Ancient Fields

Pages 622-634

An Inventory of the Historical Monuments in Dorset, Volume 2, South east. Originally published by Her Majesty's Stationery Office, London, 1970.

This free content was digitised by double rekeying and sponsored by English Heritage. All rights reserved.


In this section


Ancient fields are described in groups, numbered generally from E. to W., independently of the parish arrangements; for ease of reference the groups are given topographical names, although these cannot always be precise. These groups consist mainly of 'Celtic' fields but include later field remains where the two are so intermingled as to make separate description difficult; strip lynchets and broad ridge-andfurrow will otherwise be found in the parish inventories. Other earthworks similarly intermingled or of uncertain nature, sometimes possible settlements, are described with the fields and usually identified by letters in parentheses. The first three groups include remains within the bounds of Dorset I and parish monument numbers used in that volume have been retained; settlements and other sites omitted from that volume will be described in addenda to Dorset I which are reserved for inclusion in the last Inventory of the Dorset series.

The term 'Celtic' fields is used to indicate all fields of regular shape laid out before the Saxon Conquest. (fn. 1) Their remains cover an area of 4,000 acres in South-East Dorset (Fig. opp. p. 634). Most are on chalk and limestone but some are on the Clay-with-Flints often capping the chalk hills or on sandy and gravelly areas adjoining chalk soils (Groups 25 and 26). They occur on slopes of up to 25° facing all directions, on valley bottoms and on the summits of ridges, at heights from below 100 ft. to more than 750 ft. above sea level. The remains described below are, however, only what is left of much more extensive tracts of 'Celtic' fields. How great an area they formerly covered is unknown, since they only survive on land which was largely uncultivated in later periods. The boundaries of the groups are thus generally determined by later destruction, and some (13, 15 and 29) once extended over land now eroded by the sea.

There are many traces of later ploughing over areas of 'Celtic' fields, whether in the form of strip lynchets, (fn. 2) as in Group (19), or broad ridge-and-furrow (Group 1), both mediaeval in form, or of 18th and 19th-century 'narrow rig'. In Group (4) narrow rig overlies both 'Celtic 'fields and strip lynchets. In addition to this ploughing which, if it had continued long enough, would have destroyed the earlier remains completely, these fields have been damaged by quarrying for flints, as in Groups (11) and (15), or by landslips and weathering on exposed and steep slopes (Groups 17, 18, 29). Because of this widespread destruction any conclusions based on the distribution of the surviving fragments must be severely limited. In the only extensive stretch of undisturbed land suitable for cultivation, the Purbeck Hills, original gaps between areas of 'Celtic' fields can, however, be observed. The existence of an Iron Age settlement, probably pastoral (Church Knowle (54)), may account for one such gap.

The fields are usually defined by lynchets, i.e. scarps built up or steepened by cultivation processes, ranging in height from 1 ft. to 10 ft. or more. A lynchet 35 ft. high in Group (14) includes natural slope as well as the scarp resulting from ploughing. Slight banks are found where lynchet formation does not take place, that is, where field sides run directly uphill (Group 17), or on flat ground. Relatively massive stony banks on a limestone plateau in Group (23) suggest former walls.

The fields range in shape from squares to rectangles six times as long as they are broad, and in size from 1/6 acre to 1½ acres. There is an unusual triangular field in Group (5). On Ballard Down (Group 29) are many small fields, frequently with sides of 23 yds. to 26 yds., while some fields in Group (8) are long and narrow, 100 yds. by 21 yds. In Group (15) there are over fifty fields with a proportion of roughly two to one and a size not far removed from the Roman measure of two square actus (240 by 120 Roman feet or 78 yds. by 39 yds.).

Ancient Field Group (1), Compton Valence, Frampton and Winterbourne Abbas

The development of the groups is often complex but certain patterns can be seen within them. The long axes of fields may run in the same direction over large areas despite the existence of valleys (Group 3); a notably regular pattern may appear to replace one less regular (Group 17); fragmentary remains on steep S. slopes in three Groups (17, 18 and 29) may represent sporadic development or the effects of weathering; an arrangement of field angles staggered on the downhill side is common, and was perhaps intended to facilitate access from one level to another over a field side rather than over its steeper lower edge.

Settlements which are certainly associated and contemporary with the 'Celtic' fields during their latest period of widespread use appear to belong to the Roman period. In Group (17), however, the fields are apparently associated with the occupation of the Iron Age hill-fort of Flower's Barrow and in Group (3) the settlement remains seem from their form to be pre-Roman. In several groups, notably (1), doublelynchet tracks ran through the fields from settlements or past them. In Group (23) boundaries between the farm land of contemporary settlements can seemingly be traced and can perhaps be recognised elsewhere (Groups 3, 15, 27). There is not enough evidence, however, to assess the area attributed to any settlement.

Two Neolithic long barrows, Bere Regis (66) and Portesham (33), were used as field boundaries in Groups (30) and (5). Bronze Age round barrows, whether singly or in groups, were generally respected by the farmers cultivating 'Celtic' fields so that a group would be left on an island of unploughed ground (Group 3) or a barrow be left near the angle or side of a field.

Other dating evidence for the fields is scanty. In Group (3) their relationship to a linear ditch, probably of the Bronze Age, suggests origins in that period. A Late Bronze Age hoard found in the area of Group (15) may have been hidden in a field lynchet. Iron Age pottery has occasionally been found in the fields but Romano-British pottery is much more common. The Roman road running W. from Dorchester and the aqueduct serving the Roman town probably cut across 'Celtic' fields (Group 2). There is no evidence of continuous use of the surviving fields in the post-Roman periods, but mediaeval farmers sometimes incorporated certain of the scarps bounding 'Celtic' fields in their strip lynchets, while ploughing over others (Group 19). In Group (15) the parish boundary between Chaldon Herring and West Lulworth partly follows the line of a division between blocks of 'Celtic' fields.

Group (1): Westminstone to Town Hill (SY 59 SE, 69 SW; 600920–628920; Fig. opposite). 'Celtic' fields are virtually continuous over about 900 acres, mostly Upper Chalk but with a capping of pebbly clay and sand on the main ridge top, in the S. of Frampton and Compton Valence (fn. 3) parishes and the N. of Winterbourne Abbas parish. They were formerly more extensive and have mostly been overlaid by or integrated into a later pattern of broad ridge-and-furrow some of which has, in turn, been ploughed in narrow rig. This broad ridge-and-furrow on the downland probably represents a temporary or sporadic extension of cultivation beyond the limits of the permanent arable in the mediaeval period or, possibly, later. Modern ploughing is completing the work of destruction throughout the whole group. The remains have been further disturbed by numerous small quarries and chalk pits, now disused. Two settlements in Frampton, (a) on Little Hogleaze (613922) (Plate 231) and (b) above West Hill Bottom (605928), Romano-British at least in their final phases, are contemporary with the 'Celtic' fields as is a system of tracks probably linked to the Roman road running W. from Dorchester on a line near that of the present ridgeway road.

(i) Westminstone Down (Plate 231). 'Celtic' fields, mostly overlain by broad ridge-and-furrow on the gentle slopes of the higher ground, cover 60 acres in the S.E. of Compton Valence parish. Surviving fields have sides 40 yds. to 50 yds. long and are bounded by lynchets up to 10 ft. high or by low, flinty boundary banks. A track running from the line of the Roman road towards settlement (a) apparently formed a boundary to the S.

(ii) West Hill Bottom and Little Hogleaze, at W. of Frampton parish. 'Celtic' fields of ½ to ¼ acre in size, where complete, are overlain by disturbed broad ridge-and-furrow. Tracks 4 yds. to 7 yds. wide, mostly of double-lynchet form, run into settlements (a) and (b). It is possible that the latter was developed on old fields and certainly both settlements seem to have extended on to former fields.

(iii) Town Hill and Great Hogleaze. 'Celtic' fields with lynchets up to 7 ft. high are overlain in an area of at least 60 acres by broad ridge-and-furrow. The fields often have staggered angles but an almost continuous run of lynchets from settlement (a) to 621929 suggests a boundary. Two complete 'Celtic' fields near the centre of the area measured only 25 yds. square but others, to the N., have been thrown together to form long fields. Enclosure Frampton (20) has no demonstrable link with the fields but is probably prehistoric. Curving scarps in a copse at 621927 suggest the possibility of contemporary settlement.

(iv) Eweleaze to Higher Ground. 'Celtic' fields, disturbed by later ploughing, stretched S. from Pigeon House Cottage, near which there are remains of strip lynchets, to Higher Ground where, after a wide gap with no sign of ridging, they are overlain by broad ridge-and-furrow, and at the S.E. further overlain by narrow rig. A hollow-way (?), Frampton (21), and a faint track which joins it from the S.E., as well as a continuation W. of both, apparently belong to a phase of 'Celtic' fields which end against them. The deep hollowing of the track is unusual in a 'Celtic' field system and is probably due to subsequent wear or to the utilisation of a ditch line.

(v) Winterbourne Abbas, S. of the Roman road. Fragmentary traces of 'Celtic' fields extend for about 1 mile. Around 603919 they are apparently overlain by broad ridge-and-furrow which perhaps continued S. to meet the strip lynchets of the open-field system (see Winterbourne Abbas (12)). An undated enclosure (Winterbourne Abbas (55), Fig. p. 507), was built over 'Celtic' fields, but its relationship to the ridge-and-furrow was not clear. On the N., on the approximate line of the Roman road and traceable for over a mile, is a track with a bank on the N. separating it from the modern road. It is defined on the S. by a low scarp from which 'Celtic' lynchets occasionally run. (R.A.F. V.A.P. CPE/UK 2431: 3141–3.)

Group (2): Bradford Peverell (fn. 4) and Winterborne Monkton (SY 69 SE; 662917–667902; Fig. p. 494). Traces of 'Celtic' fields, mutilated by a system of strip fields and modern ploughing, cover some 300 acres on Fordington Down but formerly extended in all directions. Two Romano-British settlements at 667907 and 666910, now destroyed, lay within the fields (Dorset Procs., LXXVIII (1956), 80–1). Most of the fields are N. of the Roman road W. from Dorchester which appears, together with the Roman aqueduct at the S.W. of Fordington Bottom, to have cut across 'Celtic' fields. Only a few scarps survive S. of the road in Winterborne Monkton. (R.A.F. V.A.P. CPE/UK 1934: 5087–9 and 2088.)

Group (3): Little Bredy (fn. 5) and Winterbourne Steepleton (SY 58 NE, 68 NW; 595872–613893; Fig. opposite, Plate 232). 'Celtic' field remains, some very well preserved, cover about 450 acres between Little Bredy Farm, Black Down Barn, Sheep Down and Loscombe. Similar remains continue W. and S.W., and probably a further 100 acres of remains have been destroyed within the area. Strip lynchets run into the 'Celtic' fields from W. and N.E. while broad ridge-andfurrow, partly defaced by later narrow rig, crosses them at the extreme W. just off the plan.

The geology is complex. Much of the area is Upper Chalk but the Valley of Stones, named after a spread of sarsens, is partly covered by Clay-with-Flints and Bagshot Beds while the high ground of Black Down bears red sandy clay with cobbles. N. of this a mixture of gravels and sands is interspersed with patches of chalk. With the exception of Black Down and the Valley of Stones most of the area is arable.

The long axes of the 'Celtic' fields run predominantly N.W. to S.E., partly because they were laid out up and down the slopes of the valleys which here generally run S.W. to N.E. The fields lay across hilltops, valley bottoms and slopes on all aspects. Despite the abundance of sarsens no field sides are visibly marked by them but some lynchets probably incorporate walls for marking out fields. Occasionally field sides running up and down slopes are marked by low banks.

The following monuments are apparently contemporary with some phase of the 'Celtic' fields: Winterbourne Steepleton (64) settlement, (66) dyke; Little Bredy (6) enclosure, (7a, b) sarsen-walled enclosures; also hut circle (a) and settlement (b). Platforms (e, f) are of uncertain date. Enclosures Little Bredy (5a, b) and Winterbourne Steepleton (63) are almost certainly later than the 'Celtic' fields. The long barrow Winterbourne Steepleton (13) and round barrow (15) were incorporated into the pattern of fields but not ploughed over in antiquity. Round barrows are, however, mostly grouped in areas apparently never cultivated as 'Celtic' fields, as on Sheep Down and S. of settlement Winterbourne Steepleton (64). The round barrows in square enclosures, Winterbourne Steepleton (24–6), are arguably later than the early 'Celtic' fields.

(i) Little Bredy Farm. S. of the farm, near the W. end of the Group, a complex of slight ditched banks, probably relatively recent, cuts across 'Celtic' field lynchets. A probable hut circle (a) with a floor 14 ft. across lies beneath the end of a 'Celtic' field lynchet at 59438800. Further slight banks, again comparatively recent, divide the hilltop and overlie 'Celtic' field remains on the spur top N. of the farm. Four earthen circles here (Little Bredy (15), see Dorset I) are almost certainly of recent origin, possibly the remains of landscaping associated with Bride Head to the N. On the shoulder of the hill ¼ mile E. of Little Bredy Farm at 59758820 is a small settlement (b), probably pre-Roman, comprising four or five hut circles set immediately below well-developed lynchets at the edge of largely destroyed 'Celtic' fields. Strip lynchets occur N., W. and E. of the farm indicating the former extent of the open fields of Little Bredy.

(ii) Crow Hill and the Valley of Stones (Little Bredy (17)). On the E. shoulder of the heavily ploughed summit of Crow Hill are two possible hut platforms (c) and (d), the latter set into the top of a 'Celtic' field lynchet. The large valley-bottom enclosure Little Bredy (5a) was apparently built over 'Celtic' field remains, already reploughed. Enclosure (5b), also on the valley floor, may be connected with two unusual long narrow plots set within 'Celtic' field boundaries above it. Immediately W. a flat floor, 18 ft. square with a gap towards (5b), has been cut into the face of a lynchet.

Ancient Field Group (3) in Little Bredy and Winterbourne Steepleton

In the Valley of Stones at least seventeen complete 'Celtic' fields may be traced, ranging in area from ½ acre to 1¼ acres. Five are nearly square and eight are about twice as long as they are broad. Nine fields have sides about 50 yds. long, the remainder vary from 33 yds. to 125 yds. The fields are rarely found on slopes of more than about 10° and are bounded mostly by lynchets, up to 12 ft. high, but occasionally by low banks. At the foot of the W. side of the valley is a band of sarsens, the product of clearance, against which lynchets have formed.

A D-shaped, sarsen-walled enclosure (Little Bredy (7a)), in the valley bottom, possibly the site of a dwelling or of a pound, was apparently built at the edge of a small 'Celtic' field which continued in cultivation. A short distance to the S.W. is a triangular platform (e), 38 ft. by 30 ft., at the angle of a 'Celtic' field, while to the N.E. is a rectangular platform (f), about 42 ft. by 30 ft. On the valley floor 530 yds. N.N.E. of (7a) is a much smaller and less well-marked sarsen enclosure (7b).

(iii) Black Down. The ground above 600 ft. was apparently not much ploughed in the 'Celtic' field phase, but is greatly disturbed. At (g) a bank appears to run N.N.E. among deepcut hollow-ways while a slighter bank runs N.W. at right angles to it. 100 yds. farther S. a terrace-way runs approximately N. to S. for 150 yds. above well-developed 'Celtic' fields. On the N. edge of Black Down a dyke (Winterbourne Steepleton (66)) runs W. across the parish boundary with Little Bredy into the head of a small combe where it turns S. and ends abruptly below a field angle. Several shallow scoops (h) on the slope N.W. of this recall those at the Iron Age settlement on Knowle Hill (see Settlements, Church Knowle (54)). 'Celtic' fields run up to the dyke on the N. side and an enclosure (Little Bredy (6)) is attached to it.

(iv) Sheep Down and Cowleaze. The slight bank (j) with a ditch on the S., running N.W. for some 450 yds. across Sheep Down, was possibly an original boundary between 'Celtic' fields and the unploughed area containing barrows. Cowleaze was mostly covered with 'Celtic' fields, later overlaid by recent banks forming enclosures, and by strip fields. The latest phase of ploughing has often formed secondary lynchet lines under the 'Celtic' field boundaries giving a false impression of double-lynchet tracks. Some of the better preserved fields lie on the E. slopes of the narrow combe running S.W. from Loscombe Plantation. (R.A.F. V.A.P. CPE/UK 1934: 1069–71, 3095–100.)

Group (4): Portesham Withy Bed, N. of (SY 58 NE; 593863; Plates 140, 233). 'Celtic' fields, partly overlain by strip lynchets, and strip lynchets which probably obliterate 'Celtic' fields, almost all ploughed over in narrow rig, cover a re-entrant valley of about 30 acres. Other strip lynchets lie immediately N., W., and E. of the Withy Bed (see Portesham (32a)), and there are traces of scarped fields, almost destroyed, extending on higher ground W. and E. from the valley. A terraceway climbs the steep natural scarp N. from the Withy Bed where finds of pottery, iron, bronze, bone and flint (Dorset Procs. XXXVII (1916), xxxix), now lost, suggest Romano-British and earlier occupation.

Only one certainly complete 'Celtic' field, of about ⅓ acre, is detectable at the head of the valley on the W. side, where 'Celtic' field traces generally survive. The strip lynchets are confined to the valley floor and E. side where the treads are generally flat and the risers pronounced, indicating long usage. Some are high up the steep valley side on ledges. They appear to mark the limit of an open field extending from Portesham village. The narrow rig around 590860, which is set against a slope of 16½°, cuts across the end of a strip lynchet (cf. H. C. Bowen, Ancient Fields (1961), plate IIb). (R.A.F. V.A.P. CPE/UK 2431: 3288–90.)

Group (5): Benecke Wood, S. of (SY 68 NW; 609867–618868), in Portesham parish, comprises about 30 acres of disturbed 'Celtic' fields on the N.-facing slope and floor of a dry valley in the Upper Chalk, partly capped by Clay-with-Flints. Other certain but very slight traces appear N., W. and E. To the N., scarps almost ploughed-out can be seen on the chalk but not on the gravels of the Bagshot Beds just beyond, where enclosure Portesham (57) lies. To the W., on Portesham Hill, the Hell Stone long barrow (33) has been used as a 'Celtic' field boundary, and S. of it strip lynchets run into 'Celtic' fields. To the E. and S.E., in and near Hell Bottom, a few 'Celtic' field scarps survive around Bench (623865), partly overlain by strip fields, and S.W. of Shilvinghampton Barn (about 627861).

In the main field area a triangular 'Celtic' field, 150 yds. by 130 yds. by 80 yds., lies on the fairly steep valley side under a double lynchet track following the S. shoulder. It is 1¼ acres in area with lynchets up to 10 ft. high. The only other measurable fields, with low lynchets, are 67 yds. by 33 yds. (R.A.F. V.A.P. CPE/UK 1934: 1066–9.)

Group (6): Shorn Hill to Ridge Hill (SY 68 NW, NE; 631876–657869; Fig. p. 626), in Winterborne St. Martin and Portesham parishes, includes 'Celtic' fields and undated long fields. Broken remains of 'Celtic' fields cover at least 200 acres of the spurs and re-entrants N. of the Ridgeway and 20 acres immediately S. of it, about 647865. The most prominent, though very incomplete in plan and not illustrated, are at the extreme E., N. of Ridge Hill, covering 12 acres around 655868. Fragments extend N. of these on the spur top and in the valley W. of it leading to Ashton Cottages. Most are on chalk but some, S. of Shorn Hill, are on a sandy wash and pebble capping. There are only two complete fields, each of about ¼ acre, but original field sides are between 30 yds. and 60 yds. long. Lynchets are up to 8 ft. high, N. of Ridge Hill, on a maximum slope of 15°. Subsequent cultivation has flattened many former field divisions and in places has left 'Celtic' lynchets in continuous lines along the contours giving a misleading effect of strip fields, as on the S.E. side of Great Hill. A slight boundary bank and ditch, of mediaeval or later date, runs N. from a round barrow (Winterborne St. Martin (58)).

Ancient Field Group (6) on Shorn and Great Hills, Winterborne St. Martin

N.W. of Shorn Hill and S.W. of Great Hill are long narrow fields, apparently later than the normal 'Celtic' fields but bounded by contour lynchets or, up and down hill, by low flinty banks. In places they may have utilised the lynchets of previous fields as boundaries. Complete fields are usually 26 yds. or 37 yds. wide and from 100 yds. to 160 yds. long. On Great Hill faint traces of slight banked fields of similar type have been much disturbed by narrow rig ploughing. In proportion and form these long fields are akin to 'Celtic' long fields, elsewhere thought to be Romano-British, but some doubt arises from their extent and from the absence of a known Romano-British settlement. There are relatively recent downland fields of this form (e.g. near Stratton, around 652950), while mediaeval cultivation of the downland is attested for Winterborne Asshe (Ashton, Inquisitions post mortem, P.R.O., C. 142/216/29). By 1765 some of the enclosed strips of Winterborne St. Martin resembled in plan these long fields (cf. Isaac Taylor's map of Town Field Farm in D.C.R.O.). (R.A.F. V.A.P. CPE/UK 1934: 1063–66 and 3092–94.)

Group (7): Bincombe Tunnel (SY 68 NE; 673855), in Bincombe and Weymouth parishes. 'Celtic' fields are clearly seen only N.E. of the S. tunnel entrance on the S. shoulder of the chalk ridge about 400 ft. above O.D. (Fig. p. 24). Two complete fields are 88 yds. by 40 yds. Lynchets are generally about 3 ft. high, occasionally 5 ft. Very faint remains over about 20 acres continue N. over the crest. One air photograph, by Major G. W. G. Allen (Ashmolean Museum, No. 154), shows round barrows (Weymouth (416–18)) at 'Celtic' field angles. (R.A.F. V.A.P. CPE/UK 1821: 6447–9.)

Group (8): Bincombe Hill (SY 68 SE; 687847–691847), in Bincombe parish. Traces of 'Celtic' fields over 15 acres mostly on the crest and S. shoulder of Bincombe Hill include, towards the W., three long fields with lynchets about 1½ ft. high on a slope steepening to 11°. Two are 100 yds. by 21 yds. and one about 120 yds. by 30 yds. Ploughing at some uncertain period has almost destroyed fields S. of the line of barrows and removed the E. ditch of round barrow Bincombe (46). 'Celtic' fields once continued N. into the area of Bincombe 'North Field' (see Fig. p. 24) where relatively recent ploughing has obliterated almost all earlier field traces. Sherds of Romano-British coarse ware were found mostly N. of the barrows. (R.A.F. V.A.P. CPE/UK 1821: 6445–6.)

Group (8) on Bincombe Hill, with S.E. part of Barrow Group (R.10).

Group (9): Whitcombe, S. of (SY 78 NW; 712870), in Whitcombe parish. Remains of 'Celtic' fields survive over 12 acres of ground falling gently N., the former Eweleaze. Where still unploughed, lynchets are up to 6 ft. high, some cut by quarries. There are further traces N.E. (on 'New Ground' of a late 18th-century Lulworth Castle estate map, in D.C.R.O.) and the faintest traces suggest a former spread towards the Ridgeway. (R.A.F. V.A.P. CPE/UK 1934: 1054–5.)

Group (10): West Hill (SY 78 SW; 703845–710848), in the N.E. of Weymouth parish, comprises incomplete 'Celtic' field remains over about 11 acres between 350 ft. and 450 ft. above O.D. on a S.-facing spur and above the valley head E. of it, where round barrows Weymouth (419) and (420) are at 'Celtic' field angles. Re-ploughing at some unknown date has broken down former field divisions. (R.A.F. V.A.P. CPE/UK 1821: 6442–4.)

Group (11): Poxwell and Moigns Down (SY 78 SW, SE; 744836–756835) in Poxwell, Watercombe and Owermoigne parishes. Remains of 'Celtic' fields survive over about 30 acres on, and N. of, the limestone ridge from E. of Poxwell Lodge to Moigns Down and on a chalk spur N. of it. Only two complete 'Celtic' fields survive N. of Moigns Down. The area has been much disturbed by quarrying and mediaeval and later cultivation.

Group (11). Platform (a) on Moigns Down.

N. of the ridge, partly in Poxwell and partly in Watercombe, are fields marked by low N.-S. lynchets crossing the valley bottom. These were covered with narrow rig and probably represent much altered 'Celtic' fields. S. of them, on the N. slope of the ridge, disturbed terraces appear to be the remains of strip lynchets. A trackway, 12 ft. wide, associated with the remains of 'Celtic' fields, followed the N. shoulder of the ridge from 74408355 eastwards into Owermoigne parish where it joined the N. edge of a possible settlement area related to a platform (a) (753835; Fig. above ) on the S. shoulder of the ridge. This platform, 1/5 acre in area, is bounded on the S. by a low bank above scarps up to 6 ft. high from the angles of which 'Celtic' fields appear to have been laid off. A lynchet above the N. side of the platform is due to subsequent ploughing of the ridge top. (R.A.F. V.A.P. CPE/UK 1821: 4440–3.)

Group (12): Ringstead Barn (SY 78 SW, SE; 744829–754824), in Osmington parish. Traces of 'Celtic' fields between 250 ft. and 400 ft. above O.D. cover 40 acres of the gentle N.-facing slope at the extreme N.E. of Spring Bottom Hill, mostly on Wealden Beds with bands of greensand and gault. Lynchets are generally about 3 ft. high but up to 10 ft. towards the S. Four round barrows (Osmington (38–41)) on the ridge crest, only 11 yds. to 14 yds. from the southernmost lynchet line, lie on ground that was perhaps never ploughed in antiquity. The area, though adjacent to Up Ringstead, was presumably pasture for W. Ringstead, of which the strip fields (Osmington (28)) are due S. (R.A.F. V.A.P. CPE/UK 1821: 4441–3.)

Group (13): Ringstead Bay (SY 78 SE; 766817–768814), Owermoigne. Remains of 'Celtic' fields cover 14 acres of the bottom and N.-facing slope, rising to a cliff edge, of a narrow combe, and formerly extended N. and S. The inner edges of some fields, with lynchets up to 6 ft. high, are preserved on the cliff edge. (R.A.F. V.A.P. CPE/UK 1821: 2442–3.)

Group (14): Brimstone Bottom (SY 78 SE; 770825–775826), Owermoigne and Chaldon Herring. Fragmentary and partly quarried 'Celtic' fields covered about 20 acres between 300 ft. and 400 ft. above O.D., on chalk, along the steep N.-facing slope above Brimstone Bottom. Surviving lynchets are high, one scarp reaching 35 ft. (R.A.F. V.A.P. CPE/UK 1821: 4439–41.)

Group (15): Chaldon Herring, West Lulworth and Winfrith Newburgh (SY 78 SE, 88 SW; 784808– 786819–823828–808803; Plate 234), includes about 850 acres of 'Celtic' fields. Traces of others, almost destroyed by mediaeval and later ploughing, occur to the E., around Hambury Tout (815805), and between blocks (i–iii) below, suggesting that they were formerly continuous. All the remains lie between 100 ft. and 500 ft. above O.D. on chalk, 'angular flint gravel', slurry (in the combe bottoms), or sands and clays of the Reading Beds. Lynchets up to 12 ft. high indicate long cultivation and there is also evidence, especially in block (i), for change and development within the 'Celtic' field phase. Major boundary lines appear in block (iii) and in block (i), which is the most extensive and has a large number of complete fields. Strip lynchets from Chaldon Herring (24), and Winfrith Newburgh (30), lie partly on 'Celtic' fields at the limits of the open fields. There are no certain settlements but a number of possible areas of settlement are suggested in blocks (i) and (iii). The Late Bronze Age 'Lulworth hoard' was found somewhere in block (ii) during flint quarrying. The so-called 'flintfilled hollows' which occur within all blocks are probably the remains of such quarrying, frequently into lynchets and sometimes into piles of flint at field angles. Other evidence of relatively recent activity is marked by narrow rig, by brick and rubble foundations on the site of Warren House (79238106), and by tile, brick and glazed pottery at the head of Vicarage Bottom.

(i) Chaldon Herring Warren, Newlands Warren (W. Lulworth) and Chaldon Down (Fig. opposite). Three main combes— Middle, Vicarage and Scratchy Bottoms—have floors sloping gently back from cliff edges to heads which rise steeply towards the E.-W. coastal ridgeway, beyond which narrow dry valleys run N. from Chaldon Down. 'Celtic' fields are continuous for 500 acres over bottoms, sides and intervening ridges and formerly continued seaward, where land has been lost by erosion. N. of the ridgeway they have mostly been heavily ploughed.

Lynchets are mostly 2 ft. to 4 ft. high but occasionally reach as much as 12 ft., as in Middle Bottom. There are 63 complete fields around Vicarage and Scratchy Bottoms with mean dimensions of 85 yds. by 45 yds. (just over ¾ acre). On the ridge N. of (a) eight long fields average 103 yds. by 27 yds. Barrows Chaldon Herring (38) and (39) are probably at the angles of fields, the N.-S. banks of which have been flattened. A line of very small fields, including some of 1/6 acre, survives S.E. of the block at 808805. The maximum field slope is 16°. Boundary lines are indicated by continuous runs of lynchets S. to N., one on the high ground between Middle and Vicarage Bottoms and another between Vicarage and Scratchy Bottoms which for 300 yds. is followed by the present parish boundary.

There has been some strip cultivation in the E. part of Newlands Warren (about 806805), narrow rig ploughing W. of the site of Warren House, in Scratchy Bottom and E. of it, and recent cultivation over fields just S. of the ridgeway. 'Flint-filled hollows' are prominent in Middle and Vicarage Bottoms.

Settlement sites are nowhere clear but possibly existed at:

(a) (793806) on the sheltered E. shoulder of a narrow col between Middle and Vicarage Bottoms where ill-defined platforms and scoops lie over 2 acres on a gentle slope. The area is bounded on the W. by a terraceway, on the N. by a faint terrace and on the S. by 'Celtic' fields.

(b) (795806; Plate 234), 200 yds. E. of (a), a very shallow 'angle-ditch', with outer bank, 270 ft. by 230 ft., partly encloses 'Celtic' fields on a slope of 16°. A platform 18 ft. square lies above the W. end of the ditch on the shoulder of a re-entrant gully. These features are undated, but a comparable angleditch over 'Celtic' fields in Nether Cerne (5), Dorset I, lay in an area of Romano-British settlement.

(c) (80108085) at the head of Scratchy Bottom is a quarriedout, roughly flat, trapezoidal floor, with sides between 15 ft. and 76 ft. long. It lies clear of, and 20 ft. below, a 'Celtic' field lynchet and is approached by a narrow track from the S.W.

(d) (80558065), E. of Scratchy Bottom, a pentagonal area of 1/7 acre is outlined by 'Celtic' field fragments.

Surface finds from Scratchy Bottom and from N.W. of Vicarage Bottom include Romano-British, Iron Age 'C' and possibly earlier sherds.

(ii) Sleight Bottom, Winfrith Newburgh. 'Celtic' field lynchets, some massive but all very disturbed, cover about 70 acres of a dry valley. Strip lynchets just N. and very faint traces of strip cultivation immediately E. suggest that some extension of such cultivation may have created the curving lynchets and broken-down field divisions at the N. of the block. Flint pits have been quarried into lynchets crossing the valley floor. The 'Lulworth hoard' of Late Bronze Age bronzes (actually from Winfrith Newburgh parish) was found in 1903 during flint quarrying near 805824 (Ant. J. XV (1935), 449–51) and so had possibly been hidden in a lynchet.

'Celtic' Fields in Chaldon Herring and West Lulworth, Part of Ancient Field Group (15)

(iii) Daggers Gate and Marley Bottom. 'Celtic' field remains, in the combe and on the downland to either side, are continuous for over 270 acres, from 814811, N. of Daggers Gate, to 813832 in Marley Bottom, S. of Winfrith Newburgh. They include many high lynchets, but complete original fields have all been broken down. Some field banks are packed with flint and in the combe bottom E. from 813827 there is a series of small, often circular, quarries dug into the lynchets. Most of the downland was arable on the Tithe Maps but this cannot be associated certainly with surviving narrow rig.

A track, partly double-lyncheted with tread 16 ft. wide and lower scarp 8 ft. high, runs S.E. for 600 yds. from 818829 on the E. of the block. There are faint indications of another track approaching the possible settlement area (e). A continuous boundary line, formed by a massive but ploughed-down scarp, runs S. for ½ mile from about 809827, S.W. of Marley Wood.

(e) Settlement is suggested at 815818, where surviving fields on the S.-facing slope of a dry valley are different in shape and size from most of those around. Above them to the S.W., air photographs show enclosures, one with a curved side skirted by a track from the W., on relatively flat ground. The whole covers about 3 acres. (R.A.F. V.A.P. CPE/UK 1821: 2434–41.)

Group (16): Bindon Hill (SY 88 SW; 82408040–82908033), W. Lulworth. Fragmentary 'Celtic' fields on the N. slope of Bindon Hill, near its W. end, lie over about 10 acres between 300 ft. and 500 ft. above O.D. Others probably extended S. to lower ground where faint traces appear under the remains of strip cultivation around 831805. The fields only reach the earthwork on the hill-top (see Hill-forts, W. Lulworth (53)) at their S.E. limit and seem to be unconnected with it. Lynchets are up to 10 ft. high on a slope of about 10°. (R.A.F. V.A.P. CPE/UK 1821: 5432–4.)

Group (17): Rings Hill to Povington Hill (SY 88 SE; 859807–885812), in East Lulworth and Tyneham parishes. 'Celtic' fields appear on about 100 acres of the N. slope of the W. end of the Purbeck Hills. In addition, isolated and slight scarps occur on the corresponding S. slope. Almost all are in an area generally inaccessible and very disturbed so that, although extensive, the pattern cannot be worked out in detail. Ploughing of the hill-top, marked in places by narrow rig, has destroyed all but a few lynchets there. The fields are of two types and in two blocks.

(i) Flower's Barrow, immediately N.W. and N.E. of (see Hill-forts, E. Lulworth (40), Plate 215). On a slope of up to 23° are 30 acres with squarish, small fields (some of about ⅓ acre), some irregular, marked by lynchets between about 4 ft. and 10 ft. high and arranged along base-lines which mostly follow the contour. Two 'base-lines' appear to meet Flower's Barrow, one at the N.W. angle, the other at the N.E. angle of the first-phase hill-fort. The fields also extended N. and E. from this latter point to well below the natural shoulder of the ridge.

(ii) Whiteway and Povington Hills. Fields over 20 acres are arranged in regular manner with parallel base-lines 20–40 yds. apart, up and down the slope (Fig. in pocket, Part 2; Plate 233). These are marked by slight banks or lynchets about 1 ft. high, in places running unbroken for at least 170 yds. Apart from some very slight cross-divisions there are some higher lynchets. These may belong partly to earlier fields, since there are indications that this regular pattern has been imposed on fields of the type of block (i). One such field is just outside this block, at 873811. Traces of irregular fields are more clear further E., on Povington Hill, extending over 25 acres but with suggestions of the regular, banked type inserted among them. (R.A.F. V.A.P. CPE/UK 1821: 2421–5; 5425–8.)

Group (18): W. Creech and Grange Hill (SY 88 SE, 98 SW; 893818–905816), in Steeple parish. Very faint and disturbed traces of 'Celtic' fields occur as banks or slight scarps for about a mile along the ridge of the Purbeck Hills, N. of it at the W. end of this area, and S. of it at the E. There are no complete fields. Lynchets are generally low. Fragmentary low banks, some ditched, of uncertain date, run against the S. slope at up to 24°. (R.A.F. V.A.P. CPE/UK 1821: 2419–20.)

Group (19): Tyneham Coast (SY 87 NE, 88 SE; 871796– 891799; Fig. in pocket, Pt. 2; Plate 197), consists of 'Celtic' field fragments, mostly overlain and disturbed by strip lynchets, some unusually irregular, over 5 acres of the N.-facing slope of the limestone ridge about 500 yds. S. of Tyneham village. Only at about 87807975, N.E. of Gold Down on a slope of 24°, are the fields not overlain. Lynchets are, rarely, up to 12 ft. high, as immediately S. of Tyneham Great Wood, where a 'Celtic' field angle and side, much rounded by strip ploughing, are particularly prominent. Very faint traces indicate former extensions of the 'Celtic' fields E. along the ridge as well as N. on to lower ground. (Cf. Tyneham (7, 9d), mediaeval remains.) (R.A.F. V.A.P. CPE/UK 1821: 5423–4 (part reproduced as Plate 197).)

Group (20): Smedmore Hill (SY 97 NW; 925798–936795), in Church Knowle and Steeple parishes, between 500 ft. and 600 ft. above O.D. Traces of 'Celtic' fields can be detected over some 80 acres in an area rich in Iron Age and Romano-British settlement remains. An almost certain relationship is only now visible around settlement Church Knowle (55), q.v. Strip lynchets (see Church Knowle (32)) and later narrow rig have contributed largely to the destruction of the fields.

Group (21): Encombe Obelisk (SY 97 NW; 942790–948790), Corfe Castle. Very disturbed 'Celtic' fields cover some 40 acres between about 400 ft. and 500 ft. above O.D. on the summit and gentle N. slopes of a W.-E. spur about ¼ mile N. of Encombe. On the S. the remains are bounded by a sharp edge above a steep natural slope on which there are no fields; to the N. their limit is not clear. The earliest surface finds are Iron Age 'A', but a Romano-British building (Roman section, Corfe Castle (235)) was discovered amongst the fields at the S.W. In the only complete field, about 50 yds. square, a small mound (94397913) was seen under the plough to consist of Portland stone blocks associated with burnt clay and Romano-British pottery, including samian ware. The Romano-British building lay 200 yds. S.W. of this. 'Celtic' fields near it are 20 yds. to 25 yds. wide and probably represent continued cultivation. (R.A.F. V.A.P. CPE/UK 1821: 3407–9.)

Group (22): Encombe Big Wood (SY 97 NW; 941778), Corfe Castle. Remains of 'Celtic' fields with lynchets up to 3 ft. high cover a small area immediately S. of Big Wood, 2/3 mile S. of Group (21), but formerly extended in all directions. One lynchet, crossing a settlement site to the S. (Roman section, Corfe Castle (240)), seems to have accumulated in the later Iron Age and early Romano-British period (Dorset Procs. LXXXV (1963), 98–9; LXXXVI (1964), 109). (R.A.F. V.A.P. CPE/UK 1821: 3407–8.)

Group (23): Kingston Down (SY 97 NE; 957781; Fig. p. 631, Plate 235), Corfe Castle and Worth Matravers. 'Celtic' fields, many fragmentary, cover about 150 acres of plateau and valley S. of Kingston and formerly extended in all directions except, possibly, S.; an isolated block survives at 972781, N. of Worth Matravers village (Fig. p. 414). They lie between 200 ft. and 450 ft. above O.D. on Portland Stone, Sand and Clay. The best preserved, though disturbed, are on the top and W. side of the plateau spur 1 mile S. of Kingston village, beyond the limits of the open fields. Here they are associated with two neighbouring settlements (Corfe Castle (221–2), Figs. p. 510). Three other settlement sites (Roman section, Corfe Castle (237–9)) lie in areas of destroyed fields peripheral to the group. Surface finds indicate an Iron Age or Romano-British date.

Settlement features and fields are generally marked on the spur top by stone-packed banks and lynchets. The banks, possibly tumbled walls, are up to 2 ft. high and 12 ft. wide. Curved sides suggest settlement features extending into the fields and make it difficult to see the bounds of the settlements. The enclosure with bowed sides (a), 340 yds. S.S.W. of the nucleus of (221), is apparently separated from it by fields and is itself closer to a narrow, rather broken, terraced area (b), which runs across the 'funnel' end of a track from settlement (222). Excluding this area, (221) covered nearly 6 acres and (222) nearly 5 acres.

Two contemporary tracks lead from (222): the track just mentioned, 6 ft. wide and mostly of double-lynchet form with scarps up to 4 ft. high above and below, and a terraceway, 15 ft. wide, running S.W. diagonally downhill for 250 yds. before disappearing in an area of landslip.

A continuous run of field sides, with lynchets dropping as much as 4½ ft. E., extends N. from the bend in the doublelynchet track past the nucleus of settlement (221) and may have been a boundary separating its territory from that of (222) (cf. H.C. Bowen, Ancient Fields (1961), 35–6 and plate III).

Ridge-and-furrow up to 11 yds. wide, as well as traces of narrow rig, covers large areas N.E. of the group. It is also clear in the 'Celtic' fields, as N.W. of settlement (221) where sides appear to have been removed to form long fields. The more extensive destruction on the E. half of the spur top has left no sign of ridge-and-furrow. Elsewhere, re-ploughing is occasionally marked by secondary lynchets, as at 96117803 in Worth Matravers parish where the 'Celtic' fields have been generally much altered.

Some thirteen fields appear to be undisturbed. Five, on the main spur top, are all less than twice as long as broad and from under ½ acre to nearly 1 acre in area. A number of broken fields have sides the lengths of which are multiples of 40 yds. On the W. side of the spur, where fields lie on the maximum slope of about 21°, eight fields include four between ¼ acre and ½ acre and none more than ¾ acre.

Group (24): St. Aldhelm's Head (SY 97 NE; 959768– 965754), Worth Matravers. 'Celtic' fields, variously preserved and in places overlain by strip fields, are found over about 60 acres N. and N.E. of St. Aldhelm's Head between 200 ft. and 400 ft. above O.D. (Fig. p. 414). Romano-British debris occurs nearby (see Roman section, Worth Matravers (43)). There are two relatively well preserved blocks.

(i) Pier Bottom. E.N.E. from 959759, lynchets cross the valley floor, terraceways not necessarily contemporary run along the foot of the valley sides above them to both N. and S., and other fields lie on the shoulders. One of these has a round barrow, Worth Matravers (34), containing secondary Romano-British material, at its S.E. angle and a N.W. side which curves along the contour.

(ii) St. Aldhelm's Chapel, E. of. 'Celtic' fields bounded by lynchets about 1½ ft. high and stone banks, recently exposed by ploughing, lie on a slope of up to 7° in an area that could not be adequately examined. Apart from the spread banks there was no scatter of stones on the fields. On pasture S. of this, about 96557545, disturbed lynchets are up to 7 ft. high on maximum slope of 15°.

Traces of strip fields are best seen on air photographs but remains of strip lynchets occur at 964759, S.E. of barrow (34). The name 'Middle Plains' shown near here on the Tithe Map may be connected with strip cultivation. (R.A.F. V.A.P. CPE/UK 1821: 1396–7.)

Group (25): Corfe Common (SY 98 SE; 955809–960810; Fig. p. 97), Corfe Castle. Remains of 'Celtic' fields cover 15 acres of sandy Wealden Beds on a maximum S. slope of 10°. A block of at least six fields lies relatively well-preserved at about 95758086 within a total area of some 2 acres. Lynchets are up to 8 ft. high. There is no ascertainable relationship with the barrows of Group (K). (R.A.F. V.A.P. CPE/UK 1821: 2409–10.)

Settlements and 'Celtic' Fields Kingston Down, Corfe Castle, Ancient Field Group (23)

Group (26): Sandyhills (SY 98 SE; 967815–974809), Corfe Castle. Traces of 'Celtic' fields are traceable over at least 70 acres of undulating sandy ground (Wealden Beds), muchploughed, on the E. of Corfe Common just above the 100 ft. contour. Lynchets are up to 2 ft. high. A staggered angle survives by round barrow Corfe Castle (196), but there are no complete fields. An Iron Age 'C' or Romano-British site (Roman section, Corfe Castle (233)) is adjacent but not demonstrably related. (R.A.F. V.A.P. CPE/UK 1821: 2406–7.)

Group (27): Challow Hill (SY 98 SE; 962824–969823; Fig. p. 97), Corfe Castle. Traces of 'Celtic' fields cover about 34 acres, between 200 ft. and 300 ft. above O.D., of the W. brow, domed summit and N. slope of the chalk ridge which continues the Purbeck Hills E. from the Corfe Gap (Plate 81). The E. end of the group appears to be marked by a lynchet line 1 ft. to 4 ft. high crossing the ridge at about 969823. To the E. there are no signs of 'Celtic' fields until Ballard Down, Group (29), 3½ miles on. Fields are marked by banks, some confused by narrow rig, and lynchets up to 8 ft. high, though generally much lower. Only two complete fields survive, each about 60 yds. by 67 yds., immediately S. of Rollington Wood. (R.A.F. V.A.P. CPE/UK 1821: 2406–7.)

Group (28): Mount Misery (SY 97 NE; 989791–994789– 992793), Langton Matravers. Very broken 'Celtic' fields cover about 20 acres on the gentle N. slope from the limestone ridge between 350 ft. and 220 ft. above O.D. Lynchets are up to 6 ft. high. Narrow rig and former village closes add to disturbance by quarrying and ploughing. The Wilkswood Romano-British site (Roman section, Langton Matravers (42)) lay beyond the surviving fields at the foot of the slope. (R.A.F. V.A.P. CPE/UK 1821: 5403.)

Group (29): Ballard Down (SZ 08 SW), in the parishes of Studland and Swanage. 'Celtic' fields, some with the remains of mediaeval and later ploughing, survive on Ballard Down and behind Handfast Point. The remains are in disconnected blocks, totalling about 65 acres, which may once have joined. Some have disappeared into the sea near Handfast Point. Most are fragmentary but (iii), now destroyed, was notable for its well-preserved small, almost square, fields. All the 'Celtic' fields are separated from the Romano-British buildings at Woodhouse Hill (Roman section, Studland (46)), by the mediaeval open fields.

(i) N. slope of Ballard Down, W. end (019813–031813). Traces of 'Celtic' fields cover about 18 acres on a slope of up to 15°. Only one complete field survives, measuring 46 yds. by 35 yds., above a lynchet 10 ft. high. A negative lynchet 2 ft. high runs along the slope only 12 yds. below the undisturbed edge of Ulwell Barrow, Studland (29), suggesting the upper limit of the fields at this point.

(ii) S. slope of Ballard Down (030812–035811). The only remains are low scarps on each side of a natural gully, the ground sloping at 25°.

(iii) N. slope of Ballard Down, centre (035815–039814). 'Celtic' fields covered 12 acres on a maximum slope of 19° and formerly extended in all directions. Where the fields were not square, the long axes were generally up-and-down. Areas were from 1/6 acre to ½ acre, thirteen sides being between 23 yds. and 26 yds. Lynchets were up to 8 ft. high. A double-lynchet track running diagonally S.W. up the hill was blocked at the top by a scarp 2 ft. high apparently forming the lower side of a 'Celtic' field. (See H. C. Bowen, Ancient Fields, fig. 38.)

Ancient Field Group (29), area (iii).

(iv) Handfast Point, W. and S. of (050820–055825). On ground mostly falling gently to the N., 'Celtic' fields, some crossed by strip fields, lay over about 34 acres between 60 ft. and 160 ft. above O.D. and formerly continued in all directions. The N.E. of the area was known as 'Cliff Fields' on the Tithe Map. Clay-with-Flints and sand lie patchily on the chalk here. Lynchets are up to 4 ft. high. The inner sides of 'Celtic' fields lie on the cliff edge in the area of St. Lucas Leap. (R.A.F. V.A.P. CPE/UK 1893: 3305; 4300–2.)

Group (30): Roke Down and Bere Down West (SY 89 NW; 813964–831974), Bere Regis. 'Celtic' fields, mostly broken and in parts overlain by strip ploughing, cover 220 acres between 160 ft. and 320 ft. above O.D. on the sides and floor of a shallow valley running S.E.N.W. The subsoil is chalk with some clay and pebble capping. Lynchets are generally very well developed but only survive clearly at the N.E. of the group and near a probable settlement (a) at the S.W. Fields have been carefully fitted around a long barrow, Bere Regis (66), and a number of round barrows, all at or near field angles. The parish boundary which forms the W. limit of the group generally lies on a bank (see Dykes, Bere Regis (119)), which is interrupted by the track W. from (a). 160 yds. S.E. of Roke Barn, about 821963, is a low banked circle about 18 ft. across, without entrance, and probably modern.

(i) Roke Down. S.E. of the former Roke Barn, lynchets are in places up to 21 ft. high, though most are levelled. Strip ploughing is apparent and immediately S.E. of Roke Barn is marked by ridge-and-furrow 15 yds. to 17 yds. broad and up to 1 ft. high. At 81559625 is a probable settlement area, (a), of about 1½ acres, on the sheltered N. brow of an E.-W. ridge with 'Celtic' fields above and below. It consists of an almost flat strip of ground 50 ft. to 100 ft. wide with occasional low and poorly defined platforms along it, which stretches for 750 ft. on the S. side of a scarped track, now partly hollowed, up to 21 ft. wide. This track terminates at the E. end of the supposed settlement where there are two ill-marked curved platforms set into the hillside below another platform measuring about 70 ft. by 24 ft.

(ii) Bere Down. At the W. end of the Down, on ground sloping gently S.W., lynchets are frequently about 4 ft. high, though all ploughed over, and no complete fields survive. Long barrow (66), the ditches of which cannot be seen, appears to have been used as a 'Celtic' field side (R.A.F. V.A.P. CPE/UK 1934: 4117–9.)

Group (31): Elderton Clump (SY 89 NW; 843974–846973), Bere Regis and Winterborne Kingston. (fn. 6) 'Celtic' fields, now almost destroyed, occur over 12 acres between 250 ft. and 350 ft. above O.D., S. and W. of the Clump, and formerly extended in all directions. The chalk here is capped with clay, pebbles and a deep loamy soil. There is a probable settlement in the area (cf. Hutchins 1, 147) but the only earthwork to suggest it is a fragment of double-lynchet track 10 ft. wide, which runs into Elderton Clump from the direction of round barrow, Bere Regis (110), S.W. of it, at a field angle. Coarse pottery in the Iron Age 'A' tradition and some Romano-British sherds, including New Forest ware, were found with iron slag, burnt clay and daub in the ploughed scarp, 10 ft. high, at the foot of the slope 130 yds. W. of the clump. (R.A.F. V.A.P. CPE/UK 1934: 2121–2.)

Group (32): Bere Down (SY 89 NW; 845969), Bere Regis. Remains of 'Celtic' fields, with lynchets up to 12 ft. high, occur over 20 acres between 200 ft. and 300 ft. above O.D., N. and S. of a small dry combe where the Roman road from Badbury Rings to Dorchester probably cut across them. They formerly extended in all directions and almost certainly joined Group (31). A lynchet lies against round barrow Bere Regis (112). A 'settlement' marked on some maps immediately S. of the Roman road at 84559684 is a quarry area. (R.A.F. V.A.P. CPE/UK 1934: 4122.)

Group (33): Bloxworth Down (SY 89 NE; 875958–879962), Bloxworth. Remains of 'Celtic' fields with lynchets up to 5 ft. high cover 50 acres between 150 ft. and 240 ft. above O.D., mostly on moderate W. and N. slopes. A probable double-lynchet track is traceable for ¼ mile running N. from under a 'Celtic' field angle at 87589580 to just W. of round barrow, Bloxworth (36) (cf. Hutchins 1, 184). (R.A.F. V.A.P. CPE/UK 1934: 4127.)

Distribution Maps

Distribution maps of Barrows and of Iron Age and Roman Sites in South-east Dorset, to a uniform scale of ½ in. to 1 m., will be found opposite. A third map in the series, showing Mediaeval Settlements, is included with the Sectional Preface in Part 1 (opp. p. xxxvi).

Note. The only sites marked in the areas shown on the maps peripheral to S.E. Dorset as defined in this Inventory are those which cannot be dissociated from monuments included in Dorset II.

Iron Age and Roman Sites in South-East Dorset

Barrows in South-East Dorset


  • 1. For a fuller study of both 'Celtic' fields and the forms of strip ploughing that followed see H. C. Bowen, Ancient Fields (1961).
  • 2. Evidence was discovered in 1967 for Romano-British settlement on 'Celtic' fields, under strip lynchets in Poxwell at SY 736840 (Dorset Procs., forthcoming).
  • 3. Frampton and Compton Valence are in the area covered by Dorset I.
  • 4. Bradford Peverell is in the area covered by Dorset I.
  • 5. Little Bredy is in the area covered by Dorset I.
  • 6. Winterborne Kingston is in the area to be covered in Dorset III.