SECTION 2.1.2:CLAYDON PIKE WARRENS FIELD ISLAND 3

INDRODUCTION

TRENCH 14
Structure 13
Structure 14
Structure 15
Structure 16
Structure 17
Structure 19
Structure 20
Trench 14- Linear Boundaries

TRENCH 15
Enclosure 8

SALVAGE, NORTH OF TRENCH 14
Structures 18 and 21
Linear Boundaries north of Trench 14


Introduction (Fig. 2.1.3: Plan of island 3 )

The westernmost limit of occupation was identified on gravel Island 3. Excavation on this island was focused on Trenches 14 and 15. In Trench 14 a total of seven structures were examined (S 13 - S 17 and S 19 -S 20). A number of linear boundary ditches (hereafter abbreviated to the prefix LB) traversed the island, at least four of which were dated to the middle Iron Age period (LBs 451, 363, 372, 373), two to the Roman period (LB 377 and LB 414) and one to the post-medieval period, LB 360. In Trench 15, located in a low lying area to the west of Trench 14, an enclosure of probable late Iron Age / early Roman date was exposed. Some salvage work was carried out to the north of Trench 14 revealing two further structures (S 18 and S21).

Trench 14 SU 19070 / 93830

The earliest activity in this area is probably represented by Structures 15, 16 and 17. The north-west facing entrance of S 15 opposed the south-east facing entrance of S 17, suggesting that the two may have been contemporary. Both S 16 and S 17 were cut by features associated with S 20, a penannular enclosure linked to two rectilinear enclosures. At least two phases of construction were recognised for S 20, the earliest incorporating the southern side of enclosure ditch LB 372. A later recut of this ditch, however, cut the backfilled gullies representing S 20, indicating that the enclosure incorporating LB 372 outlived the structure. The western return of LB 372 cut a northeast-southwest aligned boundary ditch, LB 451, located in the northern part of Trench 14. The positions of structures S 13 and S 14 in the chronological sequence were uncertain. S 14 was cut by LB 363, which ran parallel to the southern side of enclosure 372, but no relationship was observed between S 14 and S 20 to the south. S 13 had no stratigraphic relationships with the other structures, but its south-east facing entrance opposed the north-west facing entrance of S 19, perhaps suggesting contemporaneity. S 19 was demonstrably the latest of the structures. It cut the latest phase of enclosure ditch LB 372 and LB 363, which post-dated S 14, appeared to respect the position of its entrance.

Top of page

Structure 13 (Fig. 2.1.4a: Structure 13)

Structure 13 (S 13) was represented by pennanular gully 454, enclosing an area with a diameter of 10.5 m. This feature varied in width from c 0.4 - 0.6 m and was 0.27 m deep and appeared to have been recut at least once. It was well preserved to the west but survived as little more than a stain to the east. No occupation material was recovered from the feature, which was dissected by modern field boundary 360.

A south-east facing entrance was marked by two postholes, 460 and 1057, and a short stretch of gully, 453. The entranceway of this structure opposed the north-west orientated entrance of S 19. The two entrance postholes, positioned 2 m apart were set back c 1.0 m from the gully arc. They consisted of double post sockets, the shallower part to the outside presumably designed to support a smaller, ancillary post. The maximum depth of the postholes was 0.4 m and both contained a sandy loam fill A clay-lined pit, 361, lay within the structure. It was c 0.7 m in diameter and 0.26 m deep and burnt stone was present in its fill.

Top of page

Structure 14 (Fig. 2.1.4b: Structure 14)

Structure 14 (S 14) was represented by three discontinuous gully segments, 365, 385 and 386, which together enclosed an area 10 m in diameter with gaps to the north, south-east and north-west. Ditch 363/393 cut across the structure, removing the northern terminals of 365 and 386. The gaps were 4 m wide on average and may have represented entranceways into the enclosure surrounding the structure.

The gullies had all been recut and significant quantities of occupation debris (91 sherds, 819 g of middle Iron Age pottery and 117 animal bone fragments), principally from gully 365, were recovered from their fill. This gully had the widest profile, 0.8 m breadth and 0.44 m deep, and was c 12 m in length; 385 and 386 had broadly similar dimensions, 0.54 - 0.6 m in width, 0.25 m deep and approximately 6 m in length. The south-east facing entrance of the structure was marked by postholes 383 and 489 / 490, place 2 m apart. These were paired postholes with the shallower socket to the front. Both measured approximately 0.5 m across the long axis and 0.3 m deep, and contained a fill of sandy loam with gravel. A number of clay-lined pits were exposed within the area enclosed by the gullies. Intercutting pits 387 and 388 measured c 0.7 m in diameter and 0.2 m deep. Pit 392, which lay less than a metre from the entrance postholes, was 0.88 in diameter and 0.24 m deep. Pits 389 and 488 were both shallow, flat-bottomed features with no traces of clay linings.

Structure 14 was situated immediately to the north of ditch LB 372, the earliest phase of which had a gap coinciding with the entrance of the structure, possibly sited to provide access to it. The relationship could not be demonstrated stratigraphically but some association is likely. Structure 20 clearly incorporated this earlier phase of LB 372 within its south-western enclosure, allowing for the possibility that the two structures may have been contemporary.

Top of page

Structure 15 (Fig. 2.1.5a: Structure 15)

Structure 15 (S 15) was situated on the southern edge of the settlement area, its position defined by penannular gully 404. The entrance gap was 3 m wide and the area enclosed was 10 m in diameter. The gully measured up to 0.9 m wide, narrowing to 0.4 m towards the western terminus, and 0.22 m deep. The southern stretch of the gully was cut by a Roman ditch, 414. Atypically, the structure had a north-west facing entrance. This unusual orientation suggests that it may have been associated with S 17, situated 4 m to the north-west and with a more typical south-east facing entrance. The opposing entranceways suggest some relationship between the two structures.
A narrow slot, 1071, concentric with gully 404 probably represents a wall line of the structure, defining an area c 8 m in diameter. Only the northern and eastern part of this feature survived, where it was sealed and protected by a Roman ploughsoil layer, 415. The slot measured 0.05 - 0.1 m wide and 0.05 m deep and was clearly aligned with entrance postholes 475 and 922.

The entrance was probably a double post arrangement with a shallower socket to the front. The postholes had a diameter of approximately 0.86 m and were 0.36 - 0.46 m deep and were just over two metres apart. A solitary pit, 939, lay within the area formed by slot 1071. This, in contrast to the clay-lined pits within S 13 and S 14, was relatively deep with vertical sides and a flat base. It measured 0.80 m wide and 0.46 m deep. Occupation debris from the area of the structure was sparse, comprising 60 sherds (533g) of pottery and 49, mostly undiagnostic, animal bones.

Top of page

Structure 16 (Fig. 2.1.5b: Structure 16)

Structure 16 (S 16) was situated close to the east edge of the island, c 6 m north-east of S 15. The main surviving structural feature was penannular gully 366. It formed a roughly oval enclosure measuring 11.5 m x 9 m with a south-east facing entrance. The gully was truncated by features belonging to Structure 20 and, therefore, stratigraphically earlier. The gully was c 0.5 m wide and generally only 0.2 m deep with no apparent recuts.
A 2 m wide entrance into the structure was marked by two paired sets of paved postholes located adjacent to the entrance gap. The larger postholes of each pair, 484 and 487, were c 0.5 m in diameter and 0.29 and 0.19 m deep respectively. The two smaller postholes, 483 and 485 (both c 0.25 m x 0.14 m), were set in front of the main postholes and separated from them by a gap of 0.15 - 0.20 m.
Other internal features consisted of a cluster of six small post or stakeholes close to the eastern side of the entrance. These features, 477 - 482, were small and shallow, with a depth of between 0.6 and 0.8 m and a diameter of between 0.15 and 0.28 m. A centrally positioned oval pit, 476, was 1.9 m long, 1.4 m wide and 0.21 m deep and contained a fill of sandy loam with gravel. Another two small post/stakeholes, 422 and 423, exposed at the back structure are more likely to be associated with an enclosure belonging to S 20 than with S 16. The only finds associated with the structure, three pottery sherds and two animal bone fragments, were recovered from the enclosure gully.

Top of page

Structure 17 (Fig. 2.1.6a: Structure 17)

Structure 17 (S 17) was situated on the south-western edge of the island with its entrance facing north-west, opposing the entrance of S 15. The two structures may have been contemporary. S 17 had two phases of construction, represented by gullies 374 and 375. The earliest phase, 374, was formed by two arcs of gully enclosing an area 7.25 m diameter, with gaps c 1.5 m and 1.8 m to the north-west and south-east respectively. Two large postholes, 904 and 907, were set 2 m apart and approximately 0.5 m from the south-east gap. Both were double post features but 907 had been recut at some stage or had cut an earlier, unrelated feature. Posthole 904 was 0.83 m across and 023 - 0.3 m deep. The central fill contained a considerable number of burnt limestone fragments. Posthole 907 was 0.9 m long and at least 0.3 m deep. A number of smaller postholes, 901-903, 905, 906, 908 and 380, were clustered around the entrance posts. Postholes 901-903 flanked posthole 904 and ranged from 0.3 - 0.5 m in diameter, and 0.18 - 0.28 m in depth. Burnt limestone fragments were included in the fills of 902 and 903. Postholes 905 and 906 lay to the south-east of posthole 907. They were, respectively, 0.28 m and 0.20 m in diameter and only 0.1 m deep. Postholes 908 and 380, to the north west of 407, ranged from 0.3 - 0.4 m in diameter. Posthole 380 was only 0.09 m deep and posthole 908 was 0.23 m deep.
A scatter of additional small postholes and pits were located within the area defined by gully 374. They included clay-lined pits 379 and 381, with similar dimensions of 0.6 - 0.64 m diameter and 0.12 -0.16 m deep. Pit 381 had been cut by linear gully 373 which ran across the north and west of the house site. Pit 378, located within the north- west gap of 474, was oval in plan, measuring 0.9 x 0.27 m. Postholes 910 - 917 and 447 averaged 0.25 m in diameter and 0.1 m in depth.
The second phase gully, 375, was larger than 374. It was not concentric with the earlier gully, cutting it on the north side and extending further southwards to encompass an area c 10 m in diameter. The single entrance gap was south-east facing, but not aligned with the earlier south-east entrance. Two shallow paired entrance postholes, 382 and 909, were set back approximately 1 m from the gap. Posthole 382 was 0.55 x 0.3 m and limestone packed. Posthole 909 was 0.6 x 0.18 m and contained a single large limestone slab. A narrow slot 0.15 m wide connected the gully terminals and may have held a sill beam or other doorway structure. Gully 375 was truncated by a short stretch of curving gully 376, to the south. This may represent a recutting of the southern ditch, or a later phase of construction. It was cut away Roman field ditch 377.
A relatively large assemblage of finds was recovered from the gully fills. This comprised 165 sherds (1.8 kg) of pottery, 291, mostly undiagnostic, animal bones, a briquetage fragment and three fired clay fragments, one possibly from a loomweight. A single pottery sherd was recovered from pit 378, but the pits and postholes otherwise produced no finds. S 17 falls stratigraphically within an early phase of the occupation of Island 3. Gullies 374 and 375 were both cut by an enclosure ditch associated with Structure 20. No secure relationship could be established between linear gully 373 and the penannular gullies but the linear appeared to cut clay-lined pit 381, suggesting that that at least one phase of Structure 17 predated it.
The evidence was insufficient to demonstrate the time span separating the two phases of construction of S 17. Elsewhere on the site structures or enclosures appear more commonly to have been re-established on new ground but S 20 may also have been rebuilt on its original site.

Top of page

Structure 19 (Fig. 2.1.6b: Structure 19)

Structure 19 (S 19) was situated on the west side of Trench 14 between S 13 and S 17. The main surviving feature was a length of curvilinear gully, 364, measuring c 1 m wide and 0.34 m deep. No evidence of re-cutting was observed. Gully 364 was likely, on morphological grounds, to be related to gully 450 to the south. Together the two gullies enclosed an area of c 10.5 m. A linear boundary ditch, 363, ran along the north-west edge of the enclosure and a break in the line of the ditch coincided with the terminal of gully 364, suggesting that S 19 was either adjoined to the gap in an existing boundary ditch or that the structure and ditch were contemporary. Gully 450 truncated linear ditch 416, the earlier phase of the north-west arm of enclosure ditch 372. The stratigraphic relationships, therefore, place S 19 late in the sequence of buildings on Island 3.
Two double postholes, 935 and 936, set 2 m inside the enclosure gully and coinciding with the south-east entrance gap, probably supported entrance posts. They were approximately 0.32 m deep and 0.4 to 0.52 m wide. Close to the entrance was a group of three stake or post holes, 932 - 934, measuring between 0.24 and 0.31 m wide and 0.08 and 0.11 m deep. The area immediately to the north-west was devoid of features but a scatter of pits, postholes and stakeholes lay elsewhere within the enclosed area. Four of the pits were approximately centrally located, but it was not clear whether they were contemporary. Pit 409 measured 0.8 in diameter and 0.26 m deep, pit 410 was 1.1 in diameter and 0.3 m deep and clay-lined pit 938 was 0.6 in diameter and 0.16 m deep. The fourth pit, 408, was larger at 1.5 m in diameter but was of similar depth, 0.3 m, to the other three. It cut a smaller feature, 937.
To the south-west of the pit group was an irregular alignment of three stake or post holes, 457, 458 and 459. Two groups of inter-cutting postholes, 929 and 930 / 940 / 931, lay to the north-west of the pit group. An isolated stone-packed postholes, 461, lying further to the north-west, may also have belonged to S 19. Post and stakeholes 924 - 928 lay beyond the area enclosed by the gullies and were not obvious structural features associated with S 19. But they may have formed a fence line that extended the enclosure northwards. They varied in width from 0.22 to 0.42 m, and 0.14 to 0.18 m deep.
All finds associated with this structure came from the gullies. They comprised 66 sherds (809 g) of pottery and 32 animal bone fragments, 10 cattle, five sheep and one horse.

Top of page

Structure 20 (Fig. 2.1.7: Structure 20)

Structure 20 (S 20) was approximately centrally situated within the occupied area of Island 3. The penannular enclosure was the central element of three linked enclosures, the other two lying to the north-east and south-west. Three phases of construction were identified for the penannular enclosure in the form of three arcs of gully on the east side, 394, 369 and 400. The three penannular gully segments were separated from one another on the north and east side of the enclosure but merged to the west. Where they intercut they were allocated a single context number, 401, and here at least four recuts and three termini were recognised during excavation. The entrance to the penannular enclosure was south-east facing in all phases.
The earliest phase gully, 394, enclosed an area c 11.5 m in diameter and its entrance was the most easterly of the three phases. Where isolated, gully 394 was 0.62 m wide and 0.18 m deep, with a rounded profile. A pair of possible entrance postholes, 947 and 954, were set back c 2.0 m from the entrance, approximately 2 m apart. Posthole 947 was 0.4 m deep and was cut by the later gullies, 369 and 400. Posthole 954 measured 0.6 x 0.4 by 0.22 m and impinged on but had no certain relationship with a curvilinear gully 373 that crossed S 20.
The middle phase gully, 369, was clearly linked to ditches of the ancillary enclosures, its southern terminal opposing the terminal of ditch 368 and its northern end merging with ditch 372. The internal area decribed by gully 369 was c 10.5 m in diameter. Postholes 399 and 435 were set back c 2.5 m from the entrance and may have held entrance posts associated with either 369 or the third phase structure represented by gully 400. Both were stone-packed and c 1 m in length but their shapes were somewhat different. Posthole 399 was roughly circular in plan and 0.2 m deep but 435 was oval and 0.42 m deep. The innermost, third phase gully, 400, defined an area c 10 m in diameter. The gully was c 0.5 m wide and 0.16 m deep, far shallower than gully 369. It nonetheless appeared to represent an enclosure gully rather than a structural feature.
Within the enclosed area were several small postholes, four large post-settings and a clay-lined pit. Postholes 920 and 437 and two unrecorded postholes in between formed a linear arrangement 4 m in length, possibly a fenceline, the return of which crossed into the south-west enclosure, represented by postholes 1059, 411 and 953. As such, the features would have been unrelated to a circular structure and may have pre- or post-dated it. A number of other small stakeholes and postholes, 403, 432, 434, 436, 918, 949, 951 and 957, did not form any obvious pattern. The clay-lined pit, 402, was 0.75 m in diameter and 0.18 m deep. It contained a primary fill of clay overlain by large quantities of burnt limestone, sealed by a clean silt layer.
Located centrally within the structure were four large post pits, 391, 420, 433 and 919. They were 1.3 m in diameter and 0.5 m deep, well above the size of the entrance postholes. They had been filled with stone and gravel packing. Posthole 391 produced a few fragments of animal bone and 420 contained three middle Iron Age pottery sherds. The postholes were evenly spaced and may represent the setting for a 3 m square four-post structure. The axis of this four-post structure would, however, have been askew to a central axis from the entrance to S 20 and its size, orientation and spatial position with reference to the entrance posts suggest that it was likely to have been a free-standing structure unrelated to the circular structure rather than an internal feature of the building. It would, therefore, have either pre- or post-dated S 20.


Enclosures associated with Structure 20

The entrance to the north-eastern enclosure was formed by the terminals of gully 368 and 369, the latter curving round to the south-east to oppose the former, leaving a gap c. 2 m wide. Two postholes, 422 and 423, located adjacent to the terminals, probably represented a gate structure. The postholes had similar dimensions, 0.3 and 0.34 m in diameter and 0.14 m deep. Posthole 422 contained a single sherd of middle Iron Age pottery.
Gully 395 defined the north-western corner of the enclosure. Gullies 395 and 368 had similar dimensions, 1.1 m wide and 0.58 m deep. Where gullies 395 and 369 intersected one another, the latter became deeper and more steep-sided but the fill sequence did not show evidence of recutting. Gully 395 cut the early phase gully, 394, belonging to the central penannular enclosure, suggesting that the north-east enclosure was a secondary addition to S 20. The size of the north-eastern enclosure can only be estimated, at approximately 11 m by 11 m, as the eastern boundary was not located within the marshy area and may have been removed by the Roman or post-medieval ditches (Fig. 2.1.3: Plan of island 3 ).
The south-western enclosure measured 14 m by 7 m and had an entrance to the south-east. The enclosure utilised the first phase of linear boundary ditch 372 as its north-west boundary. The south side was defined by gully 431, later recut as a deeper feature, 371, measuring 1.2 m wide and 0.53 m deep. The eastern boundary was formed by gully 425. This gully intersected the penannular gully, 369, that defined CS 20, and which was recognised as the deepest recut within terminus 401. A large quantity of rubble and occupation debris weres present in the fill at the point where gullies 425 and 369 intersected. A narrow gully, 470, had been cut across the entrance to the enclosure, running between the termini of gullies 425 and 431. It was 0.4 m wide and only 0.16 m deep and may have supported a gate structure or have been designed to block the entrance at some point. Three sherds of middle Iron Age pottery were recovered from its fill. One internal feature, pit 384, appears to have been contemporary with the south-western enclosure. The pit was circular in plan, measuring 0.94 m in diameter and 0.16 m in depth. It produced no finds.
The sequence of relationships between features demonstrably associated with S 20 and other features in the vicinity is complex. A late phase of linear boundary ditch 372 cut gully complex 401, the southern enclosure ditch, 371, and gully 369. The terminal of gully 369 had in turn cut the penannular gully, 366, representing S 16 to the south. The southern gully, 371, had also cut through gullies 374 and 375, associated with S 17. The latest phase of the gullies associated with S 20 gullies appeared to truncate a narrow gully, 373, that ran north-east to south-west across S 17, and curving round northwards across S 20. A line of three postholes, 411, 1059 and 953, formed a north-east to south-west line crossing from the interior of the enclosure and into S 20. These features were of variable width, 0.22 to 0.46 m, and depth, 0.12 to 0.18 m. They may have represented a fenceline with a north-west to south-east aligned return represented by postholes 920 and 437. The posthole alignments did not appear to be contemporary with S 20 and the associated south-western enclosure.
Several of the enclosure ditches and penannular gullies forming S 20 produced large quantities of finds. A total of 735 sherds (15 kg) of pottery was recovered, along with a small amount of fired clay, including two fragments of briquetage. The animal bone assemblage, consisting of 605 fragments, was large but poorly preserved and only 23% of fragments could be identified to species. Context 401 produced two joining fragments of a May Hill sandstone saddle quern rubber.

Top of page

Trench 14 - Linear boundaries (Figs. 2.1.3: Plan of island 3 and 2.1.8: Section through Roman ditch 377)

Trench 14 was bordered by Roman ditches to the east (370), west (377) and south (414), and by a post-medieval boundary to the north (360). Four middle Iron Age linear boundaries, LB 451, LB 372, LB 363 and LB 373, traversed the island.

The Northern Enclosure

The northern part of the settlement area exposed in Trench 14 was occupied by a rectilinear enclosure c 30 m by 20 m in size, represented by boundary ditches 372 and 451. The stratigraphic relationships between the features associated with this enclosure were complex and, in some cases, unclear. Boundary ditch 451 was the most northerly feature, running NE-SW across the island. It appeared to have been dug prior to LB 372, having been cut by it where the two ditches intersected in the north-west corner of the trench. The morphology and alignment of the two features suggest, however, that at some point both were open at the same time, together creating the enclosed area. The entrance to the enclosure was in the north-east corner, where the two ditches curved northwards, running in parallel across the marshy area, forming a trackway approximately 30 m long. Ditch 451 contained eight sherds (38 g) of middle Iron Age pottery.
Ditch 372 was c 1 m wide and 0.56 m deep, and had two distinct phases. In its earliest phase, the southern stretch was interrupted by a break of c 8 m coinciding with the entrance to S 14. This arrangement poses some questions in that the structure would have blocked access to the rest of the enclosure at this point. But access could have been gained to the enclosure elsewhere, including by means of the trackway, and, in any case, the penannular enclosure surrounding S 14 was discontinuous.
The east and west sides of the enclosure were during the early phase aligned on the natural boundaries of the island, a pattern mirrored by the later Roman ditches. The early phase of the western boundary was designated ditch 416.
During the later phase ditch 372 closely followed the earlier alignment on the north-east side of the enclosure. The alignment of the southern stretch of the ditch changed, however, continuing in a south-west direction to cross into the marshy area to the west of the island. This new stretch of ditch was discontinuous, with a break at the edge of the island and another a few metres to the west, where it was redesignated 1047. Ditch 1047 was truncated by phase 1 ditch 444. Ditch 444 was, in turn, overlain by a gravel bank, 1044, associated with a Roman boundary ditch, 377 (see below). The enclosure ditches associated with S 20 were linked to the first phase of LB 372. The S 20 ditches were subsequently infilled and cut by the second phase of LB 372. Overall, 372 contained 57 sherds (396g) of middle Iron Age pottery, and 104 animal bone fragments of which only 15 could be assigned to species, four each to cattle and horse and seven to sheep. Small numbers of quantities of iron, wood and fired clay objects were also recovered.

A linear boundary ditch complex, 363 / 393 / 469 / 106, ran parallel and c 10 m to the north of LB 372. This feature was, on average, 1.7 m wide and 0.77 m deep and preserved evidence of at least two phases of construction. The western part of the boundary ditch was designated LB 363. It appeared to respect the entrance area of S 19, its terminus abutting that of penannular gully 364. The ditch also cut the western side of S 14, but terminated at a point just within the eastern boundary of this structure, where it was truncated by the eastern ditch segment, LB 393. It was not clear whether LB 363 had originally continued eastwards as an earlier version of LB 393 as any trace of it could have been removed by the later ditch.
LB 393 appeared to curve north-west to follow the natural boundary of the island, cutting LB 372, but this relation was unclear as the point of intersection was obscured by LB 360, a post-medieval ditch. The western return of LB 363 was designated 1066. LB 1066 terminated in the marshy area at the western limit of the island. The boundary ditch produced 55 sherds (367 g) of middle Iron Age pottery and 227 animal bone fragments of which only 39 could be assigned to species, 13 to cattle, six to horse and 20 to sheep.
A narrow, shallow gully, LB 373, traversed Trench 14 running on a northeast-southwest alignment, then turning northwards as LB 398, where it was truncated by gully 400 of S 20. The western end of the gully truncated features belonging to S 17, including clay-lined pit 381. The chronological sequence, therefore, was that the gully pre-dated S 20 and post-dated S17. It produced only a single small sherd of middle Iron Age pottery.

The Southern Boundary

The most southerly boundary, LB 414, ran north-east to south-west across the island but on a slightly different alignment to the three major middle Iron Age boundary ditches. 372, 363/393 and 451. At least six recuts were evident, but it was not possible to trace the sequence in detail and to date each phase, but a number of observations were possible. One phase of the ditch clearly cut S 15. Three sherds (33 g) of middle Iron Age pottery and a fragment of saddle quern or rubber were recovered from the lowest levels of ditch 414, suggesting that the Roman phases of the ditch truncated an earlier Iron Age version on the same alignment. The alluvial top fill of the latest recut indicated that it was contemporary with the Roman field system that traversed the islands. This alluvium sealed the earlier ditches. The earlier phases of the ditch continued across the lower ground to the east and were relocated in Trench 16. Where relationships were explored by excavation, it was clear that the early versions of LB 414 were cut by all of the Roman ditches in the vicinity. On the western side of the island the early phase ditch cuts did not extend beyond the point LB 377 was observed to cut them (see below).
Where LB 414 was recut along its northern side, these phases of the ditch were found to be of Roman date. They were not traced beyond Roman field ditch 370 to the east, which either abutted or cut them. At the western limit LB 377 cut several phases of LB 414, and the latest version, filled by alluvial deposits, terminated short of this ditch
A small gully 980, cut the southern edge of LB 414. The gully was, in turn, cut by a short stretch of NNW-SSE aligned ditch, LB 1032, which pre-dated LB 370. An assemblage of 123 animal bone fragments recovered from LB 414 included only 14 pieces identifiable to species, one cat, nine cattle and four horse bones.

The western boundary (Figs 2.1.3: Plan of island 3 and 2.1.8: Section through Roman ditch 377)

The western boundary of the island was demarcated by Roman field ditch LB 377. This feature was c 3 m wide and 0.4 - 0.6 m deep and contained five fills of dark greyish brown sandy silts with sparse gravel. The ditch formed the eastern boundary to a gravel bank, 1044, which was bounded to the west by gully 1055. The bank was between 1.8 and 3 m wide and approximately 0.2 m thick. Both ditch 377 and gully 1055 contained upper fills of alluvium and the gravel bank was sealed by alluvial layer 455. Only two animal bone fragments were recovered from fragments from ditch 377, and a single fragment from bank 1044.
Gully 1055 cut a layer of brown sandy silt with sparse gravel, 1061/1062. This deposit was 0.1 m thick and extended for c 2.8 m, sealed by gravel bank 1044 to the east. The gravel bank also sealed layers 1046 and 497 in the south-western sector of the island. Both layers contained small quantities of animal bone, and 1046 produced a small collection of middle Iron Age pottery and a fragment of Old Red Sandstone, possibly from a quern. Layer 1061 and gravel bank 1044 both sealed curvilinear ditch 444. This ditch contained a small quantity of animal bone and 28 sherds (577 g) of middle Iron Age pottery. North of section 1 (Fig. 2.1.8), ditch 444 was observed to truncate gully 1047 (the western extension of LB 372) and a layer of redeposited gravel, 1045. Both extended into the marshy area to the west and each contained small amounts of middle Iron Age pottery and animal bone. An earlier gravel bank, 466, may have been associated with ditch 444.
Bank 466 was 1.8 m wide and 0.08 m thick. It sealed a sequence of layers that had accumulated in one of the Late Glacial channels that dissected the gravel terraces. The lowest of these layers, 965, was a deposit of grey Chara marl, deposited during the late Glacial period. It was sealed by a layer of peat, 923/464. Environmental sampling of this layer suggests that it formed during the Windermere Interstadial (Late Devensian Zone II) or the very early Post-Glacial (Flandrian Early Zone I). Eight sherds of middle Iron Age pottery from this layer are best regarded as intrusive. A dark grey loam, 474, overlying the peat produced a radiocarbon date (HAR-5411) of 220bc (2170+80) and four sherds (27 g) of middle Iron Age pottery, including one from a burnished bowl. This layer may be equivalent to 1063, located further to the west. Layer 474 was overlain by 472, a dark greyish sandy silt with abundant gravel and a few indeterminate animal bone fragments. Layer 472 was sealed by a sequence of layers composed of brown silt with varying amounts of gravel, 456, 468 and 473. One sherd of pottery was recovered from layer 473, almost certainly belonging to the same vessel as the sherd from 474, and small quantities of animal bone were present in layers 468 and 473. This sequence of layers was sealed by layer 467, which also contained a small amount of animal bone, and was in turn sealed by gravel bank 466.

Eastern boundary

The eastern edge of Island 3 was defined during the Roman period by LB 370. This ditch was on a parallel alignment to LB 377 on the western side of the island. The tertiary fill was an alluvium that also corresponded to the upper fill of LB 377. LB 370 cut a spread of limestone rubble, 921, which overlay LB 414 at the south-eastern corner of the trench. This layer postdated all other features, including NW-SE aligned ditch 943 / 1004. It probably served as an area of hardstanding or infill on the edge of an earlier Roman field. Another rubble spread, 1022, located at the north end of the trench, to the east of LB 370, may have been contemporary. This deposit could not be located precisely within the chronological sequence, but it pre-dated a post-medieval ditch 1021 (not shown on plan, located between ditch 1006 and ditch 968).
It proved difficult to trace some of the ditches through the low-lying area, owing in part to the indistinct nature of some fills and to the complex series of intercutting features that represented boundary and drainage construction affecting the eastern margin of Island 3 throughout the Iron Age, Roman and post-medieval periods. Several Iron Age and Roman ditches that crossed this part of the island were traced further to the south-east in Trench 16.

Top of page

Trench 15 SU 19015 / 99835

Enclosure 8 (Fig. 2.1.9: Enclosure 8)

Trench 15 was located c 10 m to the west of Trench 14 within the low-lying area adjacent to a modern stream. An oval enclosure, E8, was exposed within this 25 m x 25 m trench. The enclosure ditch, 359, was, on average, 3 m wide and 0.9 m deep and there was no evidence of an entrance gap or of recutting of the ditch. The area enclosed measured 14 m by 16 m. The function of the enclosure was unclear and there was insufficient evidence to closely date it. The top alluvial fill contained post-medieval material in addition to a small quantity of Roman pottery. A low gravel bank lay on the outside of the ditch to the west of the enclosure and gravel spreads were also recorded within the enclosed area.

Top of page

Salvage, North of Trench 14

Structures 18 and 21 (Fig. 2.1.3: Plan of island 3 )

During stripping of the site a post ring was exposed 30 m north of Trench 14. Although it was not possible to excavate this feature, a basic record was made. Twelve posts, 1074 - 1085, formed a circle. The posthole diameters were similar, between 0.3 m and 0.4 m. A south-east facing entrance was marked by two larger postholes, 1072 and 1073, set 1.5 m forward of the post ring. They were 0.6 m and 0.7 m in diameter. A single posthole, 1086, was exposed at the south-western edge of the enclosed area.
S 18 may have been constructed either as a post built structure with a projecting porch, or as an post ring within an outer stake wall aligned on the door posts. A stake-walled version would, however, have measured 10.5 m in diameter, giving a significantly larger floor area than the other structures at Claydon Pike.
Penannular gully 1087 (S 21) lay approximately 20 m to the south-west of S 18. The enclosed area produced higher than usual phosphate levels (19020/99860). Only the northern section of the enclosure survived well and the western and southern stretches were visible only as soil staining. The poor preservation makes it difficult to confidently classify S 21 as a house site but the estimated diameter of the gully was 10 m, comparable to the other enclosures. A gap of undetermined width was located on the south-east side. Four postholes, 1088 - 1091, intersect the north-east stretch of gully 1087 and probably belonged to a separate structure.
A linear gully, 1092, ran north-east from (S 21) and parallel to phase 1 gullies LB 363 and LB 372 in Trench 14. It was c 0.5 m wide and was traced for 20 m and appeared to end on the edge of the marshy area to the east of gravel island 3.

 

Linear boundaries north of Trench 14 (Fig. 2.1.2: Trench plan of Warrens Field)

A series of linear features were recorded during the commercial stripping of the topsoil. Of possible middle Iron Age date were three irregular ditches with associated gullies, which extended both north and west out of the observed area. These were contexts 1094, 1095 and more certainly 1098. The only stratigraphic relationship recovered from this area was between context 1095 and context 1097, the latter being later and of Roman date.

Top of page