The excavation of a later Bronze Age site at Coldharbour Road, Gravesend

THE EXCAVATION OF A LATER BRONZE AGE SITE AT COLDHARBOUR ROAD, GRAV ESEND* ANDREWMUDD with contributions by Alistair Barclay, Philippa Bradley, Catherine Underwood-Keevill, Mark Robinson and Fiona Roe SUMMARY 1\vo parallel linear ditches, running for more than 100 m., associated with smaller ditches and gullies are interpreted as a droveway with attached settlement. It is dated by associated pottery and radiocarbon to the later Bronze Age. The site provides the first good evidence for large-scale land division of this period in west Kent. It is suggested that the settlement was an arable counterpart to specialised transhumant pastoralism on higher and lower ground. INTRODUCTION Excavations were undertaken by the Oxford Archaeological Unit during April 1993 at Coldharbour Road, Northfleet, Kent (N.G.R. TQ 638717 - Fig. 1). The work was carried out on behalf of Safeway Stores Plc, ahead of the development of the site for a new supermarket and access roundabout. It was conducted in fulfilment of the archaeological conditions upon planning permission (Ref. GR/91/697) to a specification set by Lawson-Price Development Consultants (Archaeological Consultant Mr Paul Chadwick) and agreed by the County Archaeological Officer for Kent, Dr John Williams, on behalf of Gravesham Borough Council. The site was situated south of Gravesend and immediately north of the A2 and covered about 2.8 ha. (Fig. 2). Most of the site lay in the field to the south of Coldharbour Road, with a smaller area (the proposed roundabout) lying to the north. Topographically, it straddled a * Published with the aid of a grant from English Heritage. 363 b 􀀉 􀀊 􀀪TuRnobyraild ge ;ff,Wells ANDREWMUDD K 􀀋 Ashford􀀌- 64 65 Fig. 1 (a and b )·. Site location. 364 0 10 km. w 􀀆--==--=-==--􀀇 l.,l 0\ u, 21􀀃 fTr.B I I 1, Tr.4 ' ' I 􀀃 n , ----- 􀀄----- xm':)􀀌 􀀎 ,...... Tr.5 ........ ...... I • ..................... ' \'􀀏 . -....................... ......... ,........ Tr􀀌 Tr.6 ----- ----- . 716 ----- ----- . ........ ................ -e --........ --.. Tr.F􀀎 ----- ----- . ......... .................. . --------------------- -"I..? -------------------- ----- ----- .......................... ........ ......... Trenches A-G OAU evaluation '--... 􀀍'-..... ---------- -----, Trenches 1-10 KARU evaluation '--........ ',,.... ----- ----- TO 636 ................... 638 .................... :,.,. I --.. '!;/l416 Fig. 2. Location of trial trenches, excavated areas and major features. N t Tr.8 Marl pit 􀀻 Built up area O 50 m. -- 􀀏 tr1 􀀐 (") 􀀑 􀀒 0 z 0>Ij ► r-< 􀀓 􀀔 to ; 􀀕 Cl.l 􀀖 tr1 Cl.l 􀀗 ANDREWMUDD ridge running north-south which was followed by the former Coldharbour Lane. The height of the ridge is about 49 m. O.D. with the land running almost level to the north and dropping gradually in other directions. The site lies on Thanet Sand with underlying Upper Chalk at depth. The site had been archaeologically evaluated in September 1992 by the Kent Archaeological Rescue Unit and two areas of interest had been found. To the south of Coldharbour Road, (Area 2 of the present excavation) two trenches had revealed a substantial V-profiled ditch, which was thought likely to form part of a large enclosure of prehistoric date. On the opposite side of the road, two trenches in Area 1 had revealed a number of possible features and a pit containing early Iron Age pottery, suggesting at least part of an early Iron Age site here ('Northfleet 1992 (Safeway plc): An Archaeological Evaluation of a Site at Coldharbour Road, near Gravesend, Kent' by Brian Philp). In consideration of these findings, the specification proposed the excavation of further trial trenches to define more precisely the areas of archaeological interest, followed by area stripping on both sites. The approach was to be flexible and staged, with provision for further work contingent upon the results of the previous stages. In Area 1 provision was to be made for stripping to a maximum of 580 sq. m. in the 'footprint' of the proposed roundabout. In Area 2 the site was to be further defined by trenching and areas progressively stripped to expose the site. In the light of the subsequent excavations, the watching brief included the provision for stripping an additional triangle of land (of about 500 sq. m.) on the extreme western side of the site. The rest of the development area was subject to a watching brief during ground works. The specification did not require exhaustive excavations in Area 2, but sufficient examination to define the site's extent, date and character. The results of the excavation were synthesized in an archive (Level III) report which has been presented to the County Archaeologist, Gravesham Borough Council and Lawson-Price. This report has essentially been abstracted from it with certain modifications. ACKNOWLEDGEMENTS Both the excavation and the reports were funded by Safeways Stores. The archaeological work was built into the development operation and was undertaken immediately in advance of construction. It could not have been carried out without the co-operation of numerous contractors, of whom Peter Baxter of Walchli and Wilkins (Quantity Surveyors) and Mike Neagus of Cliffe Construction should be mentioned. Paul Chadwick, Archaeological Consultant with Lawson-Price, was largely responsible for designing the archaeological strategy, while Dr John Williams 366 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND oversaw the excavation and offered valuable advice. The illustrations are the work of Paul Hughes and Karen Nichols of the O.A. U. SITE DESCRIPTION Area 1 (Fig. 3) Methods The 'footprint' of the proposed roundabout was stripped using a 360° tracked excavator equipped with a toothless bucket, taking into consideration the location of the possible early Iron Age pit found in Trench 8 of the K.A.R.U. evaluation. A light scatter of features was found over most of the area. Under 300-400 mm. of modern plough soil was encountered a more compact mid-brown sandy clay loam with moderate amounts of flinty gravel [2].1 This sealed all the features. Up slope (i.e. towards the south) it was a very thin layer (100 mm.), but thickened to a maximum of 250 mm. in the north-west area. This was interpreted as an ancient _!_ I _[_ N t I 6388/7167 10 o 10 m. ---=---==---=--IC:I-=:!_ ____, sca:e 1:500 Fig. 3. Plan of Area 1. 1 Figures in square brackets refer to context numbers throughout. 367 6393/7166i/ / .,, I / / . . ✓,_, scale 1:1250 N t 417 ANDREWMUDD • I -j6367/7168 i \ \ ,j •-j ! ! / / / /I /_, _ _ !_ /􀀈(// ;J I ,, I I 531on1ea // / 4􀀑 #422 / 420􀀒 431 #439 / 460 I L..,_,_ -· -· -·-·-·-6,_,_,_,_,_, I 6367/7167 453 scale 1:100 Fig. 4. Plan of Area 2. 368 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND / / / /I / / -i6371/7170 ,/® / 424 see detail on feeing page --1-"\ ·- 􀀓--·-j-\ I \ I I I I ··-·- · -=j5371°i71s6 􀀋:1.,-:􀀌-􀀍,--0 1 1; \ ,__ 225 o ! 361 I 282 -i6375/7170/ 208 ·-·- I 1._____ / KARU 4A · ·-·-., ,􀀑􀀒 \ _:_􀀇/􀀈1, \ T ·!\"􀀄r 􀀬A􀀭U_4 j \ N t \ _l ,- isa75/7168 [.\,,,;,.􀀈􀀉 􀀊o·-·-· 􀀋 I 347 \ I 209 \ I \ \ _ -􀀐- -􀀑- -!.98 L=]􀀒􀀓 --\ \ 399 ,I ---/ ( \ I I I \ i...---·-i 382 373 370 i _J_ Ii I 6375/7166 i I I ! ! I i_, 10 30 m. -==-===-==-==-==------=======------ 0 scale 1:500 Fig. 4. Plan of Area 2. 369 ANDREWMUDD plough soil, but there was little dating evidence from it. The natural geology under this layer consisted of an orange sandy silt ('brickearth') with patches of clay, sand and gravel. Results The features revealed consisted of discrete pits, many of them irregular in plan. In view of the site's close proximity to the early Iron Age pit found in the evaluation, close consideration was given to them. However, none could definitely be considered to relate to prehistoric settlement here even though knapped flints were recovered from some of them. A number were certainly tree root-holes. Features of possible archaeological interest included Pits 10, 18, 19, 24, 25, 44, and 46, and Post-holes 12, 31 and 55. These tended to be subcircular or oval in shape and between 10 and 30 cm. deep with a degree of regularity which distinguished them from the other features. Selected sections are shown in Fig. 5. The upper fill of Pit 10 consisted largely of burnt flints suggesting that the feature was related to prehistoric occupation, but there was little to indicate its purpose. Area 2 (Fig. 4) Methods Trial trenches were machine-excavated initially in order to examine the extent of the large ditch located in the evaluation. Three trenches, which were subsequently incorporated into the site of the area excavation, and an additional three trenches to the south (D, E and F; Fig. 2), indicated that the ditch was rectilinear rather than curving. Three further evaluation trenches were excavated to the north and east (Trenches A, B and C) to define the limits of archaeological activity up slope. These contained no features. It was decided to strip an area centred on the rectilinear ditch with a margin on either side to reveal associated features. An area of approximately 70 m. x 40 m. was stripped revealing a broadly parallel series of smaller ditches and gullies to the west ( down slope) but no features at all to the east. An additional 500 sq. m. or so down slope was archaeologically stripped during the watching brief in order to examine the western arm of an Lshaped ditch. Following this, a watching brief was conducted during topsoil stripping and groundworks for the development. The conditions of the watching brief were unsuitable for observing subtle features, and the apparent lack of features in the remaining area of the field cannot be considered a reliable record. However, in view of the results of both the O.A.U. and K.A.R.U. evaluations, it is clear that any occupation would have been very light. The overburden everywhere consisted of 250-300 mm. of modern 370 THE EXCAVATION OF A LATER BRONZE A GE SITE, GRAVESEND Area 1 Section 2 44.98m O.D. "" W Reversed 􀀲 a E 7' --------=-9 􀀃 Disturbance 1 O Section 7 46.0Sm 0.D. 7'< SE NW 7' 􀀂9 Section 16 E W 45.37m 0.D. 7'< "'􀀳 7' 48􀀴 41 Section 6 ,.. SE NW 4􀀵9m 0.D. Dlstur􀀶 Section 9 7' SW NE 4􀀷9m o.D. 􀀃611/ 28􀀸 25 o 1 m. ilolwc=-c=-i=-i=-c:::11 Fig. 5. Area 1 sections. plough soil [200] over a lighter brown sandy subsoil of slightly varying character [201-203]. This was of variable thickness, being very thin or non-existent towards the top of the slope, particularly in Trenches B and C, and deepening to about 200 mm. lower down. Few finds other than flints were retrieved from it during stripping. Results The main components of the site consisted of the major rectilinear feature [210] running north-south, a parallel ditch [209] 10 m. to the west of it, and a group of shorter parallel ditches and gullies [205-208 and 334]. Ditch 205 formed an L-shaped feature on the western side. It and Ditch 209 were intruded into by a large ?medieval pit [241]. There were few other features of any description although three shallow charcoal-rich pits were located in the north-west part of the site within the partial enclosure formed by Ditch 205. Descriptions of the prehistoric features are summarised in Table 1. A selective description of the main prehistoric features and the later ones is given below. For more comprehensive descriptions the reader is referred to the 'Level III' archive report. PREHISTORIC DITCHES, GULLIES A ND PIT A LIGNMENTS Ditch 210. This major ditch ran north-south below the crest of the hill following a more or less level course at a little over 46 m. O.D. As a 371 ANDREW MUDD result of the trial trench evaluation, area excavation and watching brief, it was observed to cross the whole field on an approximately north-south alignment. Its course was not dead straight, but had a slight eastward curve at its southern end. In detail, too, it appeared to be slightly sinuous (hence the conclusion from two sections dug during the evaluation that it formed a circular enclosure), but insufficient of the ditch was excavated to establish the full extent of its course. It was examined by three cross sections and one small longitudinal one. Trench 4 of the K.A.R.U. evaluation which crossed the ditch was also re-excavated by machine and the sections cleaned. An additional machine-excavated section (Cut 455) was examined under salvage conditions during the watching brief. This was located at the southern edge of the site near Trench F (Fig. 2). Selected sections are illustrated in Fig. 6. The profile of the ditch was generally V-shaped, although Cut 284 was different and showed a wider, flatter base. Everywhere the ditch sides sloped at 40-50° , with the western slope consistently slightly steeper. There was no clear evidence of recutting in any of the sections, although if this had taken the form of periodic cleaning out of the ditch, it would have been undetectable. The sharp contrast between the middle and lower fills in Cut 284/356 was initially thought to represent a shallow recut in the top of the ditch. However, a consideration of the other sections suggests rather a period of ditch stability following relatively rapid initial silting. This was marked by the deposition of occupation rubbish in Cut 455 and possibly to a lesser extent in Cuts 210 and 225. The ditch was subsequently backfilled, but not made level until finally filled by plough soil. There was not the slightest trace of an earthwork in the field prior to stripping and none of the ditch fills suggested evidence of a bank. Given the large quantity of sediment to be moved in the excavation of this feature, it would seem likely that a bank was constructed, albeit perhaps a low and diffuse one. The site layout suggests that, had this taken place, it would have been positioned on the up slope side. Ditch 209. Ditch 209 ran parallel to 210 and 10 m. from it. It was observed to run to the northern edge of the field. In a southerly direction it was not traceable further than Trench E, but the conditions of the watching brief made it uncertain whether or not it did continue. It was interrupted in the northern part of the excavation, but the intrusion of Pit 241 to the north of terminal Cut 237 made the gap between it and Cut 412 impossible to determine. Like Ditch 210, the middle section of this ditch was the most substantial. Cut 240 (Fig. 7, Section 29) was 2.30 m. wide and 1 m. 372 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND Section 15 E Section 49 E Section 79 SE topso􀀉 subsoil Area 2 Ditch 21 O w 223 4􀀌3m o.o. Section 25 E \ 232 natural sand '---.... ....... 231 ...................... ·--... NW '!. •-=-==-=-c:,-=:i] m. Fig. 6. Area 2 sections. 373 w 45.64m 0.0. 7' ANDREWMUDD deep and the size of the ditch diminished substantially towards the south. The cross profile was normally symmetrical and composite, with a relatively shallow upper slope (20-40° ) and a steep lower slope (60-70° ) resulting in a narrow V-shaped base. Cut 373 showed a simpler steep-sided, round-based profile, and Recut 370 similar but asymmetrical. This section showed the clearest evidence of recutting, with Recut 370 located slightly off the line of the earlier cut. However, it is possible that the recutting can be traced throughout the length of the ditch, with a shallower and wider recut in the top of the earlier ditch resulting in the characteristic composite profile. In addition to the possible two phases of ditch, there appeared to be an earlier, shallower gully on a slightly different alignment. Cuts 402 and 274 were the only sections that could be examined, and elsewhere this gully seems to have been absent or truncated vertically or laterally. Gully 208. This was a gully running parallel to Ditch 210 and 3-4 m. from it for a distance of at least 34 m. Its northern end, near the edge of the excavation area, was truncated by a recent pit. It is not known whether or for how far it continued beyond this, and the negative evidence from the watching brief is not considered reliable. It had a southern terminal and its alignment was continued by a very shallow (0.02-0.03 m.) depression. Further south features 398 and 396 were possible gullies (not excavated) on the same alignment. Ditch 205. This was another ditch generally following the north-south alignment of the others and located 3-4 m. west of 209. However, after running south from the edge of the excavation for 32 m. it turned sharply west at a little over a right angle (100° ) and ran for 21 m. before terminating in a series of pits. It is not clear whether, or for how far, it extended northwards beyond the excavation area. It was generally 0.80-1.20 m. wide and 0.60-0.75 m. deep, but considerably shallower towards the western terminal. Clear recuts in one of the sections (Cuts 311, 317, and 319) added to the variability of dimensions and fills and for this reason the separate cuts are dealt with individually in Table 1. Selected sections are shown in Fig. 7. Pits 433/436 and 417/437 (Fig. 7, Section 70). These two elongated sub-rectangular pits (or short ditch sections) were located at the western end of Ditch 205. They can be considered as integral to this ditch and, indeed, were thought to be part of the ditch prior to excavation. The excavated sections were, therefore, not positioned to determine the relationship between the two pits although there was clearly some degree of overlap. 374 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND Area 2 Ditch 209 Section 29 E Section 42 projected edge W 289 E 43.21m O.D. A---,;:--::=----r-;,"7}r 7' w Section 60 E W 45.51m O.D. 7' -,-------,----,􀀳 7' Section 52 E Section 38 Area 2 Ditch 205 Section 31 o 1 m. lw-lw-:=-c::::J1oc::,-=:=-::jl Ii W 43.23m O.D. N s A\-=􀀴:--;::;-;:;-􀀵􀀶􀀷- 7' 7' --. ,-, ,-----------r---, - 4􀀸6m 0.D. Section 27 4'!:>􀀏 Flint 􀀐􀀑􀀒 Charcoal W 4􀀹2m O.D. 243 Section 70 S N 43.74m O.D. «::;46 􀀺 434 445 433 Fig. 7. Area 2 sections. 375 ANDREWMUDD Pit 435 and Post-holes 434, 438 and 441. These features formed an alignment on the northern side of the terminal of Ditch 205. The post-holes were all small and of a similar size. Including an unexcavated post-hole to the west, they were between 1.75 and 2.50 m. apart. While there was some overlap between these features and the fills of the adjacent pits, no stratigraphic relationships were visible. MEDIEVAL AND PROBABLY MEDIEVAL FEATURES There was a small number of later features (interpreted as being of medieval and/or later date) which appeared to be associated with chalk mining. 'Chalk Pits'. Feature 241 was a large subcircular pit , 12-13 m. in diameter, in the northern part of the site cutting Ditches 205 and 209. A trench on the south-eastern side of it was initially excavated by machine and then dug deeper and cleaned by hand. The sides were found to be irregularly stepped to a depth of 1.60 m. in the western corner of the trench (Fig. 8, Section 28). However, this was not the centre of the feature and it was highly likely that the feature continued deeper towards the north-west. It was filled with a series of highly friable sandy loams [250-253], the upper one of which, in particular, contained a high proportion of chalk lumps. Six sherds of medieval pottery came from this fill. It was considered probable that the feature was a marl pit, tapping deposits of chalk at depth. Alternatively, it might have been a well, but in the absence of occupation here later than the early first millennium B.C. , this is considered less likely. Chalk lumps and flecks were present in the subsoil [201] in this area of the site, and were also present in shallow gullies (such as 243 and 337) which were presumably associated with the mining activity in some way. Two other pits with chalky fills were discovered. Pit 406 at the northern edge of the site was of undetermined size, and 416 , observed at N.G.R. TQ 63857164 during the watching brief , was about 3 .50 m. in diameter at 0.50-1.00 m. below the surface of the natural geology. None of these pits were bottomed, and it can be noted that they must have been of considerable depth. A depth of about 3 m. of sand was seen removed from the top of the hill during ground works for construction without the chalk being reached. The Geological Survey (Sheet 271) gives the general depth of Thanet Sand as 18-24 m. 376 t,J -:i 7'" Area 2 Pit 241 & Ditch 255 Section 28 s N W 250 :·:::􀀖 􀀗;􀀘􀀙:-􀀚􀀛=;􀀜:Jr:􀀝:•􀀞􀀟*-. - -.=;·t'ff{.::􀀗. :. 􀀘·:-· 􀀙.-.-s•• Area 2 Ditch 207 E Section 56 w 7' 45.25m O.D. 7' ·:􀀈􀀉4.t't.::· Area 2 Ditch 208 Section 54 W E 45.44m O.D. 7'􀀕 6 7' 􀀄371 Fig. 8. Area 2 sections. 7'< Area 2 Pit 322 Section 37 w E 45.92m O.D. 7'" Redeposited natural sand 322 1 m. I E 44.25m O.D. 7'< 􀀁 􀀂 􀀃 􀀄 􀀅 􀀆 ► r-< 􀀇 􀀈 ► tI1 en ANDREW MUDD OTHER FEATURES Feature 342 was a short length of shallow, curving gully cut into the top of Ditch 209. It apparently terminated in a small subcircular feature 347. Neither contained any finds. They do not seem to bear any logical relationship to the prehistoric settlement here and are considered more likely to be later. However, the fills did not contain any chalk and they are, therefore, not obviously associated with the 'chalk pits'. TABLE 1: PREHISTORIC FEATURES IN AREA 2 Feature Width Depth m. Cuts Description and Fills (Lgth) m. Ditch 205 1.5 0.75 260/ 275 Brown sandy silt upper fills 261 262 263 276 277 278 279; over grey/yellow lower fills 264 280 281. 1.05 0.73 311 Light brown sandy silt (312 313 (Middle 314) with burnt clay, charcoal and Phase flint in 313; over dark silty sand terminal) (315); over light grey silt (316). 1.1 0.4 317 Single light brown fill (318) between (Early termini 311 and 319. Phase) 0.8 0.53 319 Brown loamy sand (283) over light (Late grey silt (320). Phase terminal) 1.2 0.9 327 328 329 330 (as Cut 366). 1.0 0.75 366 Light grey-brown silty sand 367, over brown silt 368, over yellowgrey silt 369. 1.1 0.65 245 246 light brown silty sand. (Late Phase) 1.1 0.7 247 248 containing quern? frag., 249 265. (Earlier Phase terminal) 0.88 0.17 429 Single mid brown fill (430) with (Early pottery. Cut? by 433. Phase) 378 THE EXCAVATION OF A LATER BR ONZE AGE SITE, GRAV ESEND Feature Width Depth m. Cuts Description and Fills (Lgth) m. Pit 433/436 1.3 0.44 Sub-rectangular pit on alignment of (2.5) 205. Fills 444 445 447 448 449. Pit 417/437 1.5 0.23 Sub-rectangular pit on alignment of (2.8) 205. Fills 418 419 450 451. Pit 206 1.8 0.5 348 331 Offset from 207. Fills 349 350 332 (4.0) 333. Ditch 207 1.2-1.6 0.65 324 385 Aligned with 205. Fills 325 326 387 (9.5) 388 389. Ditch 207 0.6 0.02 390 391 (Early Gully) Gully 208 0.8 0.28-0.5 361 364 Parallel to 210. Fills 362 363 (Neo. (34.0+) 376 404 and Beaker pottery) 365 377 405. Ditch 209 0.6-2.3 0.35-1.0 240 237 Upper fills 238 258 259 411 401 371 257 288 mid brown sandy silt, but dark clay/ 370 silt in Cut 240 (fills 267-270); over (recut) light yellowish grey primary silt. 373 399 Possibly 2 phases throughout. 408 412 343 Ditch 209 0.4-0.8 0.12-0.30 274 402 Discontinuous and/or truncated. Fills (Early Gully) 272 273 403. Ditch 210 1.7-3.4 0.7-1.55 210 225 Upper fills 211 224 226 227 352 357 356 284 297-300 456 brown silty loam, 285 282 virtually identical to subsoil 20 I; 455 middle fills 296 353 354 358 360 212-214 228 301-307 greyish brown sandy or clayey silt, probably deliberate infill; lower fills 215 and 229 thin dark grey-brown silt, probable occupation layers like 458 which yielded large quantities of charcoal, burnt clay and domestic debris; primary fills light grey/reddish grey sand and silt, difficult to distinguish from natural sand. Few finds: 8 flint flakes from 230 232 and 233 (Cut 225). Pit 242 2.5 (2.8) 0.20 242 Sub-rectangular pit cut by 209. Fill 266 grey-brown sandy silt containing 379 ANDREWMUDD Feature Width Depth m. Cuts Description and Fills (Lgth) m. flint pebbles and broken nodules, also 2 scrapers, retouched flake and fabricator. Probable flint-working hollow. Ditch? 255 1.0 c. 0.95 255 Within 'chalk pit' 241 and largely truncated by it. Close to alignment of Ditch 209 but possibly not associated. Pit 322 1.7 0.2 322 Brown clay-loam fill with charcoal, but no in situ burning and no evidence of use as a hearth. Gully 334 0.84 (5.0) 0.22 334 Aligned on 206. Fills 335 336. Gully 378 0.5 0.06 378 Running NE-SW on alignment with 460. Gully 382 0.3 0.10 382 Parallel to 207. Insubstantial and (11.0?) difficult to define. Uneven base suggests it might be hedgerow. Pit 383 0.60 (1.5) 0.06 383 Aligned on 207? Pit 420 1.0 (1.4) 0.18 420 Possible tree-root hole. Pit 424 0.75 0.08 424 Upper fill 425 c. 50 per cent charcoal. Perhaps remains of hearth. Pit 427 0.4 0.04 427 Fill 428 c. 50 per cent charcoal with smears of burnt clay. Perhaps remains of hearth. Pit 435 0.8 0.14 435 On alignment with postholes 434, 438 and 441. Fill 443. Postholes 0.25-0.3 0.09-0.14 434 438 Aligned with Pit 435 on northern side 434 438 441 441 of terminal of Ditch 205. Gully 460 0.5 0.10-0.18 422 431 Running NE-SW on alignment with 439 378, possibly originally continuous. Fills 423 432 440. 380 THE EXCAVATION OF A L ATER BRONZE AGE SI TE, GRAVESEND TABLE 2: FINDS BY FEATURE. AREA 2 Feature Cut Struck Re- L N/BKR BA Pot Other Finds Flint touched Pot (total) Flint 205 260 53 2 1 9 275 311 148 5 6 fired clay 317 4 319 130 7 37 327 44 3 1 4 366 40 5 16 245 12 1 12 fired clay 247 􀀂o I 12 I possible quern frag. 429 8 40 433 2 436 2 437 10 2 417 1 1 206 348 5 331 2 1 207 385 1 390 324 14 I fired clay 208 361 9 5 364 376 404 209 240 13 3 381 ANDREW MUDD Feature Cut Struck Re- LN/BKR BA Pot Other Finds Flint touched Pot (total) Flint 237 4 257 16 1 11 288 370 373 1 399 8 402 408 12 412 6 274 1? 343 210 210 13 6 Fe nail (from topmost fill) 225 44 2 2 Animal rib and tooth fragments 356 284 20 1 fired clay 282 2 455 184 241 241 34 4 3 med. sherds, fired clay, abraded animal bone fragments 235 9 1 1 med. sherd 242 242 7 4 243 243 255 255 2 294 294 309 309 2 382 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND Feature Cut Struck Re- LN/BKR BA Pot Other Finds Flint touched Pot (total) Flint 322 322 13 1 2 quern frags. 334 334 337 337 342 342 345 347 347 378 378 8 382 382 383 383 1 386 386 8 392 392 396 396 398 398 406 406 416 416 1 420 420 1 424 424 427 427 434 434 435 435 2 20 438 438 441 441 460 422 3 431 1 1 439 8 453 383 ANDREW MUDD FINDS A summary of the finds by feature is presented in Table 2. Detailed reports follow below. CHARRED PLANT REMAINS Mark Robinson Four samples were investigated for charred plant remains. These comprised flotations from soil samples of varying volume which had been floated onto a 0.5 mm. mesh. Charcoal fragments were identified where possible to a maximum of ten from each sample. The results are given in Table 3. In addition, an unidentified seed was noted from Sample 3, context 425. The range of charcoal identified is unexceptional, but the absence of charred crop remains is possibly of significance. TABLE 3: CHARCOAL Sample Context Sample size Pomoideae Quercus No. (1) (hawthorn etc) (oak) 1 323 (pit 222) 10 5 1 3 425 (pit 424) 5 2 8 4 428 (pit 427) 5 - 1 5 458 (cut 455 of ditch 210) 7.5 5 10 FIRED CLAY Alistair Barclay A small quantity of fired clay (175 gr.), consisting of amorphous lumps, was recorded from ditch contexts 246, 296, 313 and 325 and from context 251 in one of the 'chalk pits'. One fragment (38 gr.) from context 296 could have formed part of an object. In addition a single piece of ?briquetage was recorded as unstratified. 384 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND PREHISTORIC POTTERY Alistair Barclay INTRODUCTION The excavated assemblage consists of 387 sherds (2702 gr.) of prehistoric pottery. With the exception of seven late Neolithic/Beaker sherds the pottery assemblage can be assigned to a later Bronze Age-early Iron Age date-range. The assemblage is characterised by a high percentage of body sherds with relatively few featured sherds and the average sherd weight is less than 8 gr. METHOD The material was examined using a (x20/40) binocular microscope and fabrics were characterised by inclusion type, size and frequency. The material was quantified by sherd count and weight and is summarised in Table 4. Inclusions F = flint (white, grey and/or calcined) G = grog (crushed fired clay or pottery) Q = quartz or quartzite (white) A = sand (white, black or colourless) S = shell V = voids (mostly leached shell) Fabrics Late Neolithic/Beaker F/LN Hard fabric with sparse angular flint ( <4 mm.) GFA/BKR GAV/BKR Bronze Age GP/BA Soft fabric, with common sub-round grog (<3 mm.), rare angular (<3 mm.) flint and quartz sand Soft fabric with common sub-round grog ( <3 mm.), quartz sand and rare voids. Soft fabric with common sub-round grog and rare calcined flint (<4 mm.) 385 Fl/BA F2/BA F3/BA ANDREWMUDD Hard fabric with common angular flint (<2 mm.) Hard fabric with common angular flint (<4 mm.) Hard fabric with common angular flint (<8 mm.) FGV(S) /BA Soft fabric with common angular flint (<4 mm.), sparse sub-angular grog ( <4 mm.) and rare lenticular voids (?shell). FV(S) / BA Soft fabric with common angular flint ( <4 mm.) and common lenticular voids (?shell, <4 mm.) QF/BA Hard fabric with common angular ?quartz(ite) and flint (<2 mm.) FA/ind Hard fabric with common ill-sorted angular flint (<8 mm.) and coarse-medium sub-round quartz sand CATALOGUE OF ILLUSTRATED SHERDS 1. 362, Ditch 361 (208); Beaker. Impressed paired finger-nail. Fabric GFA/BKR. Colour: ext.: reddish brown; core: dark grey; int.: dark grey. Condition: fair. 2. 302, Ditch 282 (210); Deverel-Rimbury. Comb impressions. Fabric Fl/BA. Colour: ext.: dark grey; core: dark grey; int.: dark grey. Condition: fair. 3. 313, Ditch 311 (205); Deverel-Rimbury. Impressed combed lines. Fabric Fl/BA. Colour: ext.: dark grey; core: dark grey; int.: dark grey. Condition: worn. 4. 367, Ditch 366 (205); Bronze Age. Base from a miniature vessel. Fabric GF/BA. Colour: ext.: brown; core: dark grey; int.: dark grey. Condition: fair-worn. 5. 367, Ditch 366 (205); Post Deverel-Rimbury. Impressed combed lines. Fabric Fl/BA. Colour: ext.: dark brownish grey; core: dark grey; int.: dark grey. Condition: fair. 6. 367, Ditch 366 (205); Deverel-Rimbury, Bucket Urn. Finger-tip impressions below the rim. Fabric F2/BA. Colour: ext.: pale orange brown; core: dark grey; int.: pale orange brown. Condition: fair. 7. 458, Ditch 455 (210); Deverel-Rimbury, Bucket Urn. Applied cordon with finger-tip decoration. Fabric FGV(S) / LBA. Colour: ext.: orange-brown; core: grey; int.: brown. Condition: fair. 386 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND 8. 458, Ditch 455 (210); Post Deverel-Rimbury, hooked rim jar with slashed rim. Fabric FVG(S) / LBA. Colour: ext.: brown; core: grey; int: dark grey. Burnt residues on the interior surface. Condition: fair. 9. 458, Ditch 455 (210); Deverel-Rimbury. Rusticated, all-over finger-tip impressions. Fabric FGV(S) / LBA. Colour: ext.: reddish Orange; core: grey; int.: reddish grey. Condition: fair. 10. 283, Ditch 319 (205); Post Deverel-Rimbury. Finger-tip impression. Fabric: QF/BA. Late Neolithic and Beaker Seven sherds can be assigned to this period by the character of their fabrics. Two sherds in fabric F/LN are different from the flint-tempered material assigned to the later Bronze Age and are tentatively dated to the late Neolithic. The remaining five sherds, in fabrics GFA, GAV/BKR, can be assigned to the Beaker ceramic tradition and include one sherd of Beaker domestic ware with paired finger-nail decoration (Fig. 9, 1). Later Bronze Age Form, Fabric and Decoration The remaining 380 sherds can be assigned to the later Bronze Age. Twenty-three sherds (51 gr.) are from fine walled ( 4-6 mm.) vessels that have been manufactured from fabric F l /BA. Nine sherds from a minimum of four vessels [contexts 302 (Cut 282), 313 (Cut 311) and 367 (Cut 366)] carry simple geometric decoration consisting of bands, squares and triangles composed of impressed comb and combed lines. A base angle (not illustrated; context 265, Cut 247) and the curvature of the body sherds of Fig. 9, 5 indicate that the sherds belong to globular vessels. Some of these sherds (Fig. 9, 2-3) are consistent in fabric and decoration with Deverel-Rimbury globular urns although Fig. 9, 5 is more likely t o represent a fragment from a post-Deverel-Rimbury Class V cup (cf: Barrett 1980, 303 and Fig. 6.19). A total of 321 sherds (2560 gr.) are from coarse ware vessels, Bucket Urns and/or jars, manufactured from fabrics F2-3, FGV(S) and FV(S). This material is again very fragmentary but it includes at least three rims [Fig. 9, 6 and contexts 430 (Cut 429), 458 (Cut 455)], a sherd with an applied cordon (Fig. 9, 7) and base sherds from at least four vessels [context 246 (Cut 245), 318 (Cut 317), 430 (Cut 429) and 458 (Cut 455)]. The rim forms are simple either upright and flattened (Fig. 9, 6) or slightly expanded (context 430). Decoration consists of finger-tip impressions and was recorded either on (context 458) or below the rim 387 4 6 I I ANDREWMUDD I I I I ' I I 3 5 ,.􀀬.􀀭f!,\;>,J11! ;\:::::\:{,,f :t::\:i􀀋:;i ·::t\tif il.f ;\!1:f 1 11J o 1DJdJJ O --===---==-lll50 mm. 7 Fig. 9. Prehistoric pottery. 388 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND (Fig. 9, 6), on an applied cordon (Fig. 9, 7) and on the body (Fig. 9, 9). The cordoned sherds of Fig. 9, 7 are probably from a bucket urn. Eight sherds, possibly from two vessels, from context 458, had been decorated with all-over finger tip impressions. This type of rustication is found on Deverel-Rimbury Bucket Urns, in particular, those belonging to the Ardleigh substyle. The fragmentary and incomplete remains of a hooked rim jar (Fig. 10, 8) came from context 458. Burnt residues are present on some of the sherds from this vessel. A C 14 determination of 1225-989 cal. B.C. (two sigma uncal. 2880+/-65 BP OxA-4719) was obtained from the burnt residue. Similar vessels have been recorded from north of the Thames estuary (Barrett and Bond in Bond 1988, 29; Brown 1988), from Sussex (Barrett 1980, 311) and are a common form in post-Devere! Rimbury assemblages from the Kennet Valley and Upper Thames region (Bradley and Ellison 1975, 103; Bradley 1983-5, 27). Fig. 10, 10 represents part of a fine, slack-shouldered jar with a slightly flaring rim. The rim form of this vessel cannot be easily paralleled amongst the published material from Kent, although the shoulder form, fabric and finger-tip decoration would be consistent with a late Bronze Age/early Iron Age date. DISCUSSION With the exception of a small quantity of late Neolithic/Beaker material the majority of the ceramic assemblage can be placed within the later Bronze Age. Although this material is very fragmentary, both the fabrics and the few feature sherds are consistent with the DeverelRimbury and post Deverel-Rimbury ceramic traditions. Very little Deverel-Rimbury pottery has been published from north Kent, although both Bucket and Globular Urns have been recorded (Philp 1973). It is more common in the middle Thames Valley although Barrett (1973, 121) has previously noted the rarity of Globular Urns within the known assemblages. Post Deverel-Rimbury 'plain-ware assemblages' are notably absent from this area in contrast to 'decorated assemblages' (Barrett 1980). The chronological developments of the Deverel-Rimbury ceramic tradition around the Thames estuary have been briefly discussed by Barrett and Bond (Bond 1988, 36-7) and, if their arguments are correct, then we would expect the Deverel-Rimbury material from Coldharbour Road to date before 1000 cal. B.C. Ditch 210 contained fragments of Bucket Urn and hooked rim jar in direct association; furthermore, the vessels were manufactured from similar fabrics. These fabrics were 389 ANDREW MUDD -􀀃- \ \ 10 0 50 mm. mn--===--===-• I I Fig. 10. Prehistoric pottery. 􀀄 I I 9 different from the bulk of the Deverel-Rimbury material (Table 4). The difference in fabric need not be chronological although it is tempting to see this material as 'transitional' between the DeverelRimbury and Plain Ware traditions of the later Bronze Age. Bradley (1983-85, 28) has put forward a similar argument for material from Pingewood, Berks. Coldharbour Road also produced a small quantity of material (Figs. 9, 5, and 10, 10) which would belong within a 'decorated assemblage' of the late Bronze Age/Early Iron Age transition. 390 uJ \0 ..... Context (Cut) u/s D/2 68 (67) 246 (245) 248 (247) 249 (247) 265 (247) 261 (260) 263 (260) 313 (311) 314 (311) TABLE 4: PREHISTORIC POTTERY BY CONTEXT Fabric groups: number of sherds, weight (gr.) and illustrated sherds F/LN GFA/BKR GAV/BKR OF/BA FI/BA F2/BA F3/BA FV(S) /BA FGV(S)/BA 8, 40gr. 1, 2gr. 12, 292gr. 2, 24gr. 8, 94gr. 1, 8gr. 1, 5gr. 1, 6gr. 1,4gr. 8, 46gr. 4, 9r.g 2, 14gr. QF/BA FS/BA 1, lgr. Totals 8, 40gr. 1, 2gr. 1, lgr. 12,292gr. 2, 24gr. 8, 94 gr. 2, 13gr. 2, IOgr. 8, 46gr. 4, 9gr. 2, 14gr. 􀀃 􀀄 􀀅 􀀆 0 z0 'T.I > t"" 􀀇 􀀈 (I) til (I) t,J \0 tv Context (Cut) 318 (317) 283 (319) 328 (327) 329 (327) 330 (327) 367 (366) 430 (429) 362 (361) 363 (361) 258 (257) 259 (257) Fabric groups: number of sherds, weight (gr.) and i llustrated sherds F/LN GFAIBKR GAV/BKR GF/BA Fl/BA F2/BA F3/BA 4, 9gr. 14, 60gr. 3, 8gr. 1, 3gr. I, 25gr. 2, 6gr. 6, 17gr. 8, 156gr. 40, 472gr. I, 2gr. I, 3gr. I, 3gr. 2,4gr. 4, !gr. 7, 54gr. FV(S) /BA FGV(S)/BA QF/BA FS/BA 23, 56gr. (Fig. 10, 10) Totals 4, 9gr. 37, 116gr. 3, 8gr. I, 3gr. I, 25gr. 16, 179gr. 40, 472gr. 2, 5gr. 3, ?gr. 4, lgr. 7, 54gr. i i v) \0 v) 273 (274) 211 (210) 212 (210) 224 (225) 228 (225) 302 (282) 458 (455) 310 (309) 323 (322) 443 (435) 432 (431) Totals 2, 5gr. 3, 12gr. 2,4gr. 2, 6gr. NB: ****=sherds in lumps of cement-like mud. 1, lgr. 1, 5gr. 3, 14gr. ?l, lgr. 2,4gr. ?l, 2gr. 20,**** 23, 5lgr. 89, 458gr. 1, 8gr. 1, 3gr. 96, 52lgr. 88, 762gr. (Fig. 9, 7 -Fig. 10, 9) 1, lgr. l,44gr. 54, 816gr. 97, 524gr. 88, 762gr. 25, 58gr. 1, lgr. 1, lgr. 3, 14gr. 3, 14gr. 1, 3gr. 1, lgr. 2,4gr. 184,1 283gr. 1, 2gr. 1, 44gr. 20,***** 1, 5gr. 1, 5gr. 2, 6gr. 387, 2702gr. m 􀀂 􀀃 􀀄 0 "I1 > I:"" 􀀅 􀀆 > -􀀂 tn (/.l ANDREWMUDD MEDIEVAL POT TERY Catherine Underwood-Keevill Five sherds were recovered, ranging in date from the eleventh to the thirteenth century. Three sherds from layer 250, the upper fill of 'chalk pit' 241, are of Canterbury Saxo-Norman shelly ware, Canterbury sandand shell-tempered ware early medieval flinty ware. One sherd from layer 236 of the same pit is of Aardenburg type ware. The final, unstratified, sherd is of early medieval flinty ware. INTRODUCTION STRUCK FLINT Philippa Bradley A total of 756 pieces of struck flint and 35 pieces of burnt unworked flint were recovered. The material came from a range of ditches, gullies, pits and tree-throw holes. A small quantity of material was redeposited in the 'chalk pits'. The composition of the assemblage is summarised in Table 5. Selected artefacts are illustrated in Figs. 11 and 12 and described in the catalogue. TABLE 5. ASSEMBLAGE COMPOSITION Area Flake Irregular Core Core Chip Re- Burnt Total waste rejuven- t ouched unation piece w orked flake flint Area 1 16 1 - 3 - 1 - 21 Area 2 586 35 3 64 3 45 35 771 Total 602 36 3 67 3 46 35 792 RAW MATERIALS The majority of the flint is dark brown to grey in colour; there is also a small amount of Bullhead flint (Shepherd 1972). The flint is good 394 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND quality and flakes well. Cherty and crystalline inclusions and some thermal fractures were noted. The cortex is frequently very thin, white or sandy brown in colour. Cortication is generally very light although pieces occasionally exhibit heavy cortication. Sand glossing was noted on some of the flint. The flint would probably have been available locally. TECHNOLOGY The majority of the assemblage is the product of a simple unsystematic technology. Hard hammers were used almost exclusively, resulting in frequent hinge fractures and other mishits. Flakes are often fairly large, reflecting the availability of the raw material. Butts tend to be thick. Occasionally, thermal flakes were used as blanks for retouched pieces. Cores tend to be irregularly worked and do not seem to have been reduced systematically. Core preparation is almost entirely absent, flakes often have cortical butts, indicating that little preparation had taken place. Cores were discarded when no further flakes could be removed, usually because of hinge fractures, for example, Fig. 11, 1. Only three core rejuvenation flakes were recovered indicating that there was little concern with rejuvenating cores. Tested nodules and multiplatform flake cores dominate the assemblage (see Table 6). The average core weight is 144 gr. One or two tested nodules weighing up to 2.5 kg. from gully 460 had only a few flakes removed prior to their discard. On technological grounds, a Bronze Age date for the majority of the assemblage would not be out of place. TABLE 6. CORE TYPOLOGY Area Single Multi- Core on Tested Fragmentary/ Total Platform Platform a Flake Nodule Unclassifiable Flake Core Flake Core Core Area 1 - 1 - - 2 3 Area2 11 15 3 20 15 64 Total 11 16 3 20 17 67 Occasional blades and blade-like flakes came from tree-throw holes in Area 1 and layers within ditches 205 and 210. These and some of the retouched forms may indicate an element of earlier flintwork. Some of the blades and blade-like flakes were soft-hammer struck, perhaps indicating deliberate production rather than chance removals. One or 395 ANDREWMUDD 2 -􀀂I ' \ 3 50 mm. Fi. g. 11 Struck flint. 396 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND two cores exhibited two phases of flaking (e.g. Fig. 11, 2); these pieces tended to be heavily corticated, suggesting the re-use of discarded cores in the Bronze Age. This re-used material may be Neolithic or early Bronze Age in date. Refitting flakes and irregular waste were recovered from tree-hole 52 in Area 1 (two refitting flakes), context 320 of Cut 319/Ditch 205 (three refitting cortical flakes), context 367 of Cut 366/Ditch 205 (two refitting flakes), and context 296 of Cut 284/Ditch 210 (three refitting pieces of irregular waste). No long sequences could be established. The refitting pieces of irregular waste from context 296 emphasize the unsystematic method of flaking, the original nodule had simply been smashed in the hope of producing usable pieces. Burnt unworked flint appears to have had a restricted distribution, occurring sporadically in ditch 205 and a little more frequently in ditch 210. A few pieces were recovered from gullies 378 and 460. The 'chalk pits' contained quantities of redeposited burnt unworked flint. Heavily burnt and calcined flint such as the material from Coldharbour Road is common on Bronze Age sites. RETOUCHED FORMS (Table 7) A pick from a probable tree throw-hole (34) in Area 1 (Fig. 11, 3) was initially thought to be Mesolithic in date. However, it is made from the same flint and is in the same condition as the rest of the material from the site and may be Bronze Age. Similar examples have been found at Grimes Graves (Saville 1981, Fig. 32, F78; Fig. 36, F90). Rather nondescript scrapers are the most frequent retouched form (Fig. 11, 4-5). The blanks for scrapers are often fairly thick and retain much cortex. Retouch is usually perfunctory, end scrapers/end and side scrapers predominate. One or two scrapers from contexts 283 (Cut 319), (Fig. 11, 4) and 313 (Cut 311 - both Ditch 205) were much more TABLE 7. RETOUCHED COMPOSITION Context Heavy Semper Serrated Retouched Dcnticulate Knife Fabricator Piercer Mis􀀄 Total tool Finke Finke cellancous Retouched (Area 1) 33 1 - - - - - - - - 1 Area2 1 20 1 12 2 3 1 3 2 45 Totals 2 20 1 12 2 3 1 3 2 46 397 ANDREWMUDD 6 7 0 50 mm. I:===--===--===' Fig. 12. Struck flint. finely retouched than other examples in the assemblage. Although their dorsal faces were often almost completely cortical, the controlled, neat retouching may indicate that these pieces are earlier in date. A broken flaked axe came from context 283 (Cut 319 of Ditch 205), a form which would be better placed in a Neolithic or early Bronze Age context. A burnt serrated flake from context 328 (Cut 327 of Ditch 205) may also be earlier in date. Retouched flakes are well represented in the assemblage and are often very roughly retouched. The other retouched forms include piercers (Fig. 12, 6), backed knives (Fig. 12, 7), denticulates and a fabricator. These forms would not be out of place in a Bronze Age assemblage. DISCUSSION The assemblage contained few retouched pieces, and the presence of many cores, irregular waste and wholly cortical flakes would suggest that the material was being worked on site. The raw material would have been available locally; Upper Chalk is present within 1 km. of the site. The majority of the assemblage from Coldharbour Road is of Bronze Age technology, and this date is confirmed by the ceramic and radiocarbon evidence. A small quantity of Neolithic or Early Bronze Age material may also be identified. This accords with the ceramic evidence; some Beaker material was recovered from ditches 205 and 208. However, there was generally no distinction in the assemblages from different types of feature. A decline in craftsmanship in Bronze Age flint industries is well established (Ford et al. 1984). It has been demonstrated that flakes become broader and squatter through time (Pitts 1978) and 398 THE EXCAVATION OF A LATER BRONZE A GE SITE, GRAVESEND that knapping was less controlled. The reduction in numbers of different retouched types has also been noted elsewhere; for example, at Knighton Heath, Dorset (Healy 1981) and Micheldever Wood, Hampshire (Fasham and Ross 1978). Middle to Late Bronze Age assemblages reflect the need for expedient a rtefacts, thus retouching is often minimal and cortex is frequently left on the artefact. Locally the assemblage from Coldharbour Road can be compared with the Bronze Age material from the North Ring, Mucking (Healey 1988), some of the material from East Northdown, Margate (Smith 1988) and the Late Bronze Age assemblage from Lofts Farm, Essex (Holgate 1988). CATA LOGUE OF ILLUSTRATED PIECES Catalogue entries are ordered as follows: catalogue number, context number, feature number, artefact type, brief description, condition, small find number (if any). 1. 283, Ditch 319 (205). Multi-platform flake core, rejected because of hinge fractures. Slight sand glossing. 2. 270, Ditch 240 (209). Reworked nodule. Heavily corticated. 3. 33, Tree-hole 32 (Area 1). Pick, tip broken. Slight sand glossing. Sf. 1. 4. 313, Ditch 311 (205). End scraper, steep, neat retouch at distal end. Slight sand glossing. Scraping angle 70-80° . Sf. 16. 5. 270, Ditch 240 (209). End scraper on a core fragment. Slight sand glossing. 6. 283, Ditch 319 (205). Piercer, on a flake from a heavily corticated nodule. 7. 367, Ditch 366 (205). Backed knife, on a sub-circular blank. Minimal retouch around circumference of flake. Fresh condition. WORKED STONE Fiona Roe There are two fragments of saddle quern, both from context 323, the fill of pit 322. One has a slightly concave worked surface, the other a flat one. Both are of the same rock and may have come from the same quern. The rock is dark, purplish-brown fairly fine-grained sandstone, which seems, on macroscopic examination, to be a Tertiary iron-rich sandstone, probably obtained locally. A third fragment of the same 399 ANDREWMUDD rock, from context 248 (Cut 247), lacks any modified surface but may also have come from a quern. There is a further unstratified quern fragment, with one worked surface which has been slightly burnt. This is of a banded, light-coloured stone that has the appearance of an impure flint, which could have been obtained locally from the chalk. RADIOCARBON DATES Andrew Mudd Four samples were selected for radiocarbon dating. One was a conventionally sized sample (Godwin Laboratory, Cambridge) and three were small samples for accelerator dating (Oxford Research Laboratory). Sample No. 2 (OxA-4717) Context 313, Cut 311, Ditch 205. A small. sample from a patch of hazel/alder charcoal within secondary ditch fill. Associated with burnt and struck flints and some pottery representing a discard of occupation material. Dated to 2895 ± 70 B.P., giving a calibrated two sigma date range of 1267-898 B.C. (one sigma; 1136-983 B.C.). Sample No. 4 (OxA-4718) Context 315, Cut 311, Ditch 205. A small sample of hazel/alder charcoal from a lower fill of the ditch. Associated with burnt and struck flints. Dated to 3085 ± 75 B.P., giving a calibrated two sigma date range of 1511-1124 B.C. (one sigma; 1422-1258 B.C.). Sample No. 5 (Q-3255) Context 458, Cut 455, Ditch 210. A conventional sample from large lumps of charcoal within secondary ditch fill. Associated with large quantities of pottery representing a discard of occupation material. Dated to 2835 ± 45 B.P. giving a calibrated two sigma date range of 1135-900 B.C. (one sigma 1050-925 B.C.). Sample No. 7 (OxA-4719) Context 458, Cut 455, Ditch 210. A small sample from charred residue on the interior of a sherd from a late Bronze Age hook rim jar (Fig 10, 8). Dated to 2880 ± 65 B.P., giving a calibrated two sigma date range of 1225-898 B.C. (one sigma; 1126-974 B.C.). There is no need to doubt the validity of any of these results. The 400 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND dates from samples 2, 5 and 7 are very similar and suggest that the most intense occupation was around 1000 B.C. Sample 4 is stratigraphically earlier than Sample 5. It is thought to come from a feature of the middle, rather than the earliest, phase of occupation. INTERPRETATION A ND DISCUSSION Area 1 This area was of little archaeological interest. Most of the features were found to be tree holes or perhaps other natural disturbances. The remaining features were considered to be possibly man-made principally because of their regularity, but they formed no clear pattern and contained very few artefacts and must be regarded as dubious. Some knapped flints and a sherd of Bronze Age pottery were recovered from a few of these features, but the occasional artefact did not appear to be a guide as to whether or not features were man-made. It seems that artefacts were on occasion fortuitously incorporated into tree throw-holes during land clearance. The possible man-made features might also be related to the grubbing out of tree-roots. There is little material to aid the dating of this putative activity. The flintwork was rather undiagnostic (see P. Bradley, this report), but would not in general be inconsistent with a Bronze Age date. There may also have been earlier activity. The excavation was unable to put into context the Early Iron Age pottery assemblage from P it 6, Trench 8 of the K.A.R.U. evaluation. All that can be said is that the feature was either isolated, or related to a site lying exclusively to the north-west of Area 1. Area2 Nature of the site The main archaeological components of the site comprised mutually aligned ditches and gullies, suggesting that the chief function of the site was connected with land division and demarcation. Ditches 210 and 209, each running for well in excess of 100 m. and undoubtedly beyond the limits of the development site, can readily be seen as a ditched trackway 10 m. wide. Ditch 205 was an associated partial enclosure to the west with access to it provided by a break in Ditch 209 (terminal Cut 237). Its north-south arm included at least three recut ditch terminals (Cuts 247, 311 and 319) indicating a series of entrances on this side. Within the area of excavation, there were few finds from the main 401 ANDREWMUDD ditch (210). Relatively large quantities of flintwork and pottery from Ditch 205 and adjacent features strongly suggest occupation, probably of a domestic nature, in this area. This was probably focused on the partial enclosure formed by 205. However, there was no evidence of domestic structures, and it must be assumed that any structurally related features were very insubstantial and had been ploughed out. The material recovered from the ditches clearly related to rubbish disposal rather than in situ use. There were few other features in this area. A short alignment of a pit and four postholes near the western terminal of 205 were probably boundary-related rather than domestic. Two shallow charcoal-filled hollows (424 and 427) are considered to be possibly the remains of hearths. They yielded no diagnostic finds. Another possibly domestic feature, Pit 322, yielded two fragments of quern as well as a small quantity of flint. It should be noted that the large quantity of pottery recovered from the main ditch on the southern side of the field near Trench F during the watching brief (Cut 455, Fill 458), might well indicate a second focus of settlement about 70 m. away. Phasing The general orientation and alignments of the ditches suggest, with one exception, a single phase of site organisation. The pottery and flintwork, supported by four radiocarbon dates, indicate that the maj01 episode of occupation occurred in the later Bronze Age. Three of the dates are very close and suggest a date within a century of 1000 B.C. for the main occupation, (samples 2, 5 and 6: see Radiocarbon Dates, this report). The exception to this arrangement referred to is Gully 460, and its probable continuation as 378, whose north-east to south-west alignment suggests that they belong to a different, probably later, phase. A small amount of earlier pottery might suggest an occupation here in the later third millennium B.C. This earlier pottery consists of five sherds of domestic Beaker and two perhaps belonging to the latf Neolithic period. The scarce and fragmentary nature of the pottery suggests that it might all be residual. Tantalisingly, five of these early sherds came from Gully 208 (Cut 361) which yielded no certainly late1 material. However, the morphological unity of the site and the absence of an assemblage of pottery linking the earlier and later Bronze Age, makes a claim for a late Neolithic/Beaker origin to this boundar) system somewhat tenuous. It is unclear how early the main ditch (210) was and whether it wai at any time contemporary with Gully 208. The clearest dating came from the middle fills of the ditch, particularly in Cut 455 (Fill 458: which yielded considerable quantities of later Bronze Age pottery. Twc radiocarbon dates, one from charred residue on the interior of a vessel 402 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND - - m Early episode Late Bronze Age Mid-late episode Late Bronze Age// / Late episode Late Bronze Age/ / Later prehistoric / Medieval ? / _,_ /􀀄'- I 631011110 / I / / / / /'-. / /1 _,_ I / / -·I - A t9' .. I .-1--· I ·-·􀀐F·-·-·􀀑•--j-\ -i- N t I -ia310,11as 10 \ I I I I i _,_ I 􀀅r-· : I I t ___ .:􀀋---􀀌l. 0 scale 1:500 Fig. 13. Area 2 phase summary. 403 30 m. ·-·- r· \ I I \ I ANDREWMUDD and one from associated charcoal, indicate a date of around 1000 B.C. for these deposits (samples 5 and 7). However, they followed a deep infill of primary silting, possibly followed by a long period of stability. The few finds which came from the primary fills (contexts 230, 232 and 233) were undiagnostic flints. The ditch was positioned 3-4 m. from Gully 208 and conceivably defined the eastern side of a narrow double-ditched trackway or broad territorial boundary (perhaps with an earthwork between the two). However, its dimensions are clearly of a different order to those of 208, and it is perhaps best seen as a replacement for the latter. Gully 208 can, therefore, be interpreted as belonging to an early phase of the site. It can be suggested that an early version of the L-shaped ditch 205 was contemporaneous. The coincidence of the southern terminal of 208 with the corner of 205 11 m. to the west is noteworthy. The development of the site is tentatively suggested in figure 13. This is based on the site's form and observed relationships between features. The pottery, while suggesting some range of date on either side of 1000 B.C., cannot add any weight to this specific interpretation. Three episodes of ditch digging were recorded in the central part of the north-south arm of Ditch 205 (Cuts 317,311 and 319). The pottery recovered here was exclusively later Bronze Age. Cut 317, the shallowest, was considered to be the earliest. It might have been contemporary with Cut 429, near the western terminal which was also very shallow and yielded pottery of later Bronze Age date. Relatively early, but not closely dateable features include Gully 390 (an earlier version of 207?) which followed the southerly alignment of 205. Gully 382, located 9 m. to the east of 390, is possibly very similar as regards its alignment and dimensions and it, too, might be early. The unexcavated gullies 396 and 398, just to the north and approximately on the alignment of 208 could also be early features. An early, heavily truncated version of Ditch 209 is also apparent (Cuts 402 and 274). It is uncertain whether or not the two sections of gully were separated by an entrance mirroring the arrangement of 205, 390, 208 and 382. Ditch 209, located 10 m. from 210 and 3 m. from 205, was undoubtedly contemporary with both these features. Excluding the earliest cuts already mentioned (Cuts 274 and 402), it was probably recut at least once, although this was hard to determine from most of the sections. Its northern terminal, 237, closely mirrored terminal 311 of the middle phase of Ditch 205. The ultimate phase of prehistoric occupation seems to have been marked by the laying out of Gullies 460 and 378 on a north-east to south-west alignment. This presumably occurred after 205 and 209 had gone out of use. The dating of these gullies is very tentative, but the 404 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND relatively large quantity of flint recovered suggests that they were prehistoric and did not post-date the main occupation by a long period. Gully 378 probably respected Ditch 210, which would have remained as an earthwork for a considerable time, and perhaps even until the medieval period. The later 'chalk pits' do not respect the prehistoric occupation. While some early medieval pottery was recovered from the excavated part of 241, the dating of the origin of the features must remain obscure. They might come under the label 'denehole' - a class of deep shaft which appear peculiar to Kent and south Essex and which seem to have had a wide range of dates and purposes (Spurrell 1881 and 1882; V.C.H. Kent 1908). However, there appear to be some problems over definition. Spurrell was clear that deneholes, with deep narrow shafts, were distinct from marl pits, which were broader and shallower. The definition of a denehole in the V ictoria County History of Kent excludes 'chalk wells' and other forms of mine which were dug 'for the sake of the material extracted'. Deneholes by this definition were deep shafts used as secret hiding places, storehouses, or for other domestic purposes. Examples of shafts which might be comparable to the Coldharbour Road features have been reported from Stankey Wood near Dartford Heath (Spurrell, op. cit.; V.C.H., op. cit.). Between 40 and 50 'deneholes' were found here, all passing through Thanet Sand to reach chalk at depths between 40 and 70 ft. (14-24 m.). According to the V.C.H. these were not 'chalk wells' because there was plenty of bare chalk within a mile which would have been far easier to extract (V.C.H., op. cit., 455). The Coldharbour Road features pose similar difficulties of interpretation since surface chalk is accessible quite locally. However, it is possible that factors such as land tenure and traditional rights and uses played an important role in allocating such resources. A medieval date for these features is perhaps the safest guess, although prehistoric and Roman dates have been claimed for some deneholes. It can be noted that Pliny relates that shafts for the extraction of chalk were dug in Britain, sometimes to depths of 100 ft. (ibid., 446). The Site in its Setting The site lies on the dip slope of the North Downs within 3 km. of the River Thames at Northfleet. This topographic zone, which is labelled the 'Foothills' by Everitt, has historically been the richest and most populous part of Kent and its most important cornland (Everitt 1986, 45). Upper Chalk is the general geology of the area and is extensive to the south and east. Nodular flint as a raw material would have been available quite locally. The site is actually located on Thanet Sand. This would have been easily cultivable although, as the presence of the 405 ANDREWMUDD ?medieval marl pits would indicate, the soil might have been rather poor in agricultural terms. On the higher Downs, 3-4 km. to the south, the chalk is capped with Clay-with-flints. In general terms, the area is not well-watered. There does not appear to be a natural source of water in the immediate vicinity of the site. Spring-line sources at Springhead (2 km. to the west) and probably another about the same distance to the north are the nearest streams, but it is possible that urban development and de-watering over the past hundred years have extinguished a closer supply. There are no known contemporaneous sites in the neighbourhood. Crop-marks, about 200 m. to the south, showing rectangular ditched fields are undated. They are not on the same alignment as the Coldharbour Road ditches and may not be associated. Fieldwalking in this area (by the O.A.U. in connection with the Rail-link Project) did not locate any particular concentration of Bronze Age material. GENERAL DISCUSSION Later Bronze Age settlements are rare in Kent. A number of sites have been examined in the east of the county and pottery sequences spanning the second/first millennia B.C. are beginning to crystallise for this area, although the evidence is often fragmentary (MacphersonGrant 1991; 1992). The site at Netherhale Farm, Thanet, would seem to be the closest comparable example in terms of date and form although details have not been fully published. Here a complex sequence included a series of enclosures/field boundaries, perhaps forming part of a wider system of fields and droveways, of Late Bronze Age date (ibid. 1992, 57). Other sites ascribed in broad terms to the Late Bronze Age include the enclosures at Highstead, Chislet and Mill Hill, Deal. These, like Lofts Farm and Mucking North Rings, are of distinctive form and a recent appraisal gives them a slightly later transitional date (850-600 B.C.) (ibid. 1991, 39). It is uncertain whether this type of site would have formed part of a settlement pattern contemporary with Coldharbour Road. In west Kent the only Late Bronze Age settlement identified is that at Hayes Common (Philip 1973; Drewett et al. 1988). Here possible grain storage pits, quernstones and loom-weights suggested a small mixed farming settlement. Its situation on the acidic Blackheath and Woolwich Pebble Beds was perhaps evidence of a general intensity of settlement and land use elsewhere (Drewett et al. op. cit., 116-18). The type of site found at Coldharbour Road, closely associated with largescale land division, appears to be the first of its kind to be recognised in the west of the county. 406 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND Bronze Age land division is becoming increasingly recognised in many areas of Britain. Perhaps the nearest example comes from Mucking, where two large ditched fields were integrated within a ringditch scatter and probably predated the Late Bronze Age South Ring (Jones and Bond 1980). No contemporary settlement was identified. Possibly Bronze Age ditched fields have also been located further up the Thames valley on the gravels of north-west Surrey (Longley 1976) and Late Bronze Age boundary ditches associated with settlement have recently come to light at Uxbridge (Roberts, in preparation). Further evidence of settlements associated with fields and trackways comes from the South Downs (Drewett 1982). Here, Later Bronze Age settlements consisted of small, scattered groups of houses which are suggested to relate to joint family farmsteads engaged in mixed farming. These were located within rectangular plots defined by lynchets. It has been suggested that this type of site, presently confined to the South Downs, might originally have had a widespread distribution (Drewett et al. 1988, 96). The Coldharbour Road site offers many points of comparison with another site showing settlement closely integrated within a system of land division - that at Fengate, Cambridgeshire (Pryor 1980). Here droveways and enclosures incorporated domestic structures scattered around the ditch system . The dating of the system, which was considered to have its origins in the later third millennium B.C., is probably rather earlier than the inception of the Coldharbour Road ditches, notwithstanding some uncertainty concerning the date of the earliest episodes here. Like Fengate, Coldharbour Road was a 'coaxial' system of land division in which the main boundaries were established first before further apportionment took place. As at Fengate too, it is possible to see the linear ditches as forming a droveway for cattle or other livestock which would have connected higher and lower ground as part of a system involving transhumant pastoralism. Certainly, in historical times, the interdependence of marsh and upland in a cycle of transhumant pastoralism has been a prime feature of the agricultural economy of Kent, although the emphasis has been on sheep rather than cattle (Everitt 1986, 34-8). It is clearly implausible to suggest a Later Bronze Age origin for such a pattern, particularly in view of the instability of the north Kent coastline - the present marshes having developed only since Roman times (Devoy 1980; Evans 1953). However, it is likely that marshland resources would have been exploited, if they had been available. Sedimentary and palaeoenvironmental evidence indicates that the Bronze Age was a period of marine transgression which followed a Neolithic regression lasting until c. 4000 B.P. (Devoy op. cit.). Devoy suggests that at this time the intertidal zones of the Thames estuary would have been areas 407 ANDREWMUDD of hunting, fishing and gathering, and also adds that the salt-marsh and mud-flat ecosystem would have been 'an environment naturally suited to livestock grazing' (ibid., 144). Thus, it is possible that the Coldharbour Road 'droveway' directed movement to and from the coastal marshes, which might have supplied both grazing and other resources. The evidence is scarcely sufficient to provide a basis for suggesting the economic role of Coldharbour Road. However, there are grounds for conjecturing that, unlike Fengate, the site was not oriented towards pastoralism. The soils here, which are very sandy, would have been illsuited to grazing and a predominantly arable regime seems more likely. As far as site morphology indicates, the single partial enclosure does not appear designed to control livestock (although admittedly fences or hedges could have been used). The absence of animal bones is probably accounted for by soil conditions, but the presence of quern fragments does suggest some cereal processing, although no cereal remains were recovered (Table 3). Coldharbour Road might, therefore, be interpreted as a counterpart to a Fengate-type pastoral site, with a droveway used to channel livestock through possibly arable land between summer and winter grazing. It could be seen as a response to similar social and economic pressures on land which resulted in the need for intensive land management and boundary definition. While there appear to be no comparable sites as yet identified in west Kent, this type of site would fit in with a general trend which sees, from the mid-second millennium B.C., a change in economic organisation from economically independent units focused on ceremonial sites to highly organised mixed farming and specialised units (P. Drewett et al. 1988, 87). The lower Thames Valley would appear to be a key area in this economic organisation, but the existence of substantial and dense settlement here has been inferred, almost exclusively, from the locations of stray bronzes (P. Drewett et al. 1988, Fig. 4.1). Actual settlement sites have proved elusive. However, the results of excavations at Runnymede Bridge and P etter's Field in northwest Surrey have suggested the importance of the lower Thames as a zone of trade and contact, and perhaps also of specialised pastoralism (ibid., 110-11). This would imply a need for the production of agricultural surpluses elsewhere and paths of contact between the different zones. Coldharbour Road perhaps furnishes indirect support for this model. It provides evidence for a large-scale organisation of land use, which may have had an arable emphasis, and it hints at the importance of the routeway between the Downs and the north Kent coast which would have been an important element in the articulation of such an economic system. 408 THE EXCAVATION OF A LATER BRONZE AGE SITE, GRAVESEND BIBLIOGRAPHY Barrett 1973 1980 Bradley 1983-85 Bradley and Ellison 1975 Bond 1988 Brown 1988 Devoy 1980 Drewett 1982 Drewett et al. 1988 Evans 1953 Everitt 1986 Fasham and Ross 1978 Ford et al. 1984 Healey 1988 Healy 1981 Holgate 1988 Jones and Bond 1980 Johnston 1983-5 J. Barrett, 'Four Bronze Age cremation cemeteries from Middlesex', Trans. London and Middlesex Archaeol. Soc., xxiv (1973), 111-34. J. Barrett, 'The pottery of the later Bronze Age in Lowland England', PPS, xlvi (1980), 297-319. R. Bradley, 'Prehistoric pottery' in (Ed.) M. Bowden, 'Excavations at Pingewood', Berks. Arch. Journ., lxxii (1983-85). R. Bradley and A. Ellison, Rams Hill, BAR, Brit. Ser., xix (1975). D. Bond, Excavations at the North Ring, Mucking: a Late Bronze Age Enclosure, E. Anglian Archaeol. Rep., xliii (1988). N. Brown, 'A late Bronze Age settlement on the boulder clay plateau: excavations at Broad Green 1986', Essex Archaeol. Hist., xix (1988), 7-15. R.J. Devoy, 'Post-glacial Environmental Change and Man in the Thames Estuary: a Synopsis', in (Ed.) F.H. Thompson, Archaeology and Coastal Change, Soc. of Antiquaries Occasional Paper (New Series), i (1980). P. Drewett, 'Later Bronze Age Downland Economy and Excavations at Black Patch, East Sussex', PPS, xlviii (1982), 321-400. P. Drewett, D. Rudling and M. Gardiner, The South-East to AD 1000, (1988). J.H. Evans, 'Archaeological Horizons in the North Kent Marshes', Arch. Cant., !xvi (1953), 103-46. A. Everitt, Continuity and Colonization: The Evolution of Kentish Settlement, (1986). P. Fasham and J. Ross, 'A Bronze Age Flint Industry from a Barrow in Micheldever Wood, Hampshire', PPS, xliv (1978), 47-67. S. Ford, R. Bradley, J. Hawkes and P. Fisher, 'Flintworking in the Metal Age', Oxford J. Archaeol., iii(ii) (1984), 157-73. E. Healey, 'The Flint', in D. Bond, Excavations at the North Ring, Mucking, Essex: a Late Bronze Age Enclosure, E. Anglian Archaeol. Rep., xliii (1988). F. Healy, 'Description and Analysis [of the flint industry]', in F. Petersen, The Excavation of a Bronze Age Cemetery on Knighton Heath, Dorset, BAR, Brit. Ser., xcviii (1981), Oxford, 157-66. R. Holgate, 'The Flint', in N. Brown, 'A Late Bronze Age Enclosure at Lofts Farm, Essex', PPS, liv (1988), 209-301. M. U. Jones and D. Bond, 'Later Bronze Age Settlement at Mucking', in (Eds.) J. Barrett and R. Bradley, Settlement and Society in the British Later Bronze Age, BAR, Brit. Ser., lxxxiii (1980), 471-82. J. Johnston, 'Excavations at Pingewood Berkshire', Arch. Journ., lxxii (1983-5). 409 Longley 1976 Macpherson-Grant 1991 Philp 1973 Pitts 1978 Pryor 1980 Roberts (in prep.) Saville 1981 Shepherd 1972 Smith 1988 1992 Spurrell 1881 and 1882 V.C.H. Kent 1908 ANDREWMUDD D. Longley, 'The Archaeological Implications of Gravel Extraction in NW Surrey', Surrey Arch. Soc. Research, iii (1976), 1-36. N. Macpherson-Grant, 'A Re-appraisal of Prehistoric Pottery from Canterbury', Canterbury's Archaeology, (1991), 38-47. N. Macpherson-Grant, 'A Review of Late Bronze Age Pottery from East Kent', Canterbury's Archaeology, (1992), 55-63. B. Philp, Excavations in West Kent 1960-70, (1973), 30-51. M.W. Pitts, 'On the Shape of Waste Flakes as an Index of Technological Change in Lithic Industries', J. Archaeol. Sci., vi (1978), 17-37. F. Pryor, Excavations at Fengate, Peterborough, England: The Third Report, Northamptonshire Archaeol. Soc. Mono. 1/ Royal Ontario Museum Archaeol. Mono. 6, (1980). M.R. Roberts, 'Excavations at former Jewsons Yard, Harefield Road, Uxbridge', (Oxford Archaeological Unit). A. Saville, Grimes Graves, Nor folk, Excavations 1971-2: Volume II the Flint Assemblage, London, Department of the Environment Archaeol. Rep. xi (1981). W. Shepherd, Flint, its Origin, Properties and Uses, (1972), London. G.H. Smith, 'A Beaker(?) Burial Monument and a Late Bronze Age Assemblage from East Northdown, Margate', Arch. Cant., civ( 1988), 237-89. F.C.J. Spurrell, 'Deneholes and Artificial Caves with Vertical Entrances', Arch. Journ., xxxviii (1881), 391; xxxix (1882), 1. Victoria County History of Kent, i, Appendix 1, (1908), 446-55. 410