from the main enclosure to the north-east. The final structural phase within Trench 13, dating approximately to the later 4th century, saw the construction of a far more substantial enclosure with a masonry wall running along its inner edge. It is thought that the building and enclosure fell into disuse at the end of the 4th or start of the 5th century.
The successor to Building 7 was a rectangular building with a stone foundation,
originally measuring 13.5 by 9 m in Phase 4a. It partly overlay Building
7, apparently utilising some of the existing walls in its structure. Footings,
where they were best preserved, ranged from 0.6 to 0.7 m wide and were
of a mortared random rubble construction made up of small (0.1 m) pieces
of limestone. They were of variable depth, deepest on the west side (1556)
where they ran along the top of ditch 2502 (Section 116). The north and
south sides had the shallowest footings, the former just resting on the
gravel surface, and the latter largely untraceable, apart from a fragment
on the south-west corner (1591). It is likely that these masonry footings
and lower wall courses supported a plastered timber superstructure.
Extant walling survived only on the western side, apart from a fragment,
2107, reused from Building 7. The wall was of mortared, coursed and faced
limestone blocks with a rubble infill. The walls had been robbed to the
old ground surface, and a few sections had been completely removed, notably
a large section of one of the internal walls (robber Trench 890). The
foundations of the eastern wall, 880, survived along much of its length.
Post-demolition activity on site (from the late 4th/5th century to well
into the medieval period; Fig.
2.2.17) had further taken its toll on the surviving structure. A series
of pits (1905, 1906) in the northern rooms (1 and 7) removed almost all
contemporary and earlier levels. A fragment of wall, 1999, of a later
undated structure caused disturbance on the southern side and particularly
the south-eastern corner. A series of burials cut the eastern wall, causing
disturbance to Room 8.
Surviving demolition debris over and around the building provided evidence
of its character and status. Fragments of whitewashed, rather than painted,
wall plaster were recovered. Other mortar/plaster was recovered which
may have come from flooring. Large numbers of broken roof tiles were present
(see below), probably deriving from this structure. Small quantities of
stone roof slabs indicated that some roofing was in stone - perhaps the
late extension on the north-east side. Several fragments of late Roman
window glass probably also derived from this structure.
Although layers contemporary with this building were excavated (1919,
1558, 1930, 1933, 1908, 725, 1917, 1888, 1889) none could be interpreted
as floors with certainty. They were all of similar character - loose brown
sand loam with little occupation debris, perhaps forming bedding for either
mortar or beaten earth surfaces. The ploughsoil invariably overlay these
layers with a thin scatter of rubble, 693, forming an interface. They
were observed, due to the quantity of 4th-century material they contained,
to be truncated and disturbed floors.
Situated in the north-west corner of the building, Room 1 measured 3.3 m by 2.5 m. The division between Rooms 1 and 7 had been cut by pit group 1905, but a short length of robber trench, 1890, survived, confirming its presence. Layer 725 within this room may have been a disturbed floor layer. It consisted of decayed mortar, sand and gravel in a matrix of brown loam, and contained 870g of pottery, not recorded in detail. A small number of animal bones (25) were recovered from this layer.
Room 2, sited centrally on the west side of the building, was much disturbed by the later drainage gully, 720 (see Phase 4d below), which cut through part of the western wall, 719. It was of a similar size to Room 1, measuring 3.3 by 2.5 m. Layer 1917 on the south side of the drain, like layer 725, may be disturbed flooring. In the north-west corner an area of pitched cobbles, 1913, overlay wall 1912 which ran between Rooms 1 and 2, but abutted the western wall, 719. It would have formed an area of hard standing contemporary with the later drainage gullies. The northern part of the eastern wall, 2638, survived, the remainder cut by robber Trench 1947/2640. The south wall, 2591/1556, ran east-west for 2m before being cut by robber Trench 890. The corner section of this wall was well-defined, suggesting that it had been partly rebuilt when the drain was inserted into Room 2 (see Phase 4d below). A small collection of pottery (0.075 kg) and animal bone (13) was recovered from this room.
Room 3 formed the south-west corner of Building 8. It measured 4.5 by
2 m and a slot, 1991, was cut midway across the room. This feature showed
no trace of being tied into walls 1556 to the west or 1581 to the east,
and may therefore have been a later alteration. It divided the room into
two compartments 2.3 by 2 m and 1.7 by 2 m.
Preservation in this room was poor. The northern part of wall 1581, dividing
Rooms 3 and 4, had been robbed (1947), along with aisle post 1580 of Phase
3 Building 1. The west wall, 1556, was robbed for most of its length,
and the robber trench, 2102, continued for a further 5.7 m beyond the
southern extent of the building, following the line of the west wall of
Building 7. This suggested that a wall existed during Phase 4a, running
towards B 9 and terminating less than I m from its north-west corner.
The disturbed fill, 721, within Room 3 probably predated it and, as there
was no trace of the north wall of Building 7, layer 721 probably underlay
this structure as well. Finds from 721 included a coin dated AD 364-78
(SF 514), two quern fragments (SF 518, 1520), wall plaster, window glass
and part of a late Roman glass vessel (SF 520) along with 1.3 kg of pottery
and 35 animal bones.
This room measured 4.5 by 5 m and lay on the south side of the building.
Initial clearance across the room revealed a large disturbance, 1887,
sealed beneath medieval wall 1999 on the south-east corner, which probably
removed all trace of the villa wall in this area. The east wall of Structure
7, 2107, was reused and formed part of the eastern wall of Room 4. The
rubble footings of Building 8 contrasted sharply with the coursed footings
of the earlier structure. Layer 1558, which covered most of this room,
was thickest north of robber Trench 1977 (B 7), which it sealed. Layer
1920 lay beneath 1558 south of this robber trench and appeared to be associated
with the demolition of Structure 7.
Layer 1558, north of 1977 was removed in three thin spits, 1558, 1919
and 1949. The lower two spits contained 4th-century coins (SF 1204: AD
350-60, SF 1449: 335-41), and a copper alloy finger ring (SF 1205) and
two bracelets (SF 1202, SF 1432). The top layer (1558) produced a fragment
of late Roman window glass (SF 928) and an iron knife (SF 1385). A reasonable
quantity of pottery (2.33 kg) was also recovered, and included 4th century
Oxford colour-coat and New Forest wares. Over 100 animal bones, including
pig and domestic fowl, came from the layers within this room.
Rooms 5 and 6 lay centrally within Building 8. The partition wall, 2639,
was not bonded into wall 1911 to the north, and to the south it was less
deeply founded than the wall robbed out by Trench 890.The absence of floor
layers and related stratigraphy made it impossible to demonstrate whether
wall 2639 was an original partition or a later addition, although the
latter is more likely, as this area may have formed the entrance chamber
into the original villa. Room 5 measured 2.5 m square and Room 6 measured
2.5 by 1.8 m. The foundations of the east wall of room 6, 880, comprised
loose random mortared rubble, with large pieces of pottery found in situ.
The entrance to the Phase 4a/b villa may have been in this section of
wall.
Layers of clean brown sandy loam, 1930 and 1933, lay within the rooms
and were probably contemporary with layer 1558 in Room 4. A small quantity
of pottery (0.2 kg) and fragment of vessel glass (SF 1242) were recovered
from these layers.
Room 7 sited on the north side of Building 8 and measured 5 by 4.5 m. No contemporary levels survived within this room, due to disturbance by later pit complex 1905 / 1906. All walls apart from the division of Room 1, 1890, were respected by this later activity.
Building 9 lay several metres south of Building 8, on the same alignment.
On pottery evidence it belonged to Phase 4a-b (early - mid 4th century),
making it contemporary with the 'cottage' villa. It was L-shaped in plan
and measured 10 m by 11 m. It was divided into two rooms, Room 1 measuring
5 m by 3.4 m and Room 2, 9m by 5.3 m.
Although of late Roman construction, the building was incomplete and later
Roman and medieval features cut through it. Enclosure ditch 700 (E 21)
cut east-west through the middle and wells 696 and 697 removed the northern
part of Room 1. Preservation was poor to the south of ditch 700. There
was no trace of the west wall and the east wall survived only as a robber
trench, 2309. The south wall was represented by part of the `inner gate'
structure, 2348, but stratigraphic relationships here were unclear due
to disturbance and the ephemeral nature of the features (see below). Despite
the disturbance, the northern part of Room 2 had the best surviving floor
levels on the site. A series of stone hearths and pits were probably contemporary
with its use.
Walls and footings, where they survived (principally to the north), were
of similar construction, ranging from 0.5 to 0.7 m wide, with pitched
stone foundations and coursed masonry above. The footings were similar
to those of the later additions to Building 8, although not as well constructed.
Debris, including tile, overlying Building 9 was similar to that of Building
8, although quantities of debris from mortar flooring also concentrated
around the building and in the later ditch 700 (E 21).
This building post-dated Building 7 on the stratigraphic evidence. Robber
trench 1578 cut gully 2114, which cut the south side of Building 7 (2112).
The building must have been demolished during the mid 4th century to make
way for the inner enclosure ditch E 21. It, therefore, went out of use
before the end of Roman occupation on the site.
Room 1, situated on the north-west side, measured 5 m by 3 m. The north
part of western wall, 2118, survived for 2.5 m before it was cut by enclosure
ditch 700. A deposit of limestone rubble, tile debris and mortar in clean
orange sandy gravel, 2116, filled the interior. It was 0.2 m deep and
cut into the underlying gravel. At the south side it undercut western
wall 2117 but as 2117 ran north its footings deepened and the rubble took
on a reddish burnt hue. The later well pits removed the remaining part
of 2117 but this deepening and burning were adjacent to 2134 in Room 2,
interpreted as a stokehole. Just to the south of the stokehole a burial,
800, was later inserted, cutting through wall 2117.
Room 1 then appeared to have housed a small hypocaust system, the flooring
slightly raised and supported by pilae, fed from Room 2. Removal of layer
2116 revealed a fragment of wall, 2518, of mortared limestone rubble assumed
to belong to Phase 3 Aisled Building 1. Aside from mortar and plaster,
the finds comprised 0.5 kg of pottery, over 10 kg of ceramic tile, iron
nails and 38 animal bones. The quantity of tile contrasts sharply with
any other part of buildings 8 and 9, and undoubtedly derived from the
hypocaust system.
To the east of Room 1, measuring 9 by 5.3 m, was Room 2, which was reasonably
well preserved to the north of ditch 700. Layer 2121 spread across the
interior of this area, and contrasted with 2116 to the west in that it
was a mixed brown sandy loam with patches of clay, burning and occupation
deposits. Finds from layer 2121 included iron nails, and iron hook, fired
clay, and a little pottery (0.18 kg), tile (0.07 kg) and animal bone (17
fragments).
A shallow oblong pit, 2174, measuring 1 m by 0.6 m across and 0.18 m deep,
was situated centrally within B 9, and stokehole 2134 lay on the north-west
corner. The stokehole, narrow at its eastern end, widened and dropped
in depth to the west, adjacent to the line of wall 2117. Both 2174 and
2134 contained charcoal and ash. Adjacent to pit 2174 a spread of stones,
which appeared to be intentional cobbling, may have also been a contemporary
element of the building.
Layer 2121 did not survive south of ditch 700, although various features
might have been contemporary with Building 9. A small hearth, 2071, containing
fired clay, was cut by the ditch. Adjacent to this was an associated stone
spread, 2070. In the south-eastern part of the room against the east wall
was a structure sealed by a dark rectangular spread, 2075. This layer
covered a `kidney-shaped' pit complex, 2321 and 2322, with post settings
paired at both ends. The post settings were c 0.4 m in diameter and 0.3
m deep. These elements formed a 14 m² structure that was not keyed
into the wall and was of uncertain function.
Robber trench 1578 removed the north wall of the room and cut the late
Iron Age enclosure ditches. The east wall, 2132, most of which was removed
robber trench 2115, ran along the eastern edge of the ditches. Wall 2348
of the 'gateway' seemed to form the south wall, as its robber trench,
2079, ran westwards for c 2 m, visible as a slight stain, to align with
wall 2117, the dividing wall between Rooms 1 and 2.
Finds from Building 9 were sparse. No personal items that would indicate
domestic use were recovered but a glass bead was recovered from 2322.
The small pottery assemblage ranged in date from the 1st to the 4th centuries,
and included small quantities of Oxford colour-coat and black burnished
ware. Two coins stratified in the top of robber trench 1578 suggested
a terminus post quem for the robbing of the wall at the end of the 4th
century. A total of 223 animal bones was recovered from the building,
under a fifth of the assemblage size from building 8. Even taking into
account the different scales of the two buildings, it suggests that food
preparation, and perhaps consumption, may have been largely confined to
the northern building.
To the east and west of Building 8 were a series of pits, ditches and a well, assigned to Phase 4 a-b on the basis of stratigraphy or pottery dates. The features would not all have been contemporary but are here discussed together.
Building 9 lay to the north of a three-sided enclosure, 16 by 12 m in
size, represented by ditch 1553 (2608, 2637), which had a broad shallow
U-shaped profile up to c 2 m wide and 0.5 m deep. The south-east terminal
was cut by ditch 870 and the north-west arc had been removed by the late
Roman enclosure ditch 700 and 765 (E 21). To the north it was cut by pits
2604 and 2631, and cut Phase 2 ditch 2502 (2628) and a stone lined gully,
2650, probably belonging to Phase 3 (section 133). It cut ditch 877.
A large slab of masonry was recovered from the ditch, along with fired
clay daub and over 5 kg of pottery, including 4th century Oxford colour-coated
wares and late shell-tempered wares. A group total of 280 animal bones
was also recovered this feature. Given the stratigraphic sequence and
pottery dates, the ditch probably belonged to Phase 4a/b, although it
is possible that it could relate to Building 7 of Phase 3d.
Ditch 877 ran SW-NE approximately 7 m to the north-east of Building 8. It was wide with a flat base and measured 0.80 m wide and 0.40 m deep. There was no sign of recutting. It was cut by pit 1577 and was traced for c 5 m before being cut by ditch 1553. It is not securely dated to Phase 4 as the pottery recovered (0.3 kg) indicated a broad 3rd to 4th century date. A total of 83 animal bones was also recovered.
Ditch 870 ran NW-SE from near the north-east corner of Building 8, with which it was contemporary. It was 12 m long and c 1 m wide and 0.4 m deep with a flat base. A shallow gully, 1576, measuring 0.4 wide and 0.25 m deep (not shown on plan), ran parallel along its northern side for a short length but did not impinge. The western terminal of 870 cut ditch 1553. The extent of the eastern end of the ditch was less clear as it was cut by ditch 765 (E 21) and then by post-Roman ditch 500. In plan it appeared to join ditch 875 but the two ditches had different profiles. Over 4.5 kg of pottery from 870 indicated a broad 3rd or 4th century date, and two coins dated to AD 330-346 narrowed this down to the 4th century. Other finds included vessel glass, an iron horse fitting, numerous iron nails and 286 animal bones.
Situated to the north and north-east of Building 8 were two large pits.
Pit 1577, was oval in shape, its southern side truncated by enclosure
ditch 765 (section 1577). It measured 4 by 2.5 m and a little over 0.8
m deep. The sides were gently sloping and the bottom was irregular. Its
fill was a very mixed gravel, loam, silty clay and building rubble, and
suggested an infill of rubbish. A large quantity of finds was retrieved
from it. Pottery (8.2 kg) dated from the late 3rd century or later and
a coin dated to the mid 3rd century. It cut ditch 877 and predated the
late Roman enclosure ditch 765.
Finds from pit 1577 included domestic debris such as bone pins and vessel
glass fragments, and an amphora sherd with possible graffito from layer
3. Almost 400 animal bones were also recovered from this feature. Most
finds from this feature came from the upper layers, 1-3.
P it 1989, further to the west, was square in plan, measuring c 4 m across
and 1.10 m deep. It lay just 2 m north of Building 8 and on the same axis.
It was deeper than the Roman water table and from the conditione of the
fill (Section 118) it generally held at least 0.5 m of standing water.
The lower sides were vertical but the gravel top had eroded, depositing
gravel around the base, represented by layers 7-9. It was unclear how
long the pit stood open. There was no indication of a revetment or covering,
but this could have been removed prior to infilling. It resembled the
deep sunken chamber, 1969, within Building 8. The remains of a wicker
basket, possibly a fish trap, and another fragment of wood were preserved
on the base, predating the primary fill. Environmental samples suggested
that the original function of this pit was the temporary storage of live
fish (see Robinson, section 4.4).
The pit was deliberately filled with limestone rubble, mixed silt and
gravel. The large quantity of finds from the fill probably derived from
redeposited midden material, and included building debris (building stone,
large quantities of nails, lead trimmings, ceramic tile and some mortar),
along with a loom weight, quernstone, bone pins and a large animal bone
assemblage (1617 fragments) that contrasted markedly from most other feature
assemblages from the site (see Sykes section 4.3). Over 11.5 kg of pottery
provided a detailed chronological sequence for the feature (see Booth
section 3.2). In summary, all contexts containing pottery were deposited
during the first half of the 4th century, probably towards the middle
of the century, and there appeared to have been only short intervals between
episodes of filling. The final filling took place in the second half of
the 4th century, not much beyond AD 350. This predated the late Roman
enclosure ditch 765 (E 21). A coin from the surface was dated AD 393-395.
The pit had also been partly overlain by a cobbled surface, 1916, from
which disturbed fragments of human bone were recovered.
This well was sited 8 m to the west of Aisled Building 3, and between
the two late Roman enclosure ditches, E 21 and E 22. The opening was c
1.2 m diameter and depth was 3.6 m, with a dry stone lining reaching to
the baseof clean gravel (section 193). It was filled with gravelly loam
and some limestone rubble. The type of fill and the poorly preserved organic
material suggested that there was at least a 1 m head of water within
the well. The dating of its use and infill was uncertain. An assemblage
of 3.5 kg of pottery was recovered from the feature, that from the bottom
fills post-dating AD 240, and the rest predominantly 3rd and 4th century.
A single coin was recovered from an upper fill, 2, but was dated to the
Flavian period. Other finds included a bone pin and shale bracelet from
layer 3 and a copper alloy razor from layer 5. In total, 429 animal bones
were recovered.
The construction date was uncertain but it was likely to have been in
use during the later 3rd century and part of the 4th century, spanning
Phase 3d and 4a/b. It was unlikely to have continued in use after the
construction of enclosure ditch 780, which would have effectively isolated
it from the main domestic buildings. Poorly preserved organic material
was recovered.
During the early to mid 4th century, probably not long after the original villa building was constructed, an extension was added to the south-east corner, the east wall of which continued southwards for c 13 m before stopping 1.4 m from a 'gateway' structure attached to Building 9 (see below). This would have architecturally unified the two buildings, creating a central 'courtyard' area c 10 by 6.5 m between them. This 'courtyard' was covered by a two layer deposit of black soil, 1929. Both layers respected the northern, eastern and southern boundaries, while to the west, only the lower layer respected robber trench 2102. This suggested that the western wall of the courtyard was removed prior to the demolition of B 9 in Phase 4c. An assemblage of pottery (c 3 kg) along with vessel glass and iron nails was recovered from these layers. In contrast to the main building, no animal bones were recovered.
On the south-eastern corner of Building 8 a small room was added east
of Room 4. Although no stonework survived (except part of the eastern
wall 1587/1589) the robber trench, 2123, clearly indicated that it would
have abutted the original east wall of Building 8, along the line of wall
Trench 890. The room was rectangular and measured 5 by 2 m. No trace survived
of the southern wall. Eastern wall 1587/1589 continued south beyond Building
8 for c 13 m forming a boundary wall associated with the `gateway structure'
(see below). There were two distinct phases of robbing of this wall, robber
trench 1579 being much deeper than 1559. The northern robber trench (2123)
was cut along its length by mid Saxon grave 2104 (Fig.
2.2.17).
Layer 1908 appeared to be a contemporary infill of the room and produced
ex situ fragments of mortar, probably from a floor. It overlay cobble
spread 1990, which abutted wall 2107 of Building 7. Over 1.3 kg of pottery,
along with a copper alloy finger ring (SF 1186) a small amount of vessel
glass and 38 animal bones were recovered from layer 1908. Robber trench
1559 contained over 1.5 kg of pottery, 1.5 kg of tile, 150 animal bones
and a range of other finds including window glass, mortar/plaster, fired
clay, iron nails and coal
Situated on the southern side of Trench 13, running parallel but cutting
into the top ditch 547, was an arrangement of walling, of which at least
part appears to have formed a structure distinct from the rest of Building
9. It was represented by a 7 m long east-west aligned wall, 2348, constructed
of small limestone rubble set in at an angle. The footing width was 0.7
m. At the east end it turned north for 2 to 3 m, and another wall, 2320,
came off 2348 at an angle, forming a diagonal supporting wall. A further
angle of wall, 2311, almost completely robbed and apparently not connected
to 2348, enclosed a small area c 1.25 m² against this eastern return.
The chamber thus formed would have provided sufficient space for a person.
A 1.8 m length of north-south aligned wall, 2344, lay between 2311 and
2310 but their relationship was not established. It was unclear if the
east 'chamber' was part of the original design. Its construction technique
of coursed, mortared masonry, and dimensions (0.3 m wide) were different,
suggesting that it was a later addition.
The date and function of these features is uncertain. Wall 2348 seems
to have formed the southern boundary of building 9, but the relationship
between the two could not be proven, as the north-south robber trench
of room 2 (2309) was very shallow, and could not be traced through to
the junction with 2348. It is possible that the western part of 2348 was
integral with the original construction of B 9, the eastern angled extension
being a later addition, perhaps contemporary with the construction of
the eastern wall leading from building 9, on which it was aligned Although
tentative, this interpretation would place it in Phase 4b, as a possible
gated entrance 1.4 m wide leading into the inner courtyard.
Wall 2348 appeared to have had two distinct phases of robbing. The earliest,
represented by robber trenches 2310, 2320 and 2311, removed the east extension
of the structure. The west wall, which served as the southern boundary
of Building 9, was subsequently robbed by 2079. Robber trench 2079 and
2311 contained mortar flooring, presumably derived from Room 1 of Building
9. This entrance structure, therefore, went out of use during the mid
to later 4th century when Building 9 was demolished and enclosure E 21
was dug. Other finds from the robber trenches included iron nails, vessel
glass and window glass, a bone pin from 2311, and a copper alloy finger
ring from 2310. Just over 0.9 kg of pottery and a small quantity of animal
bone were recovered from the robber trenches.
Building 8/9 At some point not long after the middle of the 4th century AD, the villa complex underwent drastic alteration The southern area, B 9, and eastern boundary wall, 1587/1589, were demolished, and the remaining building, B 8, were enclosed by a substantial enclosure, E 21. Modifications were made to B 8, including the addition of two rooms.
The addition of a block to the north-east corner reinstated Building
8 as a simple rectangular structure measuring 14 by 12 m. The new addition
was formed by an 8 m length of north-south walling, 2648, which then turned
west for 3 m to abut footings 2595. Only the footings of the smaller east-west
section survived, and these were of obliquely pitched limestone rubble,
contrasting in technique with the original construction. Internal wall
2276 was constructed in the same way and was probably contemporary. Wall
2594 abutted the original north-east corner of Building 8 and ran north
for c 10 m, to end at the edge of ditch 700 (E 21), suggesting contemporaneity.
Its function however, remains uncertain.
The additions to the north-east corner of Building 8 resulted in the creation
of two new rooms, 8 and 9, and the modification of Room 8. The northern
wall of Room 8, 2276, was moved 1 m further north and a pitched stone
division, 1928, was built 2.5 m to the south, creating a c 2 by 2.5 m
room, 9. Layer 1888 (equivalent to 1908) within this room was disturbed
by later grave cuts, and human bone was spread throughout this layer.
A large stone packed posthole, 1918, lay close to the western side of
the room. Its relationship to grave 2279 was unclear and so it was difficult
to assign the posthole to Room 9 with certainty. The addition of the pitched
stone division, 1928, altered the size of Room 8 to 3 m by 2 m.
Room 10 in the north-east corner measured 5.5 m by c 2.25 m and was occupied
by two large sunken chambers, 1909 and 1969.
The northern chamber, 1909, was 2.5 m square and 0.8 m deep (section
120). The north-east and west sides were near vertical, cutting close
to the inside edge of the outer walls of B 8. The south side was less
well cut with a tongue of gravel extending into the centre of the chamber,
perhaps marking the position of steps. Traces of clay, 6, were found on
the floor (not on section). The fill suggested rapid infilling but gave
no clear indication of use or duration. The lower fill, layers 7 and 8,
was made up of rubble and eroded gravel. The main fill was fairly homogeneous
and only two separate fill were distinguished, the upper (1/2) a dark
grey gravelly loam and the lower (4/5) a dark grey silty clay. No evidence
of a revetment was found although it would probably have been necessary
to prevent the collapse of the walls.
Most of the finds were of building debris. Two coins dated AD 330-35 and
AD 350-400 were recovered from the top fill (1/2), and over 2.5 kg of
pottery was found spread throughout the layers, providing a general 4th
century date. Other finds included 1.6 kg of tile, building stone, nails,
a bone gaming piece, vessel glass, a whetstone and 145 animal bones. The
character of the fill and the lack of any substantial silting over the
top suggested the chamber was backfilled prior to the abandonment of the
building.
A smaller, deeper chamber, 1969 (section 121), occupied the southern part
of Room 10, separated from 1909 by a small gravel baulk. It measured 2.5
m by 2 m and was 1.5 m deep, comparable to the wells. Like 1909, it was
vertical sided with a flat base. There was no evidence of steps. The south
edge lay c 1 m from wall 2276, dividing Room 9 from Room 10. The fill
differed to that of 1909 in that the lower 0.5 m (layers 4-9) appeared
to have silted during use. It consisted of a series of layers and lenses
of clay, silt and gravel with no apparent organic matter. The upper 0.9
m of fill (layers 1-3) was a homogeneous brown loam with small quantities
of limestone rubble and was possibly a deliberate backfill. There was
no sign of a revetment.
Finds included two Roman coins from layer 4, dated to AD 335-41 and AD
345-53. Large numbers of iron nails were scattered throughout the fill.
Vessel glass and stone roofing slabs were also present, the latter from
1, possibly deriving from the roof of this extension. Over 5.7 kg of pottery
and 232 animal bones were recovered from layers 1 to 6. The lack of demolition
rubble and later silting suggested that 1969 was infilled prior to the
demolition of the building. The function of the two pits was unclear but
they may have been used for the temporary storage of fish, as was suggested
in the case of pit 1989 to the north (see discussion, print Chapter 6).
During the latter part of the 4th century the central area of Trench
13 was enclosed by a number of ditches, 700, 765, 780, 873, 894 and 1766.
The enclosure ditch, E 21, ran close to the west side of Building 8 and
cut across the middle of Building 9. The area enclosed by this enclosure
was 32 m across, with Building 8 situated in the north-west corner. A
2 m wide causeway lay on the south side adjacent to the position of the
'gateway' structure (see above). There was evidence of at least two separate
phases. The earliest was represented by ditch 765 on the north side and
possibly by ditch 894 on the south-east side. It was unclear whether 765
continued round the entire circuit as the later ditches, 700, 780 and
1766, had removed all traces on the west, east and south sides.
Ditch 765 was small in comparison to the late phase ditches, measuring
0.8 m wide and c 0.5 m deep. It was cut by ditch 700 to the west and ditch
873 to the east, and cut rectangular pit 1989 (section 118). Wall 2594
of Building 8 extended to its southern edge. The ditch fill midway along
its surviving length showed signs of deliberate infilling. Whether this
represented an actual causeway across the ditch or was an episode of infilling
was unclear.
Ditch 894, to the south-east, was larger at c 1.1 m wide and 1 m deep,
and ran from the outer late Roman Enclosure (E 22) to 780. It was cut
by both of these ditches and by rectangular pit 874 (see below). Although
780 cut 894, a shallow recut in the top of 894 cut 780, suggesting that
this feature was visible during both phases of the enclosure (Section
145). To the east, adjacent to E 22, the upper layers had been infilled
with rubble.
The later phase of the enclosure, defined by ditches 700, 780 and 873,
and possibly 894, may have been open to the north, unless ditch 765 remained
in use at this time. Ditch profiles and size were consistent at c 1.2
m wide and c 1 m deep (section 116). The western stretch of E 21 (700)
continued north from 765, to the northern limit of the trench. Its relationship
with ditch 501 of E 22 was not clear. The eastern arm, (873, 780 and 1766),
formed an S-shape in plan. The northern terminal of 873 was cut away by
linear boundary 500. It continued south and west to end just short of
the terminal of 700, to form the southern entrance. Assuming that the
recut of ditch 894 was contemporary, there would then have been two large
enclosures to the east and north of E 21, although their exact extent
is uncertain.
Ditch 700, which formed the south and west sides of E 21, was deliberately
backfilled with limestone rubble and the south-western bend also contained
debris eroded in from Building 9. The eastern side ditches of the enclosure
did not contain this rubble component but had also been deliberately backfilled.
An assemblage of over 33 kg of pottery was recovered from E 21, most dating
to the 4th century, but was too mixed to provide a more precise date.
Other finds included three coins dating from the later 3rd to later 4th
century, vessel glass, ceramic tile (c 23 kg) quern fragments, spindle
whorls and some of personal ornaments, including a brooch and pins. A
total of 1966 animal bone fragments was also recovered.
An oval enclosure, c 20 m across, lay south-east of Building 9, cutting
across the original trackway. It was represented by three unconnected
lengths of gully, 1860, 1546 and 819. On average dimensions were 0.6 m
wide and 0.4 m deep. The enclosure was open to the south with a 20 m wide
gap. Another break in the gully was suggested on the north-western side.
Gully 819 terminated on the southern edge of ditch 620, and gully 1860
could not be traced cutting across ditch 547, suggesting an opening on
the same line as and with the same dimension as the trackway.
Finds of 4th century pottery (c 3.2 kg) and a coin dated to AD 300-335
confirmed a late Roman date but the spatial evidence suggested that it
was more likely the enclosure belonged to Phase 4c. The southern trackway
ditches were cut by the enclosure but the latest Roman enclosure ditch
501 (E 22) clearly cut it. A small quantity of animal bone came from gully
1546.
Although well 697 was the latest of the excavated Roman wells, it was most poorly preserved. It had been cut through Room 1 of Building 9 during the later 4th century, but had been heavily disturbed by the construction of well 696 during the medieval period. It was slightly smaller than well 502, measuring 0.6 m in diameter and 1.7 m in depth, but was of similar stone lined construction. Little of the northern side remained intact and only the lowest levels were uncontaminated. Five coins were recovered from this lower fill with a date range of AD 260-341, all of which predate its construction. How long it was in use cannot be determined but its site must have been an influencing factor for the siting of the adjacent medieval well 696. Organic material was recovered from the surviving levels including part of a small wooden bowl. Over 2.7 kg of pottery and 245 animal bones were recovered from this feature.
At some stage following the construction of the sunken chambers, a drain
was cut through the west wall (719) of Building 8 into Room 2 (section
720). The drain, 720, 1922/708, ran westwards cutting across E 21. The
earlier cut, 1922/2708, petered out and was lost, while the later cut,
720, ran into a large well or sump, 2721, which measured c 2.8 m across
and 1.2 m deep. Both phases of the drain measured 0.6 m across and ranged
from 0.2 to 0.3 m deep. This modification was the latest change to Building
8 and may have involved major rebuilding since the surviving walls either
side of the drain were of different widths. Wall 719 was 0.45 m wide and
wall 1556 was 0.55 to 0.60 m wide. Furthermore, wall 1556 was well constructed
with a faced outer corner, in the area just to the south of where the
drain entered the building, possibly suggesting that the room was left
partly open to the west. It appeared that wall 1912, between rooms 1 and
2, was removed at this time, as an area of hard standing, 1913, in the
north-east corner of Room 2 overlay it.
The various cuts of the drain produced a large assemblage of finds, including
over 4 kg of pottery, almost 2 kg of ceramic tile, 335 animal bones, iron
nails, mortar and plaster, stone roofing slates, glass beads, a quernstone
and part of a limestone column. A large number of finds also came from
sump 2721, including over 5 kg of pottery, 1.1 kg of tile, 366 animal
bones, three 4th century coins (latest issue AD 388 - 402), vessel glass,
stone roofing slabs and a copper alloy bracelet. Large numbers of iron
nails and some wooden planking were also recovered.
The latest Roman phase was marked by a large subrectangular enclosure,
E 22, represented by ditch501, clearly visible on aerial photographs.
It was roughly concentric with E 21, but was much larger, measuring c
50 m across. A shallow shelf ran along the inner edge of the ditch (section
501). Along some stretches of the southern and western sides a narrow
dry stone wall survived up to three courses. The ditch was on average
between 1.5 and 3 m across and 0.6 to 1.2 m deep. The shelf in places
was up to 0.8 m across and c 0.25 m deep. There was an apparent gap of
c 5 m in the wall on the southern side, aligned on the entrance to the
inner enclosure. The wall may have continued on the north and east sides,
judging by the presence of a rubble-filled robber trench. The outer side
of the wall was faced, the inner face left rough, and was c 0.4 m wide.
There was no evidence of a bank behind the wall, but modern ploughing
had cut into the Roman ground surface and may have removed any evidence.
An entrance gap of 7.5 m was identified on the north-west side of the
enclosure. The ditch was originally continuous but was modified later
to construct the entrance.
The upper fill of ditch 501 was a distinctive alluvial deposit up to 0.3
m deep. It produced 13th century pottery and a coin dated to AD 1205-1215.
Roman finds from the lower levels included a bone pin, copper alloy thimble,
brooch, querns and vessel glass along with building debris. An assemblage
of 16 kg of pottery, 5 kg of ceramic tile and over 2000 animal bones were
recovered from the enclosure ditch.
This enclosure ditch clearly survived as a prominent earthwork well into
the medieval period and formed the focus for later activity on the site
(see below).
After the villa complex fell into disuse, in the late 4th or early 5th century AD, renewed activity is attested by a number of finds and features, although many of these were not well dated. A group of burials cut through the eastern walls of B 8, and two pits, 1905 and 1906, were dug within Rooms 1 and 7, apparently while the outer walls were still standing to some extent (Fig. 2.2.17: Trench 13 phase 5 features). Two distinct medieval ceramic phases were identified, dating broadly to the 13th - 15th centuries. A stone-lined box well, 696, and section of walling, 1999, were the only structural features associated with this medieval phase.
A group of five east-west aligned inhumation burials overlay the late
Roman villa (800, 1971, 2105, 2129, 2277), and another inhumation burial
(702) lay c 4 m to the west (Fig.
2.2.17: Trench 13 phase 5 features; see Witkin section 4.1). All lay
within grave cuts and most were in good condition apart from skeleton
2277, which had been extensively disturbed post-mortem, probably because
it occupied a shallower grave. A child burial, 702, was represented only
by parts of the skull, vertebrae, pelvis and femora.
The four adult graves cut the eastern part of Building 8. Three were fully
extended supine burials but 2777 was too disturbed to be certain. An isolated
crouched juvenile burial, 800, was cut into the hypocaust room of Building
9 just to the north of late Roman enclosure E21. It may have had stones
deliberately placed around the head. The infant burial was also isolated,
cutting Phase 3 cobbled surface 687 to the east. There were no finds associated
with any of the burials, although radiocarbon dating of three adult skeletons,
2129, 1971 and 2105, confirmed a mid to late Saxon date (see below). The
isolated infant and juvenile burials may have been earlier, possibly contemporary
with the Phase 4 settlement, as infant burials are well attested within
Roman settlement sites (eg Barton Court Farm: Miles 1986, 15).
A complex of pits, 1905 and 1906, cut the northern side of Building 8.
They cut through the internal wall dividing Rooms 1 and 7 but otherwise
appeared to respect the Roman building plan. They were subrectangular
in plan, measuring between 1 m and 2 m across and c 0.5 m deep, with a
fill of stone and silt. The pits clearly intercut but most had initially
cut into natural gravel rather than adjacent pit fills. They may have
been gravel extraction pits as the sterile fills suggested no other function.
The pits were not fully backfilled or had subsided, leaving and a depression
that was levelled by gravel free silt, probably alluvial, which contained
pottery of the 11th - 13th century AD. A Saxon coin of Alfred was also
recovered from this late silting layer, 694. A short distance to the west
another late Saxon coin of Baldred of Kent, dated to AD 823-824, was recovered
from the rubble layer (693) immediately below the topsoil. The pits were
broadly dated to between the destruction of Building 8 in the late 4th/5th
century and the late Saxon/early medieval activity. The coins may have
been contemporary with the burials, allowing a more precise date in the
9th century.
The spatial and ceramic evidence allowed the medieval activity on the site to be divided into two distinct phases. The focus of both phases of activity was identified within Trench 13 and it may be significant that the late Roman enclosure (E 22) would have been a visible earthwork during this period. The poor preservation of the southern side of Building 8 may be explained by the later activity on the site.
The earliest phase of activity was represented by several features and
a scatter of domestic pottery, centred on the site of Building 8. The
pottery was derived from layer 694, which sealed pits 1905 and 1906, the
southern part of Building 8 and the area between Building 8 and Building
9.
Three postholes, 2296-2298, which cut the southern edge of Building 8
contained earlier medieval pottery, and a nearby posthole of similar dimensions
may have been contemporary. A small pit, 1926, which cut the south-western
corner of the robber trench of Building 8, contained similar pottery.
No structures were defined but the pottery suggested limited domestic
activity.
The 13th to 15th century date of the later medieval phase, based on ceramic
evidence, was confirmed by two stratified coins dated to 1205-1215 and
1473-7. Pottery of this phase was more scattered but still confined to
Trench 13. Sections across the west, south and east sides of the late
Roman enclosure ditch produced this type of pottery. In every case it
was recovered from the top ditch fill and in three cases was associated
with alluvial deposits. A 13th century cut half penny also came from the
top fill.
Two features within the central area of Trench 13 were also assigned to
this phase. A wall fragment, 1999, which ran NE-SW across the south-east
corner of Building 8, sealed a deposit containing domestic pottery of
the earlier medieval phase. Well 696 lay some distance to the south-west
of the wall but was constructed on the same axis and cut the late Roman
well 697. It was constructed by inserting a stone-lined box, 0.6 m²,
in the bottom of an excavated hollow. A series of steps led down from
the north-east side The well was 1.9 m deep, the box occupying the lower
0.7 m.
A post-Roman date was confirmed by the recovery of a 15th century coin
from the lowest level of the well. It also produced well preserved organic
material.