rrftaolffjjiH fljantiana — + — THE EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB AT ADDINGTON, KENT By JOHN ALEXANDER, M.A., PH.D. A GROUP of large sarsen boulders near the long barrow at Addington has been known since the eighteenth century. It has always been included in the Medway Group of megalithic tombs although quite shapeless1 and there were no records, as at Little Kits Coty a few miles away, of it having been disturbed. I t stands in the greensand belt 100 feet above sea level, on the southwest slope of a hill crowned by a wood known as " The Chestnuts." In recent years this name has been applied to the stones but its earlier name • was Stony or Long Warren.2 Excavations were carried out for five weeks in August and September, 1957, through the initiative and financial support of the owner, Richard Boyle, Esq., with the encouragement and help of the Inspectorate of Ancient Monuments, the Ministry of Works.3 The majority of the work was carried out by volunteers4 with local help.5 It is the intention of the owner to repair and preserve the monument and the finds have been placed in Maidstone Museum. Before excavation in 1957, the stones looked as they did in the eighteenth century.6 The field round them had also been long undisturbed7 but there was no record or trace of the barrow which the old name of Long Warren suggested. In 1953, the field had been prepared for horticulture and although the stones were left undisturbed on an island of turf which preserved the slope of the field (Plan I), the rest 1 O.S. 65. 3592. 2. It had been scheduled as an Ancient Monument and discussed by Daniel, Megalithic Tombs of England and Wales, 81, and Piggott, Neolithic Communities of the British Isles, 112, and Evans, Arch. Cant., LXITI, 1950, 63. 2 Tithe Map, Addington Church. 3 Especial thanks are due for the plan and elevation of the stones prepared before excavation and now deposited in Maidstone Museum. 4 They were supervised with skill by P. Ozanne and L. Barfleld, without whom the excavation could not have been carried out so successfully. 6 Especially Mr. Healy. 0 Engraving in the Gentleman's Magazine, Aug. 1824, p. 111. ' It was a paddock in the late nineteenth century and had remained under grass, except in wartime, until 1953. 1 £* facade Robber Pbuflh Soil Mixed. SandR At humus w is 'V A gfcl^T -=AS* 5*S C0a**5 B nun i w n w -+ SURVIVING PAVEMENT TRENCHES XlltXIB Wall [IF MaiwKwCs .Bumm ft^»4 ^V &*+ s*i A to —r^Bwjo!..c K'4 (?.%' TOliOs S a n d r-** MotltaJ 'Whits Jsei-itf ^xm*.•T !T«T;7l rrnr Mesolitkk & Neolithic Levels M M ' — ^ —'—r1 I 0 I J 3 FEET MAIN ROBBER PIT TRENCHES XWU .XXItX Stone MflnV H \ Z I Mottled s o n i _ ® % * tlUtlULJ A Brown &. WS! A T 7 A i*t Bwwn.** Wiifs MmW^us A — •— !* \>—i • C 3 Stones pulled vertical. buC ocherwlse unmoved. / [ T ] fesifcioa uncertnia. ---—-"/ 'CO Rpbber ats. ;': Neolithic sherds. ^EiE R}Sthole5. O StDnekoles. •%• Undisturbed small stones. PLAN II. 7 OD EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB stone. The wallstones (F, H, V and S on the plan) were still in position although inclined at various angles and were not moved, beyond pulling them back into a vertical position, during the excavation. They had been chosen, apparently, for their flat bases for they had been set up without pits on the hard level sand of the bedrock reinforced by greensand blocks,1 which were still in position under them. None of the stones looked dressed, although they were of even thickness and had probably been selected for their shape.2 The largest stones were used here, the longest, " S ", being 12£ feet long and the heaviest, " B ", weighing over 10 tons. The chamber was trapezoid, about 12 feet long and 1\ feet wide. The east trilithon was the less damaged ; both the wallstones (S and H) were virtually intact and the capstone was " B ".3 Stone " H " must have taken most of the weight, for " S " has a narrow top, but it is possible that " G ", the entrance blocking stone, and even " V " shared it. The west trilithon was very damaged, and both the wallstones (V and F) had broken into several pieces. Stones " L " (certainly) and " A " (probably) were parts of " V ", and " D ", " Q " and " K " parts of " F "; the cap-stone was " C ". A large stone, " G ", lay diagonally across the east end of the chamber where it had fallen when the chamber collapsed ; when standing, it must have almost completely blocked the entrance. There was no large sarsen stone at the west end but instead there was a wall of greensand blocks4 (Plan IV and Fig. 1). The blocks had been piled up and, although disturbed by the collapse of the wallstones, some were still in position. At least four piles, in one case six stones high, were found and are the remains of a drystone wall which would have blocked the whole width of the chamber (see Fig. 9). In the centre of the chamber, buried by the mediaeval robbers' spoil-dump, the smallest sarsen, " E ", was found (Fig. 9). It was unbroken5 although only four feet long, and can never have formed part of the trilithons. It is unlikely to have been dragged in by the robbers and is probably a medial stone dividing the chamber into two. Its precise position is uncertain, for it lay in the robber's spoil, but it probably stood on its broader end. Such a small stone might well have been pushed over to make more room and the spoil-dump continued on top of it. 1 When pulled vertical in the excavation they needed no further support. The pointed base of " S " probably needed a pit but robbing had destroyed the evidence. a Some of the boulders scattered on Bluebell Hill a few miles away were of similar shape. 3 See Appendix I. 4 Rock of this sort does not lie near the surface here and it must have been brought in. 6 This is beyond doubt, see Appendix I. 8 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The inside of the chamber had been thoroughly turned over by the robbers but the stratigraphy of the spoil suggests its arrangement (Fig. 9), and this is confirmed by the undisturbed area at the entrance. Here (Plate II) there was a pavement of greensand slabs partly set in clean yellow sand above white sand. Neither the clean yellow sand nor the greensand slabs occur naturally at the site and must have been brought in. Both were found mixed as a stratum in the spoildump in the western half of the chamber, and from it came the bulk of the pottery and cremated bone. It is likely, therefore, that the whole of the chamber had a pavement set in yellow sand and that on it the burials were placed. The pavement at the entrance was some six inches higher than the old ground surface in the forecourt. The height of the chamber is Likely to have been about 10 feet. On the pavement a number of bodies were placed, for among the paving slabs and in the robber pits were more than 3,500 scraps of human bone (see page 52) ; all except two teeth were burnt and owe to it their survival in the acid soil. A minimum of 10 individuals, including at least one child, have been identified. From their position in the spoil dumps, the bulk of the bodies are likely to have been in the eastern half of the chamber. Also from the spoil in and around the chamber1 came objects likely to have been placed with the dead. The 34 sherds found belong to four, possibly six, pots, two of which had fingernail ornament and one a flat base. There were also two very fine barbed-and-tanged flint arrow heads, a petit-tranchet derivative arrowhead and a clay pendant (see pages 49-51). A lamb's tooth and a bovid's tooth and rib might belong here but are more likely to have come at the time of the robbery. The finds in the forecourt are also likely to have come from the chamber. Here on the old ground surface under the barrowlay 100 sherds of neolithic (Windmill Hill) ware (see page 38). They formed parts of at least eight bowls, had been broken elsewhere, and had with them a few scraps of cremated human bone and charcoal. They had not been disturbed since the final blocking of the tomb. It seems likely that they had been thrown out of the chamber to make room for other burials. A rather dubious leafshaped arrowhead, a denticulated flake and a thumbnail scraper (Page 51 and Fig. 12.6) from the disturbed levels higher may have been associated with them. The Fagade. and the Forecourt Four large sarsen boulders (T, J, M and R) framed the east end of the chamber, forming a curve with the blocking stone " G ". Only " J " had a flat base and the other three were all packed round with stones. The two northern ones (M and R) had holes dug for them, of 1 At the entrance, by stone " V " and between the wallstones. 9 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB which the lower part still survived (see Plan II). Stone " M " had a pointed base and had been eased into position against 2 inch thick stakes of which the holes of four survived. Deliberate symmetry seems to be given by the triangularity of the two outside stones (" T " and " R "). The eastern sides of all four stones were smooth and " M " may possibly have been dressed since it had a fluted surface to the east and was very rough to the west.1 Part of the entrance to the chamber was not disturbed by late digging ; it had been shielded by the fall of stone " G " and the mediaeval robbers had not previously been interested outside the facade. An area of about 70 square feet remained and was covered by a rough, but fairly level, pavement of greensand and two sarsen blocks (Plate II, Fig. 2 and Plan III). Seventy-four stones were still in place, although buckled and broken where " G " had fallen across them. On the west side they lay in yellow sand not found naturally on the site. They were 10 inches above the mesolithic ground surface and six inches above the neolithic surface of the forecourt so that the level of the chamber had probably been raised artificially. A small pit dug into this pavement before " G " fell might have been an original feature (or a robber testpit) for it contained a fragment of burnt bone and a flint flake. The pavement probably continued over the whole width of the entrance, for the robber pit north of " G " contained many more greensand blocks. In front of the facade the forecourt strata were intact under the plough soil. Sealed under the barrow (Fig. 1) was the old ground surface on which, 10 feet east of the pavement, was the concentration of neolithic sherds already described. Although there were some on top of others, all were horizontal and undisturbed ; they gave the impression of having been strewn there and were not at all weathered. With the other sherds scattered through the forecourt (see Plan II) they were probably throwouts from the chamber and so belong to the first burials. Near them, but three inches deeper, a group of 31 small greensand fragments lay horizontally ; a sherd of Windmill Hill fabric lay among them making it unlikely to belong to the mesolithic settlement already described. A similar group of stones was found outside the barrow to the north. The Barrow (Plan II) Nothing of this could be seen above ground although the old field name of Long Warren suggests that some memory of a mound survived into the eighteenth century. The north and east edges of the barrow were found buried beneath a brown sand formed by the erosion of the' 1 Five sarsen fragments stratified under the pavement at the entrance might indicate dressing. It is unlikely that there was time for the stones to weather before the pavement was laid. 10 PLATE IV The mediaeval robber pit under the south wall. PLATE V The robber pit under the Facade stones. EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB hillside, probably through mediaeval ploughing (Fig. 4 and 6). Levelling and deep ploughing had destroyed all evidence west and south of the barrow (Fig. 5), although extensive explorations were made (Plan I). Trenches were taken 12 feet beyond the barrow to the north and east without any sign of a ditch being found, there was also no trace of a peristalith at the edge of the barrow.1 The barrow was almost certainly D-shaped or trapezoid, for in Trench XXXI the north edge seemed to be making an angle with the east edge. At this point it was 30 feet from the axis of the chamber and suggests a total barrow width of about 60 feet ; at its widest, opposite the facade, it may have been about 64 feet. Its length could not be determined but probing and testpits found no trace of it 40 feet west of the chamber, even where the depth of hillwash had been expected to preserve it ; its overall length may well have been about 50 feet. It is certain that the barrow covered the forecourt area and its edge was traced 17 feet from the facade ; here it ran roughly parallel to the facade.and must have made a near right angle at the north corner. Nowhere was more than the bottom two feet of the barrow preserved. This was normally a mottled brown/white sand, much disturbed by roots and animals,2 but it is possible that separate tips are represented in Figs. 4 and 6. It contained the very numerous mesolithic industry already described and is likely to have been made from locally scraped up topsoil since the ground surface under the barrow contained many mesolithic flints while the surface outside it contained very few. Fragments of charcoal in it were of Quercus sp., Corylus sp., and Pyrus (possibly Malus)? Six to seven feet of the original slope of the barrow were preserved (Fig. 4) and show that the mound is unlikely to have been very high but must have been greater than the 10 feet suggested by the surviving slope. The Construction of the Chamber and Fagade In the absence of pits for most of the stones it is likely that the barrow was made first and used as a ramp up which the wall-stones and then the cap-stones were dragged into position. The bedding of stones on the hard sand would have been sufficient to hold stones chosen for their flat bases. The chamber would presumably be filled with sand for extra stability until the capstones were on, and then cleared.4 1 A broken and heavily weathered barbed and tanged flint arrowhead and group of 16 greensand fragments were found on the old ground-surface outside. They were the only signs of post barrow activity. 2 The Stones were once a famous local warren. 8 Identified by Miss C. Western. 4 This was the method suggested independently by army engineers who advised on the reconstruction. 11 £dae of Barrow SURVIVING BARROW (NORTH) TRENCHES VI&V1I >• I'UJ-> iJ> i TJ• l ' Chamber m*m?* Unexcavat2cL Soft; Browa'SartttT \i7hlte Hard Brown' Sand SURVIVING, BARROW (SOUTH) TRENCH I DisOarbgd^_bs -the-—fioanK Brown Hard. Brown Sand. Brown &foxon THE BARROW and the ROMANO-BRITISH HUT TRENCH XX Edoe of Barrow RouaK sol Ulwash mm 'White Sand Bedrocly Conventions on oil sections Edae of Barrow Barrow lost-Barrow FIGS. 4, 5, 6. Rprribil Oh Foundation, Trenck I 0 I 2 3 4 FEET EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The facade was probably made from the forecourt for the stone with the best preserved hole, " M ", had stakes on the west side. Both sarsen boulders and greenstone slabs still occur naturally within a few miles of the site. INTERPRETATION AND DISCUSSION The following sequence of events seems probable : First a D-shaped or trapezoid barrow 64 feet at its widest, and perhaps 50 feet long, was raised of local topsoil, a pit being left for the chamber. Into this were dragged wall and capstones and the chamber formed, floored with imported yellow sand and paved. More blocks were used to wall up the west end and a medial stone with a naturally polished surface was erected in the centre of the chamber. A facade of five stones was placed at the east end, the centre stone acting as a blocking stone for the chamber. In this tomb a number of bodies, of which 10 have been identified, were placed—some inhumed and some cremated. The earlier burials were accompanied by Windmill Hill pottery. These were later thrown into the forecourt when new burials, this time certainly cremations, accompanied by Late Neolithic/ Early Bronze age gravegoods were placed in the eastern compartment of the chamber. The barrow was then rebuilt in front of the entrance and the tomb left sealed for more than two millenia. The Chestnuts must now be set against what is known of the other Medway tombs. Although partial excavation has taken place at three of the five,1 it is only from Coldrum that finds and facts have been preserved and that detailed comparison is possible. Discussion of the others has gone as far as it can without more excavation, in the work of Daniel2 and, more recently, Evans.3 The latter's division of the tombs into groups is supported by the excavation of the Chestnuts, which falls into his Group B (Coldrum and Little Kits Coty) to which the destroyed tomb at Warren Farm may well have belonged as well.4 All the Medway tombs have in common terminal chambers (where the plans are known, simple rectangular ones) set in the east end of their barrows on similar alignments. Apart from this, the Chestnuts seems to have nothing in common with Evans' Group A ;5 in Group B the similarities between the Coldrum and the Chestnuts are very great. The chambers are built of a few large undressed but near rectangular stones arranged as trilithons on the same alignment. 1 For Kits Coty and Addington Long Barrow see Daniel, Chamber Tombs of England and Wales, 81, p. for the Coldrum, J.R.A.I., XLIII, 1913, 79. 2 Op. cit., 80, 161 and 233. 3 Arch. Cant., LXIII, 1950, 63 f. 4 Arch. Cant., LXT, 1948, 135. 6 Except the blocking stone at Kits Coty. Both the Chestnuts and the Coldrum may have had generally similar ones. If this parallel holds, then there was no false portal at Kits Coty. 13 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB They were erected, without pits being dug for the stones, at the east end of the barrows, had medial and entrance blocking stones and were paved. The barrows were of the same width and may well have been the same shape. There are two main differences ; the Coldrum had a peristalith while the Chestnuts had only an east facade, and the burial rite at the Chestnuts was cremation as well as inhumation while there were only two periods of inhumation at the Coldrum.1 The Chestnuts seems to differ as much from its near neighbour the Addington Long Barrow as Kits Coty differs from the neighbouring Little Kits Coty. With the Chestnuts firmly related to Group B of the Medway tombs, the wider implications of the new evidence from it must be considered. In the past Daniel2 and Piggott3 have looked across the North Sea for the origins of the group. The former's Scandinavian, and the latter's Dutch parallels were mainly morphological ones based on the Bennet reconstruction of the Coldrum (Daniel, Eig. 22) and the proportions of the Addington Long Barrow. Bennet presumed that the Coldrum was built on two levels but this is so unlikely that Evans' reconstruction,* with the chamber at the east end and the stones at the bottom of the slope as having fallen, must be accepted. This has been made more probable by the Chestnuts revealing a similar plan. Eor this plan North Eastern Europe provides no close parallels ;5 the proportions of the Coldrum and Addington barrows remain the only possible link with Northern Europe. This cannot be pressed, for cairns as narrow as the Addington Long Barrow and as square as the Coldrum are also found in the West6 and, what is more, have chambers at the east ends. This latter feature, held in common by all the Medway tombs, is so conspicuously absent in Scandinavia and the North Sea coasts7 that it would seem the strongest single argument against the derivation. The objects of Scandinavian type found in Kent8 are, with the exception of the Maidstone gouge, far from the tombs and cannot be accepted as evidence but the rusticated sherds from the Chestnuts and the local beaker evidence suggest contact with the mouth of the Rhine.9 Although suggestions for a north European origin were reasonable from surface indications and distributions, they needed the support of 1 The sherd from the Coldrum cannot be closely matched at the Chestnuts although some of the stray sherds found were rather similar. The denticulated flakes were also found at both sites. 2 Op. cit., 161. 3 Op. cit., 269. * Arch. Cant., LXIII, 1950, 73. 6 Noted by Evans, Arch. Cant., LXTTI, 1950, 76 f. 6 L'Anthropologic, XLIII, 1935, p. 251. Giot. Britanny, p. 99. 7 In the Dutch tombs the characteristic features are entrances in the long side of the barrows and chambers partly sunk in the ground. 8 P.P.S., 1938, 101 and Arch. Cant., 1937, 284. 9 De Laet, The Low Countries, p. 105. 14 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB excavation and this has not been forthcoming. Instead, the Chestnuts in tomb plan, burial rites (cremation) and gravegoods (pottery, pendant and arrowheads) has produced an assemblage unlike anything in North- Eastern Europe but with good parallels further west. If the suggested late date for the construction of the Chestnuts is correct, then this too is an argument against a Scandinavian origin. A western origin has already been supported by Crawford, and Evans,1 both of whom looked to the Cotswolds for analogies. This is the nearest English group and there is a general similarity in the terminal chambers and also in the barrow plans of Daniel's Tinkinswood-Manton Down group. Evans showed that the proportions of his group A (the long barrows) can be matched there and several atypical barrows, notably Manton Down itself, approach the proportions of Group B. Other similarities include straight facades of stones at the east ends,2 near rectangularity of peristaliths,3 entrance blocking over pottery and sometimes bones4 and perhaps siting the tombs in pairs.5 It may also be significant that the North Wessex longbarrows near the line of the Ridgeway as it approaches the Thames Valley are mostly chambered.6 The closest single parallel in Wessex is with the West Kennet Longbarrow7 where aspects of the tomb morphology, the inhumations and cremation rites and the pottery are similar. With this one exception the grave goods and the cremation rite at the Chestnuts are without good parallels in Wessex. No complete, as opposed to partial, cremation is certain,8 Secondary Neolithic pottery is rare9 and the Windmill Hill ware provides no parallels for the boss ornament.10 No barbed and tanged or petit tranchet derivative arrowheads have been found, although denticulated flakes11 and pendants, of different shape and materials occur.12 Daniel, opposing this western connection, argued from the barrow plans, the absence of sites between Berkshire and the Medway, and the Medway tomb riverine distribution.13 On present evidence the case for an origin in North Wessex is certainly not strong and is not much 1 Arch. Cant., LXIII, 1950, 81 and quoted by Daniel, op. cit., p. 160. 2 e.g. Waylands Smithy and the final facade of West Kennet. 3 Daniel, op. cit., p. 69; Piggott, op. cit., p. 132. 4 e.g. Notgrove, Piggott, op cit., p. 137. 6 Piggott, op. cit., p. 130. 0 Grinsell, Wessex, p. 36 and map I. 7 Antiquity, 1958, p. 235. 8 Daniel (op. cit., p. 99) reports it at three sites, but Piggott (op. cit., p. 139) only at one. None are well attested. 8 Three sites (Daniel, p. 166). The best attested is West Kennet where both fingernail and fingertip ornament was used (Piggott, op. cit., p. 144). 10 Piggott, op. cit., p. 142. 11 Piggott, op. cit., p. 145. 12 Piggott, op. cit., p. 86. 13 Daniel, op. cit., 160. 15 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB strengthened by this excavation. A distant inland parallel for the burial rites comes from Derbyshire for from Five Wells, cremated bone, Windmill Hill and Peterborough, sherds and a barbed and tanged arrowhead came from the chamber.1 Daniel's suggestion of a waterborne arrival is a good one but it would allow coastal movement round the headlands into the Channel as well as across to Holland and Scandinavia. This Channel route needs to be considered, for megalithic tombs on both sides of the Channel in Dorset and Normandy are little further from the Medway than the North Wessex ones and also have parallels. In Daniel's Penwith Group a number of tombs have terminal rectangular chambers in barrows, entrance blocking stones, and near straight facades.2 Although called " long barrows " one certainly (The Grey Mare and her Colts) and one probably (The Hellstone) are D-shaped, while at Zennor there were successive pavings of the floor as at the Coldrum. Also from Zennor and West Lanyon Quoit3 came cremations and Piggott suggests that these tombs were still being built in the Bronze Age.4 This agrees much better in time with the Chestnuts than does the North Wessex group which he thought to be out of use when beakers arrived locally.5 The bank barrows of Dorset, although otherwise different, do have a certain resemblance in proportion to the Addington Long Barrow. Since Daniel and Piggott6 both derive some of th Channel tombs from Ireland and East Scotland respectively, it ma well be relevant that the best parallels to the Chestnuts actually com' from Northern Ireland. In Londonderry and Tyrone a series oi chambers similar in size and shape and construction are set flush in straight facades and have D-shaped cairns with peristaliths ; multiple cremations come from them accompanied by barbed and tanged flint arrowheads and flat-bottomed pots.7 Other parallels are the medial stone from Labacalle, the numerous pendants of the Boyne Group and the Petit Tranchet derivative arrowheads from the Scottish tombs.8 The tombs along the French coast of the Channel may also be related to the Chestnuts. The most easterly groups in the Paris Basin, Lower Normandy and Maine offer few parallels,9 but better ones come from the Channel Islands. There, cremations come from three tombs 1 Reliquary and Illustrated Antiquary, 1901, 229. 2 Daniel, op. cit., 93 and Fig. 30. 3 Hencken, Archeology of Cornwall, 42-3. The Irish affinities of the Zennor pottery are worth noting here. * Op. cit., 266. They were certainly Bronze Age in date on the Scilly Isles where cremation was also used. 6 Op. cit., 148. 0 Op. cit., 154 and 179 respectively. 7 Piggott, op. cit., 191. 8 Piggott, op. cit., 251. 9 Arch. Jnl., 1950, 1 ff. In Lower Normandy there are rectangular chambers set in rectangular peristyles but the rites and gravegoods are unknown. 16 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB THE CHAMBER (EASTV TONCH XVII A UVT rrrrrn wkite •<•> CJTW m Sand o Sand -Jjrc'wrL :fi 111.1 i i i i_i Concentration of Bone THE CHAMBER (CENTRE) TKENCH XVII THE CHAMBER and BARROW (EAST) TRENCHE5 II * XVII V (Rwcttm) !r./'/ / 2z z z nt TTrrrrri< | M' I' I " f rm '. fKunui ffectht Brown Robber Sana Yel Re nd ™«3*S : \ So fl x / filcdfrq Stone Cremated Boat F I G S . 7, 8 9. and in one (La Varde) were associated with pots ornamented with bosses.1 Barbed and tanged flint arrowheads also came from five tombs2 while one (Le Couperon) had a rectangular peristalith.3 In Brittany, although further away, there are many parallels.4 A number of tombs have small rectangular chambers set in rectangular peristaliths5 and from them as well as from passage graves have come 1 Kendrick, Arch, of Guernsey, 116. 2 Kendrick, op. cit., 110. 3 Hawkes, Archaeology of Jersey, 255. 4 Giot, op. cit., 99 ff. 6 Arch. Jnl., 1950, 10. LXsLw. 1 0 2 3 4 5 FEET 17 7 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB cremations with barbed and tanged arrowheads and boss ornamented pottery.1 The arrowheads, in particular, are closely paralleled at the Chestnuts. Some of the Chestnuts' gravegoods have good parallels in the late Neolithic of S.E. Britain. At Walton, on the Essex Coast of the Thames Estuary, parallels for much of the pottery, the arrowheads and even the mesolithic industry have been found2 and they also exist at Woodhenge. This evidence suggests that the Medway tomb group is to be seen as an outlier of the Atlantic Coast complex, although possibly influenced from Holland. Isolated from the nearest tombs in the Upper Thames Valley and the channel, it is likely to be due to trade seaborne along the Channel or riverborne along the Thames and is probably later in time than the North Wessex group with the exception of the West Kennet Long Barrow. The original inspiration for it might lie as well in Ireland as Brittany, with Dorset and the Channel Islands as intermediate stations. Piggott has already claimed one Dorset tomb (the Grey Mare and her Colts) as an outlier of the Clyde-Carlingford Group, and it would be interesting if the Medway Valley had another, an historical context in which connections of this type were possible certainly existed in the Early Bronze Age. Since no general consideration of the Kentish Late Neolithic/Early Bronze Age has been made for 30 years,3 an attempt will be made here to set the megalithic tombs against their local archseological background. Disappointingly little is known of the neolithic period in Kent and what is known does not suggest that it was an area of early settlement. This is surprising, for Kent must always have been one of the first landfalls from the continent, but may be due to most of the known sites being on the north coast. This northern distribution is similar to that of the Early Bronze Age sites and it is likely that the two groups of sites are closely linked, should be considered together, and represent a limited period of time in the first half of the second millenium B.O. This is also suggested by the similarity of their overseas connections. Windmill Hill pottery from six sites seems to belong to the later phases of the culture. Dr. I. Smith4 considers both the Grovehurst and Wingham pottery to be typologioally late and the sickle and igneous rock axe from Grovehurst, if they really were associated with the pottery, confirm this.5 The forecourt sherds, the possible leaf- 1 VAnthropologic, 1934, 491, Figs. 8, 9 and 12. 2 P.P.S., 1936, 183 f. 3 The last was Jessup, Arch, of Kent, 36 f. 4 I am indebted to her for advice and comment. 6 The sickle in particular, for they have been found with a bronze axe in Holland. Le Laet, The Low Countries, 114. 18 EXCAVATION OP THE CHESTNUTS MEGALITHIC TOMB shaped arrowhead1 and the boss ornamented sherd from the Chestnuts, and the Danish thin-butted axe from the Chilham Long Barrow (an early middle neolithic type) also suggest a late phase of the culture, as do the clay spoons from Ightham and, on general grounds, the rest of the megalithic tombs although the few finds are inconclusive.2 The little known sites of Drowhill (Chatham), with its leaf arrowheads and squared ground flint axes, and Keston and Pembery might well be of the same period.3 Most of the secondary neolithic sites in Kent are near the north coast or in the Medway basin. At Ebbsfleet and Tankerton the North European influence discussed below is already apparent, while from East Mailing and Holborough4 come sherds of Rinyo-Clacton and rusticated wares respectively. The Chamber pottery from the Chestnuts, the bossed sherd and the petit-tranchet derivative arrowhead might well be Secondary Neolithic types as may the bowl from Maidstone. The Chestnuts megalithic tomb, with the Coombe Hill causewayed camp and the Badshot long barrow show that such constructions need not, in South-east England, be associated with the Windmill Hill cultures. The majority of the neolithic sites in the Medway basin are in the Aylesford region. This suggests that the combination of chalk down and greensand ridge, river-ford and access to the sea were already making this a focal point of human settlement which during the Early Bronze Age continued and the Chestnuts hints that there was no cultural break. The Pinal Neolithic/Early Bronze Age in North Kent is well represented, 12 sites having beakers.5 Nine of these sites are in the Medway Basin ; four beakers from Maidstone, Ightham, Allingham and Tovil, four inhumations from Aylesford6 one with bronzes, and a number of gold objects and a bronze axe from the river there, the Chestnuts cremations with barbed and tanged arrowheads, the West Tumulus, Free Down, Ringwold with its incense cup and cremations7 and the plano-convex knife from East Mailing.8 1 The scatter of this kind of arrowhead through the county should also belong here. One was found near Kits Coty (Evans, Ancient Stone Implements, 378). More important, are those from the " south surface " at Lower Halstow (P.P.S.E.A., 1925-7, p. 291) and Linton (Maidstone), (V.C.H.I., 317). 2 The sherd from the Coldrum is generally held to be of neolithic date. 3 P.P.S.E.A., 1927, 214 f. and Arch. Cant., 1951, 57 f. 4 Arch. Cant., LXX, 1956, 90, fig. 5. I am indebted to Mr. P. Ashbee for this reference. 6 Arch. Cant., XLV, 1933, 174, supplementing Jessup, Archaeology of Kent, 90 f. 0 Jessup, op. cit., 116, supplemented by the Annual London Institute of Archaeology XI 37 (hereafter L.I.A.A.). The three crouched burials in cists may be contemporary. 7 Jessup, op. cit., 121. The incense cup is paralleled further West at Normanton barrow IX. 8 Arch. Cant., LXIV., 1951, 161 and Fig. 4.1. 19 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB At both the Neolithic and Bronze Age sites there is evidence of widespread trade connections. These reached as far as- Scandinavia and the Low Countries on one hand and Ireland and Britanny on the other. The longest time span is suggested by the Scandinavian objects ;x these include the Northern Early Middle Neolithic axe from Chilham Long Barrow and the ring pendant paralleled in the Swedish Stone Cists from the roundbarrow at Sittingbourne.2 The objects are concentrated in the Thames estuary and although three come from the Medway basin they are not associated with the megalithic tombs and the find of Scandinavian megalithic pottery is from far away Orpington.3 There is, at present, no reason to see the Upper Medway as a centre of Scandinavian influence. The imports from North Germany and the Low Countries are more concentrated in time and consist mainly of beakers and their associated objects. Like the Scandinavian imports, they have a coastal and riverine distribution. Of the three beakers in the Upper Medway one is of Bl and two of the type B2 for which an origin near the mouth of the Rhine is likely.4 The tanged Faversham dagger appears, on analysis,5 to be of Central German copper, and further afield the Wye Down chisel is probably related to the North European Axe series6 and must be the earliest recognized metal import in Kent. The rusticated ware and the Beaker type barbed and tanged arrowhead from the Chestnuts might also come from the Low Countries.7 The trade to the Thames estuary was only part of a wider trade of which the main lines are well known but the details have not yet been worked out. Scandinavian, German and Dutch objects are found scattered along the Channel coasts as far west as Brittany and must represent the same trading complex. Especially significant seem the thick butted axes,8 the collared flasks,9 the beakers10 and metal objects of the Wessex and Armorican Early Bronze Ages.11 There can be no doubt at all that the Channel was used as a trading route at this time by the same people who visited the Medway. 1 P.P.S., 1938, 101, supplemented by I. Butler, Trade between the British Isles and N. Europe in the Bronze Age (unpublished), 287. 2 L.I.A.A., XI, 37. It was associated with a tanged knife dagger with a single rivet. 3 Arch. Cant., XLIX, 1937, 284. 4 Most recently discussed in L.I.A.A., XI, 34 and 41. 6 Discussed by Butler, op. cit., 271. « Arch. Cant., LXV, 1952, 182. 7 There is a wide distribution of barbed and tanged flint arrowheads in North Kent. 8 Kendrick, op. cit., 46, and Butler, op. cit., 287. 0 Giot, op. cit., 72. 10 Curwen, op. cit., 147. Man, XIII, 12. 11 Giot, op. cit., 144. 20 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB A reverse trade from eastwards along the Channel is equally well attested, from the Atlantic coasts with an additional overland route to the Upper Thames which was then followed to its mouth. The earliest recognizable objects are the igneous rock axes of Irish,1 Cornish2 and Armorican3 origin found along both coasts of the Channel and along the North sea coasts from East Anglia to Holland. In Ireland and Brittany4 these were used by megalithic tomb builders and cross Channel Knits have been suggested for the Group VII British axes.5 The finding of a Tievebullaigh axe at Sittingbourne and jadeite axes, probably from Brittany, at Canterbury and Southend6 show that the reverse trade reached North Kent. The Channel route is also suggested by the distribution of bronze and gold objects from the far west. The decorated flat and flanged axes have been studied in most detail.7 Two from Aylesford and Ashford, and perhaps one from Maidstone too, show the western connections of the Medway very plainly. One, a decorated type I I I axe, has led Ashbee to distinguish a separate Aylesford type8 which he connects with Brittany. The other, a flat axe from the grave discussed below is an Irish type paralleled by Butler with one from Killaha East (Kildare).9 The flat axe was in an inhumation grave with two flat daggers,10 one of ApSimon's Bush Barrow type which has close parallels in Brittany. As he points out, the group has parallels among the Armorican burials. Another related dagger came from Cuxton, further down the Medway, and a halberd from Faversham and decorated axes from Swanscombe and Buckland are of Irish type. Much of the Chestnuts evidence, the tomb shape, the cremations, the pendant and the Armorican type barbed and tanged arrowhead suggests the far west. Another group of objects from Aylesford also link it with Ireland or Wales ; these are the 11 gold objects from the Medway.11 Although they have not yet been studied in detail, some are undoubtedly of Irish type and show that the connection lasted at least through the Middle Bronze Age. They are part of the surprisingly large number of gold objects (37) found in North Kent. 1 P.P.S., 1941, 67 f. Especially Group I I I and VII, and P.P.S., 1951, 109 and 122. 2 Information for the Isle of Wight kindly supplied by Mr. J. Jones, Curator of Carisbrooke Museum. 3 Giot, op. cit., 80-81 and P.P.S.E.A., 1932-4, 154. 4 Ulster Journal of Arcliaeology, 1938, 74 and Giot, op. cit., 78 respectively. 6 P.P.S., 1951, 122. 6 Copley, Arch. ofS.E. England, 56 and P.P.S., 1951, 125. ' P.P.S., 1938, 272. The Buckland hoard, with its parallels at Arreton Down (I.O.W.) and Plymouth (Devon), is particularly significant. 8 Arch. Cant., LXV, 1952, 181. 0 Op. cit., 77. 10 L.I.A.A., X, 41. The larger dagger undoubtedly had six rivets. 11 Arch. Cant., V, 1863, 41 f. and Jessup. op. cit., p. 112 ff. 21 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The wider aspects of this eastward trade to Northern Europe have been studied by Butler, who sees its first climax1 in the axes, halberds, gold2 and possibly tin objects which are found in Northern Europe in the second half of the Middle, and throughout the Late, Neolithic. This climax, he considers, was followed by the establishment of some kind of Irish metal-working colony in Northern Europe.3 Against this background of trade the association of megalithic tombs and bronzes in the Medway Valley, both derived from the Atlantic coasts and both in all probability contemporary, assumes a greatly increased importance. It is not unlikely that they form a link in the chain connecting Ireland, and perhaps Brittany, with North Eastern Europe and might even indicate an actual colony from further west. The Aylesford region would have been chosen for its accessibility to the Channel4 and the Thames route as well as for the good land route of the North Downs. A possible additional factor may have been the presence of easily accessible sarsen boulders. SUMMARY In the late Neolithic and Bronze Ages, in the neighbourhood of the ford at Aylesford, lived a community which buried its dead in megalithic tombs. During the whole of its existence it was in contact with communities as far afield as the coasts of the Atlantic and the Baltic although its original inspiration came from the West, probably via the Channel. The Chestnuts at least belongs to the Late Neolithic and Early Bronze Age. THE ROMANO BRITISH HUT The first to the fourth century A.D. are represented at the site by a scatter of sherds through the mediaeval and later robber trenches and by part of a settlement found undisturbed in the northern lea of the barrow. Although earlier sherds were found, only fourth-century forms are common and the hut found belonged to the fourth century. This might well mean that the area was under cultivation throughout the period but that for only a short time was a hut here. The settlement lay up against the barrow edge (Plan II), the hut sleeper-beam trench being some seven feet away. It was erected on a level ground surface (Fig. 3) on which there had been little accumulation since the Early Bronze Age, and was covered by nearly three feet of brown sand caused by later erosion of the hillside when under plough. Some 60 square feet of the settlement were uncovered (Plan II). It 1 Op. cit., p. 280. 2 Brittany was involved in this trade as well, De Laet, op. cit., 102. 3 Butler, op. cit., p. 293. 4 The absence of any harbours along the dangerous Kent and Sussex Channel coasts may well have made the Thames estuary a more suitable anchorage. 22 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB consisted of a four inch thick stratum of grey sand full of sherds, charcoal, iron nails, burnt clay and bone, and flint fragments. All that remained of the hut was a sleeper-beam trench six inches wide and four inches deep, of which a 12 foot length was exposed. The hut must have been more than 10 feet wide and inside it was a possible hearth. On both sides of the trench, but not in it, were found 42 iron nails. The best preserved had round heads of £ inch, and shanks of i inch, diameter of which only J inch survived j 1 many were rusted in pairs. They lay at all angles and presumably came from timbers above ground. A number of pieces of burnt clay and wood were found but were too small to show wattle marks or to be identified as beams. There were no bricks or tiles, although one roof tile came as a stray from one of the robber pits. Potsherds were small but numerous (750), and were little weathered. Most of them belonged to the fourth century and at least 15 pots were represented, the majority came from inside the hut. No stratification was observed within the four inches and sherds from all depths fitted together. The sherds are discussed in detail on page 42. In the charcoal, are represented Quercus sp. (four fragments), Acer sp. (four fragments), Buxus sp. (two fragments), Populus or Salix sp. (two fragments).2 One small fragment of burnt bone was found. The charcoal was concentrated in a circular area of discoloured sand inside the hut and may represent a hearth. Only nine fragments of flint came from the stratum and three of them were struck flakes. All were heavily patinated and contrasted strikingly with the great numbers and lack of patination of the flints found in and under the barrow. This seems to confirm the suggestion made earlier that the original topsoil had been scraped up to make the barrow. There were no sarsen fragments from the settlement, the lowest pieces found being two feet higher in the hillwash above it. This suggests that the stones of the chamber were still intact and protected in the fourth century A.D. I t is possible that the hut was not long occupied, for comparatively few pots are represented and the stratum is not thick. It might well have been allowed to collapse, for this would explain the scatter of nails, charcoal and burnt clay. It could have been some sort of field shelter or temporary shielding but was apparently used by men rather than as a byre or barn. A considerable settlement of this period is known at Snodland3 and this might be an outlying hut. 1 Presumably of Cleere's type " e ", Bull. Lond. Inst, of Arch., I, p. 58. 2 I am indebted to Miss C. Western for the identifications. She comments that this is the first time that Acer (probably A. campestre) has been recorded from prehistoric Kent. The Buxus is a little doubtful but Malus is common in some localities in the southern counties. 3 Arch. Cant., LVIII, 1945, 68 f. and LXVIII, 1954, 5 f. 23 EXCAVATION OP THE CHESTNUTS MEGALITHIC TOMB THE PAVEMENT I swmy smsr* m t - i \ $ * A C : - . . / • • S . t f J K V ' " ' ' . '' (jreensand CI> Sarceri' HH Disturbed lallm enttance BLocfena 5tbne PLAH III. THE MEDIAEVAL ROBBING AND DESTRUCTION The eleventh to thirteenth centuries are represented by some 200 sherds (see page 45), two hones and 17 fragments of daub scattered through the disturbed top soil and recent pits of the turf island. Three pits and the chamber itself produced no post -thirteenth-century sherds. The pottery includes rim fragments from at least 14 pots, most of them, including all the early pieces, being small abraded sherds. They are likely to have come to the field when it was cultivated as part of the 24 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB arable of the manor and this is supported by the three feet of hillwash1 which silted against the north flank of the barrow. A few pots which are almost half complete and a number of unabraded sherds came from the chamber and from under the fallen stones. They are a homogeneous group and belong to the second half of the thirteenth century. They are the latest sherds sealed by the collapse and fix the period at which the tomb was robbed and destroyed. The evidence is particularly good from the chamber for here, once the stones had fallen, it was impossible for later robbers to penetrate. Quite apart from the evidence of the pottery, a mediaeval date for the destruction would have been suggested. The barrow was intact in the fourth century A.D. when the hut was built against it, and it had been utterly forgotten by the eighteenth century when the stones were thought a natural outcrop by the villagers. This implies that the robbing took place some considerable time before the eighteenth century and the twelfth to fifteenth centuries would appear to be the most likely period. The later thirteenth century has the support of literary evidence and of other recent excavation. The Close Roll of 12372 ordered the opening of Isle of Wight barrows to search for treasure and the excavation of a barrow on Arreton Down on the island3 has confirmed that this was done. The same entry refers to similar action having been taken in the Duchy of Cornwall and it is possible that the destruction of the Chestnuts was a continuation of the treasure hunt in Kent. The study of the pottery showed that one of the robbers' jugs might have come from as far afield as Oxford or Southampton. Since the instructions of the Privy Council referred to were not addressed to a local magnate, it is possible that the Chestnuts may similarly have been opened by some sort of special commissioner. The expertness and thoroughness of the robbing imply considerable resources well used and make it unlikely that the villagers organized it on their own.4 To carry this speculation still further, it is possible, now that barrow robbing in several counties can be associated with Henry I l l ' s reign, that some kind of organized treasure hunt over much of southern England was undertaken by a king hard pressed for bullion. The state of the other Medway tombs suggest that they were pillaged in a similar way. The disappearance of most of the barrows from Kits Coty, the Coldrum and the Addington Long Barrow and the wrecking of the three chambers could be the work of the same commission. 1 Before levelling took place there must have been six to 12 inches more of this stratum north of the barrow. 2 Grinsell, Ancient Burial Mounds, 155. 3 P.P.S., 1960, 273. 4 The manor was, however, held at this time by the rich and powerful Huntingfield family who would themselves have had sufficient resources and were also in royal favour. 25 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The Robbing of the Tomb The robbing was obviously systematic. The barrow all round the chamber was dug away and an entrance forced into the chamber from the north-west end.1 The whole of the chamber was then systematically cleared down to the hard brown sand of the bedrock. The spoil, which contained most of our neolithic finds, was dumped behind the diggers or, as they approached the east end, flung out between the wallstones. Some of the bones and sherds were flung out between stones " S " and " M ", " F " and " H ", and " S " and " V ". During this trenching the medial stone was pushed on to the spoilheap and covered over. A pit was then dug in the centre of the chamber, perhaps where the medial stone had stood ; presumably it was a test-pit. The chamber was trenched right up to the foot of the entrance blocking stone " G ". Deep pits were dug agamst the walls from outside (Plan II). The irregular south pit was some 10 feet in diameter and at its deepest undermined the chamber wall to a depth of 5 | feet (Fig. 3 and Plate IV). From its deepest part came a few fragments of cremated human bone, a barbed and tanged flint arrowhead and a number of greensand blocks, all of which were probably flung out of the chamber between the wall stones. These things were found only at the bottom of the pit which must, therefore, have been dug and empty when the chamber was being cleared. Mediaeval sherds were found in its upper levels and it was filled before the final collapse of the chamber. Since the pit was sealed by the two fallen capstones on its west, and two fallen facade stones on its east, sides it was not disturbed by any later digging. The north pit seems to have resembled the south one but, since it had not been sealed by fallen stones, was badly disturbed by later digging. I t was about 12 feet wide and was deepest behind the facade stone " R ", where a fire2 had been lit at the bottom of it just before " R " fell. Nearer the chamber, debris had been thrown into it as into the south pit, so that it was probably dug at the same time. When these pits had been finished, several more were dug against the facade stones. The largest was against " T " and " J " (Plate V); this went no deeper than the hard brown sand, so that they had presumably realized by this time that it was bedrock, but went under the bases of the two stones which had been pushed over on to the spoil dumps behind them. Another pit was dug under the end of "T" and a small one in front of " G ". " M " was investigated from the west side and was pushed or leant eastward. " R " also fell or was pushed back on the spoil behind it. Thirteenth-century pottery came from all these pits, and except in the case of the disturbed north pit, there were no later types. 1 Possibly because it was easiest to break in through the top of the drystone wall. 2 Of oak boughs. I am grateful to Miss A. C. Western for this identification. 26 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The Destruction of the Tomb The removal of the barrow round the chamber must have been the main reason for the collapse, for the stones had no holes and no other support. When the robbing finished, however, it is likely that only the deliberately pushed over facade stones were out of place. The angles at which these leaned showed that nearly all the barrow behind them had gone. The whole chamber seems to have collapsed like the house of cards it so much resembled. The blocking stone " G " went first, falling outwards across the pavement from which most of the barrow had been removed and sliding a little into the pit dug up against it inside the chamber. This must have happened soon after the chamber was cleared, for the end of " G " was resting on the bare bedrock which, on this site, would soon have silted over. The west trilithon then fell. The two wallstones fell inwards against each other and both broke. The breakage was worsened by the capstone sliding off to the south, perhaps breaking the wallstones still more as it did so. The fragments of the wallstones and the capstone fell on the robbers spoildumps, sealing much of them from later disturbance. One fragment " K " fell on, and partly displaced, the drystone wall. At this time the eastern trilithon fell. Both the wallstones inclined to the north (Plate I) and the capstone slid down outside to the south breaking " F " still further and driving a part " D " into the ground in front of it. Most of the interior of the chamber could not now be reached without removing the stones and later disturbance was made difficult. The stones must have been left uncovered after they fell, for they are frost fractured and the top soil all round is full of sarsen chips. The Post Mediceval Disturbances In the last five centuries a considerable amout of desultory digging and perhaps at least one serious effort to open the tomb have been made. Since it was a famous rabbit warren, much of the digging may well have been unrelated to robbery. Some 50 sherds, clay pipes of the seventeenth to nineteenth centuries (Page 54), marbles of stone and clay, bottles (eighteenth to twentieth centuries), brick tile and stone fragments (including a fragment of imported French marble1 were found scattered through the top 18 inches of soil. The marbles, a child's mug, a champagne bottle and traces of many local fires confirm local memories of the stones as a favourite picnic place. During these centuries at least three pits were dug and at least one effort was made to get into the chamber. The latter is the most likely 1 I am grateful to Dr. I. Forbes for this identification. 27 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB explanation for the dragging five feet westwards on to the spoil of " L ", the western fragment of " V " (Plans I and II). This must have been done after the collapse of the chamber. " L " could not have been moved until the chamber had collapsed, and in its new position a glazed sherd and pipe fragment were found under it, and it partially sealed a post-mediseval pit. The view of further closely packed stones this revealed must have discouraged further removal but, at some time long after 1900, a pit that was almost a tunnel was made between the the fallen stones here. It reached a depth of about two feet and left a tin, champagne bottle and a piece of plate on top of the medial stone. Outside the chamber pits were dug at the entrance and behind the west and north walls. A large pit at the entrance tried to enter the chamber between " S ", " M " and " G " but was foiled by the length of " G ". It contained cremated bone, sherds and mesolithic flints and it may have been from here that Larkin made the finds already mentioned. The pits outside the walls were less precise but there was disturbance down to bedrock against " S " and " V " and a well defined pit just west of the chamber (Fig. 1). THE WE5T WALL r ( OCaCteredL st&rtfis » • \ A \ i i /~ < Pvtr R t e .pes of PLAN IV. 28 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The only large missing piece of sarsen is the top third of the facade stone " M ". There is nothing in all this robbery and destruction to suggest iconoclasm. It looks as though the chamber collapsed from weakness after it had been robbed, rather than that it was deliberately thrown down, and the pits dug under the stones strongly suggest treasure seekers. THE FINDS THE MESOLITHIC FLINT INDUSTRY1 Total Assemblage Waste 201 Cores (187 typed, rest broken) 26 Core rejuvenation flakes 1,734 Primary flakes 3 Axe sharpening flakes 7 Microburins 174 Mis-hits and unclassified 3 Hammer stones Artefacts 15 Microliths 6 Truncated flakes 3 Sidescrapers 11 Endscrapers 6 Small blunted flakes 11 Hollow scrapers 6 Awls 8 Saws 2 Points 4 Flake axes 8 Core axes 39 Core scrapers 27 Miscellaneous retouch 98 Utilized Excluded from this total are five arrowheads and a thumbnail scraper found near the chamber. Proportion of utilized to waste = 1 0 - 1 per cent. 1 I am indebted to Mr. E. Higgs of the Department of Archeology, Cambridge, for reading and discussing the text with me and for valuable suggestions. 2402 Mr. Boyle's Collection (Excluded from Total) 26 5 5 3 3 5 6 1 2 15 6 29 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB I Inch FIG. 10. The Mesolithic Industry. 30 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB Apart from the higher proportion of utilized pieces, the total assemblage is identical with that found undisturbed on the old ground surface under the barrow and is likely to form part of the same mesolithic industry. The flint was mostly grey although some honey-coloured fragments were found ; it was fresh, unpatinated and nodular. The Artefacts Where the forms of these are well known they have not been drawn ; all unusual or particularly important pieces are shown in Fig. 10. The nomenclature is that used for the Downton industry (P.P.S. 1959, 216 f.). Microliths A.l Blunted obliquely down part of one edge : Right hand side, Complete 1 ; Broken 1. Left hand side, „ 0 ; ,, 2. B. Blunted down whole of one edge, ,, 3 ; ,, 2. Blunted down whole of two edges, „ 1 ; (Fig. 10.1). D.2 Crescents, arc blunted, ,, 2 ; D.5 Sub-triangular ,, 2 ; „ 1. Nine of these show signs of use. From Mr. Boyle's surface collection (not included above) : A.A, 1 ; B.l. 2 (broken, 1) ; D.2, 1. (Fig. 10.2). Small Blunted Flakes Blunted along whole of left side (Fig. 10.18) ; Complete 1 ; Broken 1. Blunted along part of right side ; „ 2; ,, 1. Retouched along both sides from different faces ; „ 1. Four of these are under one inch long ; they have been separated from the microliths because the bulbs of percussion remain intact. Truncated Flakes These retain the bulb of percussion and are truncated obliquely to the long axis ; two are concave and one convex (Fig. 10.6). The concave ones are close in form and size to the D.5 microliths, the only difference being that they are made on the base of the blade ; an unusual one from Mr. Boyle's collection is shown in Fig. 10.10. One of the points (Fig. 10.5) is really a truncated flake with additional working. 31 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB Scrapers There were 25 scrapers on flakes as against 39 on cores. Most were cortex free. 1. Side Scrapers Three flakes are steeply retouched along one edge, two to a convex and one to a straight edge. One, the largest flake, was retouched, in part bifacially, round three-quarters of its circumference (Fig. 10.12). One had been retouched on two sides (Fig. 10.17). 2. End Scrapers Nine are normal end-of-flakes scrapers with steep retouch (Fig. 10.7). Three are more than half covered with cortex and one was made on a heavy core rejuvenation flake. One of the remaining three was worked on the butt end, and the others were broken ; one was worked bifacially, one was also used as a hollow scraper and one had an undulating working edge matched in Mr. Boyle's collection (Fig. 10.9). 3. Hollow Scrapers All 11 are flakes from 4 to £ inches in length. They vary from small notches to long convex edges. One has two, and the rest one, notch. Five are retouched along the edge with the notch, one having it on the reverse face (Fig. 10.4). 4. Core Scrapers Thirty-nine of these make it the most numerous class. Most utilized the edge of the platform from which flakes had been struck, but two had been retouched to make a new edge (Fig. 10.13). One from the old ground surface had been reworked on two sides. Awls Four of the six awls were on small flakes or blades (Fig. 10.3) ; one of the others was on a heavy squat flake and had two points, and the other was on a core rejuvenation flake (Fig. 10.8). The pieces described as points could have been used as awls. Core Axes and Sharpening Flakes There were parts of eight axes ; three butts, three working edges and two complete. The latter two had S-twist along the sides (Fig. 10.14 and 16). There were three sharpening flakes. Flake Axes All four of these were broken but none are likely to have been more than 6 inches long. One (Fig. 10.23) had been made through the 32 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB patina of a much older flake and two others (Fig. 10.22 and 24) had large areas of cortex. One had a transverse sharpening blow. Saws Six of the eight saws were on blades, one on a sideblow and one on an ordinary flake (Fig. 10.20). The blades had short lengths of serrations near the centre of one side and three had wear on the opposed sides. All were well-used. On the flakes the serrated areas were up to 1 | inches long and were near the ends (Fig. 10.21). Points Two flakes had been made into lopsided points by truncating the flake and then retouching the opposed side (Fig. 10.6 and 11). A related, but not truncated, flake came from Mr. Boyle's collection (Fig. 10.5). Miscellaneous Retouch Twenty-seven pieces cannot be assigned to any specific type, their shapes often being fortuitous. Nineteen have small areas of retouch on one side only, but two of the larger flakes have been retouched on both sides (Fig. 10.15). One flake has been retouched from two directions. An interesting knife edge (Fig. 10.19) has been worked against a natural back and one small flake was worked to a trapeze. Unclassified Utilized Pieces There were 98 of these, all on flakes. They included several very fine large flakes, one very worn by use. Waste and Knapping Technique Cores The following types were found : Mr. Boyle's collection (not included in total) No platform 7 1 A. One platform 74 8 B. Two platforms 104 16 49 opposite each other 55 at right angles C. Three platforms 19 7 D. Flaked to form a chopper edge 1 9 These were nodules of which the largest measurable was 8 inches long ; most were under 6 inches long. Those without platforms 33 8 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB had had up to six flakes removed. The great majority had one or two platforms prepared by one or two transverse blows. Of those with single platforms 14 had four, and 22 five, flakes removed. Of those with two platforms 54 had between eight and 13 flakes removed. Over half the two platform cores were blade cores. Microlithic flakes had been taken from the depression of the negative bulbar scar on the platform. Three hammer stones were found and the cores without platforms had been stone trimmed. Twenty-six core rejuvenation flakes were found : A. Struck from same plane as platform 6 B. Struck at right angles to platform 4 C. Struck obliquely to platform 7 D. Struck from base to remove apex 9 E. Ridge flake with battered keel 5 F. Ridge flake batter along part of keel 1 Three of type E came from Mr. Boyle's collection (not included in total). Two of type A, four of C and four of E had been utilized. Two had been retouched to make scrapers (Fig. 10.17) and one an awl (Fig. 10.8). Microburins The following were found : 1. Butt (a) Notch on left, butt being uppermost 2 (b) Notch on right, butt being uppermost 2 2. Tip (a) Notch on left 1 (b) Notch on right 1 Form worked but unseparated 1 Incomplete 3 Doubtful 3 Pmnary Flakes Of the 1,734 only 32 had 75 per cent, cortex. Although the length and breadth indices of all the flakes had been worked out, they have not been published for they merely demonstrate that flakes of J to 1 inch in breadth and § to If inches long were commonest and were preferred for artefacts. The detailed figures have been deposited in Maidstone museum. CONCLUSIONS This assemblage, although not a stratified one, forms a homogeneous niesolithie industry, the pieces from the barrow having been caught up from the surrounding topsoil. 34 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The main characteristics of the industry are : the small number of microliths and their lack of variety, the few microburins, the absence of burins, the large number of core scrapers, the presence of saws, flake axes and a considerable group with miscellaneous retouch. These indicate that it belongs to the South East English group with both macrolithic and microlithic elements. A study of these has recently been made by E. Higgs1 in which he noted variations within the group and it is with his Downton type that the Chestnuts industry agrees best. At Downton all the features mentioned above are also present and even their relative proportions are similar. Detailed parallels exist between the microlithic classes,2 the truncated flakes3, the scrapers, saws, core and flake axes,4 the core types and the methods of working them. The differences between them are the absence of fabricators and chiselheaded arrows from the Chestnuts and the nature of some of the awls, the small blunted blades and the points there. The parallels seem even closer with Peacehaven,5 which is also in the Downton Group, for there burins were absent, microliths were fewer and the awls similar. Very little at the Chestnuts suggests his other group which had Farnham as its type-site. The small awls from the Chestnuts" and perhaps the two lopsided points7 are better paralleled there. In Kent, with the exception of Lower Halstow, there is little with which to make detailed comparisons. The finds from the neighbouring East Mailing8 and Ightham9 can be paralleled at the Chestnuts but are too few for further comment. The Northern Floor at Lower Halstow has in common the core scrapers, the awls and tranchet axes.10 The microlithic point which Clarke regarded as typical of the few microliths11 was also found once at the Chestnuts. Roughly flaked discs and the burins are, however, missing from the Chestnuts and, apparently, flint axes and microburins from Lower Halstow. The S twist on the side of the Chestnuts core axes was noted on one from Tovil in Maidstone Museum. Typologioally, therefore, the Chestnuts industry is closer to Downton and Peacehaven than it is to Farnham and Lower Halstow. This seems to reinforce the evidence both from under the tomb and from the 1 P.P.S., 1959, 209 f. 1 Ibid., Fig. 4, 31, 40, 55, 56. 3 Ibid., Fig. 5, 69, 70, 77. 4 Ibid., Fig. 8, 115 and 116. 6 Sussex Arch. Collections, 1924, 224. 8 Research Paper II Surrey Arch. Society, 521. 7 Ibid., 45. 8 Clarke, Mesolithic Age in Britain, 70. » Ibid., 70. 10 P.PJS.E.A., 1925-7, 172, 217 and 296. 11 Clarke, op. cit., 63. 35 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB excavations nearby, that it is a late site. Downton, it was suggested, did not long precede local neolithic1 while Lower Halstow certainly did.2 This in turn leads to the interesting possibility that the Downton group of industries might be late over a wide area of Southern England. THE PREHISTORIC POTTERY Two hundred and seventy-one sherds,3 coming from at least 16 pots, are of Neolithic/Early Bronze Age date. All are likely to have come from the chamber and suggest a sequence of burials. They fall, both by fabric and distribution, into two group?, the earlier being of Windmill Hill and the latter of Rusticated ware. It appears that the Windmill Hill pots were thrown out to make room for the others. The first group was found on the old ground surface in the forecourt of the barrow and had not been disturbed since it had been scattered there before the final blocking of the tomb (see Plan II) ; it was exclusively of Windmill Hill sherds. The second group came from the spoilheaps and pits left by the robbers in and immediately around the chamber. Only one sherd from the chamber was of Windmill Hill type, the rest were of Secondary Neolithic Rusticated ware. It seems likely from the pottery that the tomb was in use over a considerable length of time in the late Neolithic and Early Bronze Age. The earlier users had connections with Western Britain, possibly via the Channel ; the later may be more local or may have had connections with the Low Countries. The Forecourt Group Here, 140 small sherds were found in an area of about 50 square feet with a scatter over the whole forecourt (Plan II). They were of very friable, poorly fired, flmt tempered grey ware but were unweathered and could never have been long exposed to the elements. Although they lay close together few joined, and parts of eight pots were present ;4 since these were mixed and incomplete the breaking took place elsewhere. All the sherds except three were of a thin (dia. 2 mm.) sandy flinttempered ware and could be further sub-divided. Three pots (Fig. 11.5, 7 and 9), represented by 123 sherds, had a large sparse temper (dia. 1 mm.). All except one of them were plain, the rims were simple and there were no carinations or lugs. The pots seemed wide mouthed bowls. The one decorated sherd had two small applied circular lumps. 1 P.P.S., 1959. 229. 1 Clarke, Mesolithic Age in Britain, 63. 5 250 flint-tempered and 21 sand-tempered. I am indebted to Mr. I. Longworth for discussing them with me. 1 There were four fabrics, one being without a flint-temper. 36 12 13 14 FIG. 11. The Prehistoric Pottery. Scale 1:2. EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB Two more pots (Fig. 11.2 and 10), represented by 17 sherds, had a much more numerous and much smaller temper (dia. under 1 mm.). These too were plain with simple rims and seem simple bowls like the previous ones. At least two more bowls in this ware (Fig. 11.8 and 11), represented by 14 sherds, were scattered over the forecourt away from the main concentration while five more sherds came from disturbed levels. Among the big concentration just mentioned were two sherds of sandy brown/grey ware without a flint temper. Three similar sherds from the disturbed levels round the chamber probably belong to the same pot. The fabric, flint temper, plainness and rims of this pottery suggest the earlier Windmill Hill wares of Southern Britain and are generally similar to those from the lower floor of site 109 at Clacton on the other side of the Thames Estuary •} only the thinness and the single sherd with applied ornament are exceptional. Both these characteristics are found in the pottery at Woodhenge but the parallels must not be pressed ;2 there the applied ornament is normally in cordons and the rare circular lumps are larger and have depressed centres. The rims are also similar but are too simple to be helpful.3 The only other pottery with this land of ornament comes from the megalithic tombs of the Channel Isles4 and Brittany. The distantly related vase from Maidstone5 has very large crude bosses and very different fabric. The stratified Chestnuts pottery cannot be linked with the other Kentish Windmill Hill ware. The Grovehurst, Coldrum and Chilham Long Barrow sherds are much thicker and coarser but resemble some of the stray sherds from round the chamber. The single sherd which might show Rinyo-Clacton6 or Channel Megalithic Tomb influence would have little importance if there were not other evidence of connection with the west. In the event the group of sherds might well be of later date than the plainness of the ware would otherwise suggest. The Chamber Group Thirty-four sherds were found in the chamber and the pits and spoilheaps round it ; they are likely to have come from the chamber at the time of the robbing and so to have been connected with the burials then in it. Four, possibly six, pots are represented, made in three different fabrics ; only one sherd is of the fabric found in the forecourt. Thirty of the sherds are thick (dia. 4 mm.) with a well-fired skin of light i L.I.A.A., X, 28 f. 2 Woodhenge, pi. 26, 1, 2 and 4. 3 Woodhenge, pi. 32, 43 and 38, 93. 4 Kendrick, Arch, of the Channel Islands, I, 88 and 90. 6 Jessup, Arch, of Kent, fig. 14. 6 Piggott (N.C.B.I., 328) derives the Rinyo-Clacton motives from the west. 38 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB ] I I I O i Inch FIG. 12. Objects from round the chamber. 39 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB brown clay over a soft core. There is a very large but sparse flint temper (dia. 2 mm.) and a brushed on self-slip covers many of the blemishes. Only one rim was found (Fig. 11.1) and this suggests a simple round bottomed bowl. Four of the body sherds were ornamented with fingernail, and two with fingertip, impressions (Fig. 11.12-13) and the difference in the ware would suggest different pots. Another pot is probably represented by the flat base with a groove at its edge (Fig. 11.14) and two body sherds. These sherds, so different from the previous group, are best paralleled among the rusticated wares found on a number of Late Neolithic/Early Bronze Age sites. The nearest good parallels are from the ploughed out round barrow at Holborough (Snodland) where there were rusticated sherds, one with Ebbsfleet and another with Holdenhurst traits,1 and Clacton on the other side of the Thames Estuary2 where there were similar Windmill Hill sherds, and arrowheads. Further afield its parallels at Arminghall3 and Brighton4 and again Woodhenge (loc. cit.) show it to belong to the late Neolithic/Early Bronze Age. The indiscriminately used fingernail ornament is also found on the rim of the small urn discovered with a cremation in the west tumulus at Ringwold ; segmented faience beads come from the same barrow.5 These sherds accord well, therefore, with the other finds from the tomb, the barbedand- tanged and petit-tranchet derivative arrowheads and the cremations. Outside Britain similar ware has been found in the Dutch megalithic tombs6 but not with the same form or associations. From the chamber came three sherds of an ungritted, rather corky, ware and 12 others came from the disturbed levels but there were no rims. These are likely to be of neolithic date. Other Possible Prehistoric Sherds Twenty-three sherds from the disturbed levels round the chamber might belong here. Six of them are flint tempered but are harder and less sandy than those found in the forecourt. Four of them have been burnished on the outside and all might date from the Early Iron Age. It is among these sherds that the single one from the Coldrum is best paralleled. The others have no flint temper but do not resemble the mediaeval or Roman wares. Detailed Description of the Figured Pottery 1. Coarse hard brown ware with rather sparse, but large, flint temper and a brushed on self-slip. The rim is simple with irregular indent- 1 Arch. Cant., LXX, 1956, 90. 2 P.P.S., 1936, 187, Fig. 2 and pi. 39. 1. 3 P.P.S., 1936, 18 and fig. 9. 4 A.J., 1934, 112, 119 and Fig. 54 and 7. 5 Jessup, op. cit., 121. 0 de Laet, The Low Countries, 105. 40 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB ing below it on the outside. Fifteen other sherds were of identical fabric ; all were plain and seemed part of one coil-built bowl. They were found in the chamber and the spoilheaps immediately south of it. Sherds from inside and outside the chamber joined. 2. Thin, rather sandy, grey ware with a plentiful but very fine flint temper. The thickened rim was made by folding over the edge and smoothing it on to the pot. From the forecourt. 3. Rather thicker grey ware with heavier flint temper ; self-slip on outside ; probably part of an everted-rim bowl. Sixty-two bodysherds are of the same ware, some showing the curve of the belly. From the forecourt. 4. Similar to 2 ; possibly part of same pot. Rim broken on outside but might have been thickened. From the forecourt. 5. Grey ware, sometimes red at core. Poorly fired and having large flint temper. Twenty-two body sherds were of this ware. From the forecourt. 6. Similar ware to 3, but with slightly everted rim. Thirty-seven plain body sherds, some with red at the core, seem to go with it. From the forecourt. 7. Similar ware to 3 ; could be part of same bowl but very thin rim. From the forecourt. 8. Ware similar to 2 but rim everted. From the forecourt. 9. Ware similar to 3 but red at core. Two small applied lumps just below the rim of the open bowl. Fingermarks where they were pressed on are visible on the inside. From the forecourt, found with 6 and 7. 10. Same ware as 2 and possibly part of the same bowl. From the forecourt. 11. Similar ware to 2 but thicker and less sandy. The best formed rim in this ware. Forty-seven body sherds are similar to it. From the forecourt. 12. Coarse brown with large sparse flint grit ; rather like 1. Brushed on self-slip inside. Outside ornamented by 10 shallow finger-nail impressions at all angles. There are 10 body sherds in similar fabric, four with similar impressions. They were found with 1 in the chamber. 13. Coarse, poorly fired, buff on outside, brown on inside with a tendency to split along the joint. Sparse but large flint temper. The outside has at least four very deep finger-tip impressions made regularly from one direction, quite unlike 12. Four sherds of the same ware were found with it in the pit at the entrance to the chamber. 41 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 14. A hard black-cored flint-tempered ware badly damaged on the outside. The base probably met the wall at an angle rather than in a curve. There appears to have been a groove round it. It came from the forecourt, but away from the main concentration. THE ROMANO-BRITISH POTTERY Some 830 sherds belong to this period ; they are the first to be noted from the parish. The great majority (750) came from the undisturbed hut floor sealed below nearly three feet of hillwash in the lea of the barrow. The remainder, often weathered, came from the disturbed levels and the later robbing pits. Those from the hut floor are much the most important for they show that it was inhabited in the fourth century. This need not have been a lengthy settlement but at least 12 pots were broken on the site. A few sherds suggest that there was some earlier local activity and this is supported by the evidence from the disturbed levels which spans the second to fourth centuries. This scatter of earlier sherds might mean no more than the regular cultivation of the area. * * ? V ^> 7 O Q ^ 7 7 1 7 _i> C.4 8 V FIG. 13. Bomano-British Pottery. Scale 1:4. 42 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB Pottery from the Hut Floor (750 sherds) These all came from a four-inch thick stratum. Fifteen pots were identified from their rims, and the majority of the body sherds were of the same wares. Six different fabrics and forms could be recognized ; the common fabrics and all the forms are drawn and described in Fig. 13 and below. The bulk of the sherds (680) were of a well-fired grey sandy ware and flanged basins, jars with everted rims and lids were made from it (Fig. 13.1, 2, 3 and 5). Two of the body sherds had vertical lines of burnish on them. Forms 1 and 2 can be securely dated to the fourth century, and the others, although they have a longer life, were still in use then. The commonest form, the everted rim jar, was also found in a soft red ware (Fig. 13.6). Three other fabrics were represented only by single body sherds ; a ' thin hard smooth grey ware, a soft red ware with a black slip of both faces and a hard grey ware with a sparse flint temper. The first two are common Romano-British fabrics, the first being the only piece of fine ware from the hut. The flint tempered sherd might be a stray from the barrow, for it is generally similar to one of the wares from the chamber but it can also be matched among late Roman wares. Standing apart from the others is a group of 30 thick corky textured, rather weathered, sherds (see Page 44 and Fig. 13.4). All seem fragments of one carinated bowl of first-, or early second- century type. Pottery from the Disturbed Levels These include three new fabrics and four new forms. The most important is the fine hard buff ware in which a mortarium, probably of second-century date, was made (Fig. 13.8) ; it should be earlier than ' the hut. Three other small sherds were of colour-coated ware and so should be later than the second century. Finally, there were sherds of grey cored redware in which a flanged bowl of fourth-century type had been made (Fig. 13.7) ; this would be contemporary with the hut. Description of the Figured Pottery (Fig. 13). From the Hut Floor : 1. Flanged Basin Fine, smooth, sandy ware fired grey over brown ; the core remained grey because of underfiring. The surface was finished in horizontal rilling. Wheel-made with rather a thick flange. This is a well known fourth-century type. In Kent it was dated to the middle and later fourth century at Richborough (Bushe-Fox, Report 1, PI. XXIX. 121 and page 104), and to the late fourth and fifth centuries at Canterbury (Williams and Frere, Butchery Lane, Fig. 9.9 and page 43 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 30). A similar date is given for it at Leicester, where it was found with No. 2 below (Kenyon, Jewry Wall, Fig. 55.5 and page 205, type C), and at Southampton (Cotton and Gathercole, Clausentium, Fig. 29.11 and page 120). 2. Jar with outcurved rolled Lip A rather thinner, fine smooth grey ware. Wheelmade, well fired and finished with rilling on shoulder and rim. This is also a late third- and fourth-century type. In Kent, at Richborough, it was given that date (op. cit. Report IV, PI. XCIII. 465 and page 268) and at Canterbury was found in a well with mainly fourthcentury pottery (Jenkins, Burgate, Fig. 15.98 and page 29). A late fourth-century date is given at Leicester (op. cit. Fig. 55.18), and Southampton (op. cit. Fig. 27.12 and page 117). 3. Bowl (?) Grey, sandy well-fired ware. This was too small a fragment for close identification. It is a simple and probably long-lived type, either as a bowl or lid. Possible parallels at Richborough (op. cit. Report 1, PI. XXVII.85) and Canterbury (Williams and Frere, Butchery Lane, Fig. 14.30 and page 31) belong to the late first and early second century, but at Southampton it belongs to the late fourth and early fifth century (op. cit. Fig. 29.9 and page 120). 4. Carinated Bowl A soft coarse ware with many impurities and a soapy texture. Wheel-made and underfired to a mottled pink/buff with a grey core. Weathered. This bowl differs from the others, both in ware and form, and might belong to an earlier period. It preserves pre-Roman traditions and does not seem to occur later than the early second century. In Kent it was found at Richborough in mid-first and early secondcenturv levels, (op. cit. Report 1, PI. XXVI.73 and Report III, PI. XXXIV224 and page 170). 5. Everted-Rim Jar Rather coarse grey sandy ware. Fired to a mottled brown/red surface. Wheel-made. This type of jar had a long life. The rather square rim is paralleled at Richborough in the late first and early second centuries (op. cit. Report III, PI. XXXVII.269), but it was also found at Leicester in the first half of the fourth century (op. cit. Fig. 52.39). 44 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 6. Everted-Rim Jar A soft red sandy ware with a mottled red/brown surface. Wheelmade. This is very similar in form to No. 5 and at Richborough (op. cit. Report III, PL XXXVII.267 and page 175) and Colchester (Hull, Roman Colchester, Fig. 119.268a and page 285) belonged to the late first and early second century respectively. At Southampton, however, a similar jar was dated to the fourth-fifth centuries (op. cit. Fig. 29.13). From Disturbed Levels : 7. Flanged Bowl Fine grey ware with thick red slip. Wheel-made. This imitates the Samian form 38 and is a well known middle to late fourth-century form. It was dated to this period at Richborough (op. cit. Report I, PI. XXVII.109), Southampton (op. cit. Fig. 29.38 and page 122) where it was found with No. 1, and Leicester (op. cit. page 80). 8. Mortarium Well-fired smooth buff ware with three quartz fragments set in the inner rim. This is too badly damaged to be certain of its form. The slight bevel on its inner rim is rare and is not figured from Richborough or Canterbury. The best parallel is from Southampton (op. cit. Fig. 24.5 and page 103) where it belongs to the late first, and most of the second, centuries. 9. Ledge-rim Lid or Bowl Rather coarse grey ware. The fragment is too small to say more. THE MEDLEVAL POTTERY Some 200 sherds found scattered through the topsoil and the robber pits are likely to belong here. The majority of them are small abraded body sherds, but the 20 rim fragments indicate at least 14 pots and possibly 90 of the body sherds belong to them. Thus, 114 sherds may be classified into four groups. Group 1 The 40 sherds of this type make it the largest group (Fig. 14.1^4, and described below). It is a coarse corky fabric made into cooking pots with ledge rims ornamented by finger printing, having shallow vertical ribs and sagging bases. At least six pots of this type are represented. The type is well-known at Maidstone1 but is best dated at Canterbury 1 Stray finds in Maidstone Museum. 45 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 1 \ FIG. 14. The Mediaeval Pottery. Scale 1:4. where it belongs to the thirteenth century.1 Here pit M 15 had the closest parallels and belongs to the second half of the thirteenth century ; elsewhere Frere2 suggests that the fully developed ledge and increasing angularity of the shoulder are indications of lateness and both are found at the Chestnuts. No examples with fingertipping on the rim were found at Canterbury and the typical pricking found there is missing from the Chestnuts. An earlier date for some of the less developed rims seems likely, for at Oxford3 the type with the same shallow vertical ribs i Arch. Cant., LXVIII, 1954, 133, Fig. 18, 29. 2 Roman Canterbury, No. 4, 43. 3 Well-filling at St. John's College. Oxoniensia, 1950, 54. 46 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB is as early as the late twelfth century, although the type there runs on through the next century.1 This group of sherds is the most significant for the dating of the destruction of the tomb. Specimens were found immediately under the fallen facade stones " J " and " T " and must have got there immediately before the collapse ; so many fragments of one pot (Fig. 14.1) were found, unweathered and freshly broken, that it must have been smashed not long before. Group 2 Seventeen sherds of this type belong to at least four pots (Figs. 14.5, 7 and description below). The ware is smooth grey/black and sandy, and all four have narrow necks and are probably jugs since one at least had a handle. They too had shallow vertical ribs and sagging bases. At Canterbury jugs of this type came from Butchery Lane2 and are dated to c. 1200 A.D. A closer parallel for Fig. 14.5 comes from Chalk near Gravesend where it is dated to the thirteenth century3 and a closer one for Fig. 8.7 from Windcliffe (Isle of Wight) in a thirteenth-century midden.4 This group was more widely scattered than group I and no sherds came from the chamber. Group 3 Eighteen sherds were of this type (Fig. 14.6 and 8 and description below). Probably two pots are represented. Both are jugs, one with a fine handle with rope and finger printed rim, of a fine grey cored orange ware. Handles of this type are common in the Oxford region on partially glazed tripod pitchers of the late twelfth and thirteenth centuries5 and were made by the same curious technique. They are also common at Southampton. This is a very rare type in Kent; a fragment came from Chalk in a general thirteenth-century context6 and another from Maidstone.7 It might possibly be an import from further up the Thames Valley but is very rare at London8. The handle was found, in three pieces, at the bottom of the chamber and had been freshly broken not long before the collapse. 1 Oxoniensia, 1943-44, 102 ff. 2 Roman Canterbury, No. 4, Fig. 17.11 and page 42. 3 Arch. Cant., LXVIII, 1954, Fig. 3.38. 4 Isle of Wight Nat. Hist. & Arch. Soc. I l l , Part. 2, 1939, Fig. 1.5 and 6. 6 Oxoniensia, 1950, 47. 0 Arch. Cant., LXVIII, 1954, Fig. 3.39. This particular piece was not dated by the excavators. 7 I am indebted to Mr. A. Grove, Curator of Maidstone Museum for this information and for much other help. 8 Information from Mr. N. Cook, Keeper of the Guildhall Museum. 47 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB Group 4 The 16 sherds of this group might be related to Group 2. They are of a sandy/grey brown ware and four have rouletting as well as shallow vertical ribs (Fig. 14.9). They come from under the fallen stones and the chamber. It can be seen that these sherds form a fairly homogeneous group belonging to the later twelfth and thirteenth centuries. The more numerous and freshest are those of the later thirteenth century and may give an approximate date for the destruction of the tomb ; they include at least one jug which probably came from a distance. Description of the Figured Sherds (Fig. 14) 1. Flat-rimmed Jar A grey, sand tempered, poorly cleaned ware with a corky texture. A selfslip has been brushed on, both inside and out, but has failed to hide the many blemishes. It is well fired. The pot is hand-made with a number of irregularities, especially on the belly. The sagging base has been knife trimmed. 2. Flat-rimmed Jar The ware is identical with that of No. 1. The form differs in having a narrower rim and in thumb impressions along its edge. Hand-made. 3. Flat-rimmed Jar The ware is similar to Nos. 1 and 2 but has a slightly more granular texture, presumably from a higher proportion of sand. Its form differs in the concave slope to the inner rim. Hand-made. 4. Flat-rimmed Jar The ware is similar to Nos. 1-3 but is blacker and with fewer blemishes ; this may be due to a thicker slip. Its form differs in having a much narrower rim. 5. Jug A pale grey, well-cleaned, clay with a granulated texture. Self-slip. The attachment scar of a handle could be seen. Hand-made. 6. Jug A grey cored, red surfaced, ware, with a granulated surface due to a sand temper. Not completely fired through. A badly weathered sherd. 7. Jug Coarse grey ware with soapy texture. Sparse chalk temper. Abraded sherd. 48 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 8. Handle A buff surface, grey cored, ware, probably from the same pot as No. 6. Sand tempered with a granulated texture. Not fired all through. The rim has been ornamented by fingertipping and the handle by excising a long strip and inserting two intertwined ropes of clay. 9. Body Sherd A well fired, sand tempered, brown/grey ware, hand-made and self-slipped. Poorly finished so that the joins between the coils show. I t is ornamented by horizontal rouletting near the neck and by thin vertically applied cordons pressed into ridges. THE NEOLITHIC AND EARLY BRONZE AGE OBJECTS OTHER THAN POTTERY Five arrowheads, a scraper, and a pendant are considered here. FLINT From the Excavation : Five arrowheads were found, three barbed and tanged, one petittranchet derivative and one unfinished or leaf-shaped. Barbed and Tanged : Straight sided (Fig. 12.1). Pressure flaked, unpatinated and in mint condition. The barbs and tang are in a straight line. It came from the bottom of the South robbing pit. Convex sided (Fig. 12.2). Pressure flaked, unpatinated and in mint condition. The barbs are square and are in line with the tang. It came from the robber pit at the entrance to the chamber. (Fig. 12.3). Pressure flaked but smaller and less regular than the others. Heavily patinated, one barb was broken off before patination took place. The barbs and tang were not in a straight line. It was found on the old ground surface outside the barrow to the north. Petit-tranchet derivative (Fig. 12.4) : Unpatinated but of more opaque flint. It is also thicker and is not pressure flaked. It is retouched bifacially and steeply on the base and the left-hand side and unifacially on the upper part of the right side. The tip is broken. It was worked obliquely across a large flake, the butt being removed. From the Surface of the Field : Unfinished (?) or Leaf-shaped (Fig. 12.5) : Rather crudely made from a large flat flake. The whole dorsal circumference is steeply worked but is cruder at the lopsided point. 49 9 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB The reverse is worked all down one side and half the other. There is a small tang but it can never have been of any use. This comes from Mr. Boyle's collection. The Thumbnail Scraper (Fig. 12.6) : This is unlike any of the mesolithic scrapers and might belong here. I t is of translucent grey flint and is made from the snapped end of a flake. The end and one side have been steeply retouched, leaving an area of cortex at the centre, and the whole edge shows signs of use. This also comes from Mr. Boyle's collection. The barbed-and-tanged arrowheads belong to the large class found in Beaker and Early Bronze Age graves in Britain. Their variations in shape have often been studied without their significance becoming clear. Smith1 noted that specimens with the barbs and tangs in a straight line are often found with beakers, but did not consider the straightness of the sides significant. This may indicate lateness for, when from closed groups, these belong to the later part of the Early Bronze Age.2 The most elaborate specimen from the Chestnuts (Fig. 12.1) is of this type. Grimes, although doubting the uses of detailed typology, suggests an origin in Brittany for this type3 and good parallels occur in Giot's First Series of Barrows.4 Although the concavity of the sides and the slightly reduced tang often found there are missing at the Chestnuts those features are present at the related site of Waltham across the Thames Estuary and might show some connection with the west.5 Some of the Armorican barrows contained megalithic chambers and the arrowheads were also found in other types of megalithic tomb.6 They were associated with boss ornamented pottery and flat riveted daggers.7 One from the chamber of Five Wells megalithic tomb had the small tang common in Brittany.8 Equally well made is another specimen (Fig. 12.2) which differs only in the slight convexity of its sides. It may well have been buried with (Fig. 12.1) for at Breach Farm9 the two types were found together and they both come from Waltham in Essex.10 They are more common than the straightsided ones, are found with beakers11 but also come 1 Archceologia, 1927, 81-106. * Ibid. Rudstone with a food vessel, p. 99; Alton Parva with a grapeeup and dagger, 100; Everley with a cremation, p. 102. » P.P.S., 1938, 116. * Giot, op. cit., 130. 6 P.P.S., 1936, 205, Fig. 12.18. « L'Anthropologic 1930, 491. Fig. 12.10 and 13, Fig. 8.8. ' Op. cit., 113, Fig. 6.9 and 12. 8 Reliquary and Illustrated Archaeologist, 1901, 229. » P.P.S., 1938, 113 f. 10 P.P.S., 1936, 205. » See, for example, S.A.C., 1936, 23. 50 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB from the megalithic tombs of the Channel Islands,1 Ireland2 and Scotland.3 On Jersey they are of Early Bronze Age date4 and they also come from the Bronze Age Gallery graves of Londonderry and Tyrone.5 The association of these arrowheads with megalithic tombs, bossed pottery, cremations and pendants in Kent and the West is unlikely to be fortuitous and they should be seen as part of a single tradition. The small broken patinated arrowhead (Fig. 12.3) is not certainly connected with the barrow since it was found outside it on the old ground surface. It is likely to have been dropped in the Early Bronze Age and suggests that the type was in everyday use. The Tranchet Derivative arrowhead is, in its way, as fine a piece as the others. It is one of Clark's rare type " H ", which he lists only from Woodhenge and Hungry Bently.6 Near at hand, related forms come from Farnham7 and Waltham (Essex)8 and, further afield, from megalithic and sub-megalithic tombs in Derbyshire and Scotland.9 On the Woodhenge and Hungry Bently evidence this would be a late Neolithic/Early Bronze Age type and might have been connected with the rusticated pottery from the chamber which also has parallels at Woodhenge and Waltham. The two other pieces—the unfinished arrowhead and the thumbnail scraper—cannot be associated directly with the tomb but are of the same general period. The arrowhead (Fig. 12.5) can be matched at Drowhill (near Chatham) where it was found with ground stone axes of square cross section.10 The thumbnail scraper is a common Neolithic/Early Bronze Age type, found, for example, at Waltham.11 The Clay Pendant In front of the entrance, in the barrow material at a depth of one foot three inches, was found a pendant. It is made of slackly baked buff clay with a large sparse flint temper and is similar in texture to sherd No. 13. It is oval with a long axis of 1 | inches and is slightly constricted on either side of the eccentric hourglass perforation. One 1 Kendrick, Arch, of Guernsey, 53 and Hawkes, Arch, of Jersey, 60. 2 Piggott, N.B.C.I., 208, 178 and 191. 3 Ibid, p. 230 and 251. 4 Hawkes, loc. cit. 6 Piggott, loc. cit. 0 Arch. Jnl., 1935, 52. At this latter site it was found with a cremation and a jet ornament. ' P.P.S., 1939, Fig. 24.6. Further west along the greensand. 8 P.P.S., 1936, 110. 205, Fig. 12.14. Across the Thames Estuary. 9 Piggott, op. cit., 180 and 251 and P.P.S., 1938, 317. 10 P.P.S.E.A., 1926-7, 214, Plate A3. " P.P.S., 1936, 206, Fig. 13.4. 51 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB face, presumably the front, is much flatter and smoother than the other. The perforation was probably made before firing and must have been bored from both sides. Pendants are common both in megalithic tombs throughout Western Europe and also in the Neolithic and Early Bronze Age of Southern Britain, but none have been noted that were made of pottery ; stone or bone are the common materials and the hourglass perforation suggests that the Chestnuts specimen was a copji- of something harder. Its constricted shape is also without parallel but round pendants are common in megalithic tombs in Ireland and Western Britain,1 and one from Ty Isaf has a similar eccentric perforation.2 The most that can be said for it is that it would be quite at home among the furniture of a megalithic tomb. THE MEDIEVAL OBJECTS Several hone fragments of hard sandstone from the robber pits probably belong here. They are parts of pyramids c. 1 inch square and perhaps 9 in. long. THE HUMAN SKELETAL REMAINS By L. BAREIELD, B.A. More than 3,500 fragments of cremated human bone weighing approximately 4,490 gms. were found. They were in and around the chamber and had been disturbed by the robbers. Most of the fragments were from 1-5 cm. long, but a few of the long-bone shafts measured 9-10 cm. and were therefore considerably less broken up than in many prehistoric cremations.3 Ashy-white to buff, darkening to grey or blue-black in the centre, was the predominant colouring. Occasional bones were black throughout or heavily charred on the outside. No pathological abnormalities were noted and no bones could be sexed with any certainty. The minimum number of individuals represented could be calculated from fragments of the petrous portion of the temporal bone, which at this site was the part of the skeleton most resistant to cremation. Out of 21 fragments eight belonged to the right, and nine to be left temporal 1 Piggott, op. cit., p. 202 and 207. 2 Ibid., p. 146. 3 For example Barclodiad Y Gawres where the fragments were under 2 cm. long, (F. P. Lisowski in Powell and Daniel, Barclodiad Y Gawres, p. 63) and Dorchester where all were under 2 in. long (J. G. Weiner in Atkinson, Piggott and Sanders Excavations in Dorchester, p. 129). 52 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB bone so that at least nine skeletons were present. Several fragments of immature cranium suggest that one or two infants could be added to this total. Two uncremated human molars were also found. These make it possible that the tomb contained inhumations as well as cremations.1 TABLE. Cremated Cranium Mandible and Maxilla Teeth Vertebrae Ribs Scapula Metacarpals, -tarsals and phalanges Long-bones Total Uncremated Teeth 2 Wt. (gms.) 1,136 85 28 57 56 28 90 1,848 4,490 (74-4 %ot total 25-3 1-9 0-6 1-3 1-3 0-6 2-0 41-4 % of this was ide %of identified bone 34-1 2-6 0-8 1-7 1-7 0-8 2-7 55-6 sntified) ACKNOWLEDGEMENT I am grateful to Mr. D. Brothwell for his comments and advice. ANIMAL BONES Report by L. BAREIELD, B.A. None cremated. Bos Left side of pelvic girdle with acetabulum (depth 10 in.). Tibia. Milk molar. 1st and 2nd molar teeth set in fragment of mandible. Sheep Two molar teeth in fragment of mandible. Deer Molar tooth. None of these bones were in undisturbed deposits. I doubt if any of them are very old. 1 The acid soil of Addington would have destroyed the uncremated bones for several animal bones of recent date had already been reduced to the consistency of butter. 53 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 2 I 8 9 FIG. 15. I Inch CLAY PIPES Report by L. BABBTELD, B.A. Among the numerous fragments of clay pipes found in the top soil the following are worthy of note : 1. Resembling Oswald's group 8b1 1680-1720. 9 ^* a a a a 3. Probably nineteenth century. 4. Resembling Oswald's group 9a, back of bowl stamped, B (IorU) RCH (EL?) L. Birchell or Burchell. The stamp has been applied twice and is partly illegible. 1 A. Oswald. " The Evolution and Chronology of English Tobacco Pipes ". The Archaeological Newsletter, Vol. 5, No. 12 (1956), 243. 54 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB 5. Oswald's group 10 with flat spur, eighteenth century. Stamped TH. 6. Nineteenth century decorated. 7. Stamp of no. 5. 8. Stamp RH on a similar foot, slightly flatter and wider. 9. Stamp on a fragment of a flat spur similar to no. 5. 10. Resembles Oswald's group 10? Similar to 5. Stamped I on one side and C on other. 11. Similar to 10. Stamped with rosette. Also the feet of three other unmarked bowls similar to no 1, Oswald's type 8b. APPENDIX I THE STONES OF THE CHAMBER AND FACADE The 18 large boulders of the chamber and 4 small, but unbroken, ones in the drystone wall and the pavement, were of sarsen. Similar boulders are scattered over the surrounding countryside and some have the same rectangularity and even the small patches of polish of the chamber stones. A few miles east of the site in the Aylesford region, Mr. A. McCrerie1 has located 197, most of them in large undisturbed groups. North of the site more are found in the direction of the Coldrum and beyond it to the Chalk escarpment.2 The stones used need not therefore have been brought far and were probably unshaped. Detailed Description of the Stones3 A. Length 5 ft. 9 in.; Breadth 5 ft. 0 in.; Thickness 1 ft. 9 in. Very weathered. Broken over half of circumference. Penetrates only six inches into present turf with mediaeval sherds underneath. Probably part of " V ". B. Tength 12 ft. 0 in.; Breadth 7 ft. 10 in.; Thickness 3 ft. 1 in. The largest stone. Badly frost fractured on upper side.4 Near rectangular and of even thickness but unshaped. Small area of polish on under side. Capstone of eastern trilithon. Not sunk deeply into turf, mediaeval sherds underneath. C. Length 8 ft. 8 in.; Breadth 7 ft. 8 in.; Thickness 1 ft. 9 in. Badly weathered on upper side and north edge. Trapezium of • even thickness but no sign of tooling. Not sunk into turf at all. Capstone of west trilithon. 1 I am indebted to Mr. McCrerie for this information, his maps are deposited with Maidstone Museum. 2 J.R.A.I., 1913, p. 80, Map G, supplemented by information from Mr. R. Boyle. 8 All measurements given are the maximum ones. 4 All stones are given their pre-excavation positions (Plan 1). 55 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB D. Length 5 ft. 6 in.; Breadth 3 ft. 2 in.; Thickness 1 ft. 5 in. Very broken edges. Driven into ground by " B " when it slid off. Probably upper part of " F ". E. Length 5 ft. 1 in.; Breadth 4 ft. 0 in.; Thickness 1 ft. 9 in. No broken surfaces and can never have been larger. Whole of upper face polished. Narrows to knife-edge at west end. In robbers' spoilheap in chamber. F. Length 7 ft. 6 in.; Breadth 7 ft. 9 in.; Thickness 1 ft. 7 in. Badly broken but still sub-rectangular. Whole of top missing. Bedded on stones, still in original position. Wallstone. G. Length 10 ft. 0 in.; Breadth 5 ft. 1 in.; Thickness 2 ft. 2 in. Surface badly broken but was of fairly even thickness. Underside flatter than upper and the north end was polished. Fell across pavement. Entrance Blocking Stone. H. Length 9 ft. 8 in.; Breadth 6 ft. 7 in.; Thickness 2 ft. 9 in. Badly weathered on upper side and most of top missing. Near rectangular shape. Bedded on stones and unmoved. Wallstone. J. Length 7 ft. 6 in.; Breadth 5 ft. 0 in.; Thickness 2 ft. 8 in. Rectangular with conic top. Little weathered. Pushed over by robbers. Little damaged. Facade stone. K. Length 4 ft. 0 in.; Breadth 5 ft. 10 in.; Thickness 1 ft. 3 in. Very broken, fragment of larger stone, probably " F ". Fell on drystone wall and mediaeval dump. L. Length 6 ft. 0 in.; Breadth 5 ft. 0 in.; Thickness 2 ft. 0 in. Badly broken on two sides. Part of larger stone, almost certainly the other half of " V " . Dragged, after collapse of chamber, on to spoil dump. M. Length 6 ft. 5 in.; Breadth 6 ft. 1 in.; Thickness 2 ft. 4 in. Badly damaged, top third missing. Tapers at base into hole packed with stones. Stake holes from the erection found. East face much smoother than west. The broken top is the only piece of stone not found. Facade stone. N. Length 2 ft. 10 in.; Breadth 2 ft. 0 in.; Thickness 1 ft. 0 in. Small fragment of a larger stone disturbed by bulldozing. P. Length 2 ft. 6 in.; Breadth 1 ft. 10 in.; Thickness 2 ft. 0 in. Small fragment of larger stone lying on present humus. Q. Length 3 ft. 8 in.; Breadth 2 ft. 9 in.; Thickness 2 ft. 4 in. Broken on three faces, part of larger stone, probably " V ". Fell on to mediaeval spoil dump. R. Length 8 ft. 10 in.; Breadth 5 ft. 3 in.; Thickness 2 ft. 9 in. Triangular and of even thickness except at base. In pit packed with stones. Facade stone. 56 EXCAVATION OF THE CHESTNUTS MEGALITHIC TOMB S. Length 12 ft. 6 in.; Breadth 5 ft. 6 in.; Thickness 2 ft. 6 in. The longest stone found. Tapers at base and apex. Little weathered and of regular thickness. Stones under base. Wallstone. T. Length 7 ft. 4 in.; Breadth 8 ft. 0 in.; Thickness 2 ft. 2 in. Triangular and of even thickness. Little weathered. Pushed over by mediaeval robbers. Facade stone. V. Length 4 ft. 0 in.; Breadth 5 ft. 0 in.; Thickness 2 ft. 2 in. Badly broken, probably a quarter of its original size. Of even thickness and flat bottomed. Bedded on stone. Wallstone. W. One additional small fragment was buried south of the chamber in 1952. 57
Previous
Previous
Annual Report (1961)
Next
Next