Earthworks: Long and Round Barrows

EARTHWORKS

LONG and ROUND BARROWS

The barrows in the Inventory are numbered individually within each parish, generally in the order S.W. to N.E. Each barrow group is described under that parish in which the predominating number of barrows occurs, those in adjacent parishes being crossreferenced thereto; the different parishes are indicated by subheadings in the group entry. All the groups, except those on the Ridgeway, are given reference letters (A, B, AA, AB, etc.) in the order that they are described under parishes in the Inventory; R is reserved for the Ridgeway Group (see p. 425), its component groups (R. 1–14) being numbered from W. to E. In addition, groups are given names of local derivation wherever possible. A register of barrow groups is given on pp. 429–30. The titles of round barrows described individually are printed in italics, those of barrows under group headings in ordinary type.

The position of each barrow is given exactly by a grid reference and generally by a topographical description. Relative distances are therefore only given between and within groups and clusters. Measurements between round barrows are from centre to centre, unless they are actually or almost touching when the phrase 'immediately adjacent' is used.

Since most barrows have been damaged, their character and even their identification as barrows may be in doubt. A question-mark is used to mean 'probable' or 'probably': thus 'Barrow (?)' is a mound which is more likely to be a barrow than anything else, as distinct from 'Mound, possible barrow'; a question-mark after a type description means that the doubt is only about the type, e.g. 'Bell (?)'. 'Ploughed' implies that the mound has been spread and is therefore lower and of greater diameter than it was formerly. Unless otherwise indicated, the diameter given is that of the mound only; dimensions of other barrow components are also quoted where possible. Heights of mounds on a slope are given as a mean measurement.

Most of the barrows have already been numbered in L. V. Grinsell, Dorset Barrows (1959). Correlations with this and other systems of numbering, principally in C. Warne, Celtic Tumuli of Dorset (1866), and the E. Cunnington MS. (c. 1890) in D.C.M., are given below in concordances (long barrows, p. 433; round barrows, p. 474). References to other main sources occur at the end of individual barrow descriptions and, if it is uncertain exactly which barrows are involved, in the introductions under the parish headings. Omission from this Inventory of examples shown as 'Tumuli' on Ordnance Survey maps or listed in Dorset Barrows means that evidence exists that these features are not barrows; some of them are included under Mounds (see below, p. 480).

In the statistics quoted below, all numbered barrows, certain, probable and possible, are included. Multiple barrows (see p. 422) are treated as single monuments.

Long Barrows: Introduction

There are nine, possibly twelve, long barrows in the Dorset II area, all presumably burial mounds of the Neolithic period. They comprise: five earthen long barrows and three more mounds which are probably earthen long barrows; one bank barrow; the Maiden Castle 'long mound' (probably best regarded as a bank barrow); and one chambered long barrow and a second, doubtful, example. Three other long barrows, already described in Dorset I, are considered in relation to the huge group of barrows on the Ridgeway (see p. 425); these are the bank barrow Long Bredy (8), and the 'bank barrows' Kingston Russell (6d) and (6i), now reclassified as long barrows. In making this reclassification we are guided by the definition of a bank barrow which requires a length greater than that of the normal long barrow, parallel sides, and parallel side ditches which do not return round the ends. (fn. 1)

All the long mounds are situated on chalk, except for Portesham (33), which is on limestone. Their siting varies from 200 ft. above O.D. at Bere Regis (66) to just over 600 ft. at Corfe Castle (181), and most appear to have been deliberately placed just off the highest point in the locality. From a distance therefore they appear in silhouette only from certain directions, though modern features such as hedges often make them less prominent. The bank barrows particularly seem carefully placed for visibility from or towards certain points. It must always be borne in mind, however, that in most cases we can never know for certain how visibility might have been affected by vegetation or artificial structures in the prehistoric period.

Most of the mounds are between 100 ft. and 300 ft. long, the four shorter examples being damaged or doubtful; the bank barrow Broadmayne (19), however, attains 600 ft. and the Maiden Castle 'long mound', Winterborne St. Martin (23), 1790 ft. They vary in width from 40 ft., Corfe Castle (181), to 88 ft., Winterbourne Steepleton (13), the latter being the only true wedge-shaped example. In height they range from 1¼ ft., Winterborne Monkton (4), to 9 ft., Church Knowle (34), but the former has been much ploughed and the latter is an unusual, oval type. The majority are between 4 ft. and 7 ft. high and some are higher at one end, always the eastern. In cross-profile they vary from an almost triangular steeply ridged outline at Bincombe (12) to the flat-topped, steep-sided shape of the Broadmayne bank barrow, but ploughing and other destructive activities have often altered the original profile.

Only the 'Hell Stone', Portesham (33), is a certain chambered long barrow, though the existing chamber is a 19th-century reconstruction. With its chamber at the E. end and traces of a peristalith it resembles 'The Grey Mare and her Colts', Long Bredy ((15) in Dorset I), less than 2 miles away to the west. Small loose sarsen blocks in the disturbed portions of Bere Regis (66) perhaps indicate that this barrow is related to those with chambers. The stones Portesham (59) and Winterbourne Steepleton (65) cannot be regarded as the undisputed remains of chambered long barrows (see Stones below).

With two exceptions, Winterborne Monkton (3) and (4), the long barrows are aligned within 45° of an E.-W. line. The significance of this is uncertain, but it seems probable that in some instances they are directed towards local features rather than towards the sun, moon or stars. The long barrow Winterborne Monkton (3), for example, points exactly towards the W. end of the Maiden Castle 'long mound' which is itself apparently placed to be visible from the area of the long barrow. The 'long mound' was built across the ditch of the causewayed camp (see under Hill-forts, Winterborne St. Martin (142)), and nine other long barrows are within 6 miles of this monument. This placing of barrows in relation to one another or to other earthworks suggests deliberate siting.

Round barrows, generally later and sometimes much later than the long barrows, were often deliberately placed near them. A round barrow, Broadmayne (20), actually lies over the W. end of the bank barrow, as does another probable barrow over the eastern end of the long barrow Whitcombe (5) (cf. also Conquer Barrow, West Stafford (22), built on the bank of a henge monument). The group of round barrows on Ailwood Down is clustered round the long barrow Corfe Castle (181), and in some other groups the lines of round barrows prolong the axes of long barrows, notably the Culliford Tree Group (Plate 209). Visibility from a distance may however have been as important as close proximity and many examples are provided in the remarkable concentration of barrows along and near the South Dorset 'ridgeway' (Ridgeway Barrow Group, 'R', see pp. 425–9).

The long mounds, Bere Regis (66), Winterbourne Steepleton (13) and, probably, Portesham (33) have been incorporated in 'Celtic' field patterns, the first two at least apparently forming field boundaries. Around Bincombe (12) the negative lynchet on its N. side also suggests a field lay-out earlier than the present one.

Round Barrows: Introduction

The round barrows in South-east Dorset have been the subject of much unsystematic study, a situation largely corrected by the publication of L. V. Grinsell's Dorset Barrows (1959), which drew in part on the material then unpublished in the Commission's files. Amongst the round barrows listed below are a few not noted by Grinsell under the parishes included in this Volume, and there are differences and additions concerning details such as grid references, dimensions, descriptive observations, earthwork relationships, and interpretation of the evidence from the many particularly ill-recorded barrow excavations (see pp. 425, 428). But these differences of detail have not led to any radical revision of the generalisations and lists provided in the first part of Dorset Barrows, though recent studies elsewhere of the Middle and Late Bronze Age in general and the so-called 'Deverel-Rimbury culture' in particular have invalidated much of the earlier work collated by Grinsell which sought to explain and date 'post-Wessex culture' Bronze Age material largely on the basis of pottery typology. (fn. 2)

Round barrows are predominantly the burial mounds of the Bronze Age, though they were occasionally used, and in a few cases actually built, in the following Early Iron Age, Romano-British and Pagan Saxon times. It can, however, be taken that in the area covered by this Volume all the barrows described, with but a few possible exceptions, were built during the second millennium and first half of the first millennium B.C. Barrows covering primary burials accompanied by bell beakers and long-necked beakers were probably built in the first three centuries after c. 2000 B.C., but the great majority probably originated between c. 1700 B.C. and c. 1000 B.C.

The barrow typology on which the descriptions are based is that now generally accepted. (fn. 3) A bowl barrow is simply a round mound with or without a ditch (here specified where recognisable) immediately surrounding it. Only one certain example (Winterbourne Steepleton (20)) has a bank outside the ditch. A bell barrow is always surrounded by a ditch, separated from the foot of the mound by a flat or sloping ledge called a berm. A disc barrow consists of the same elements—mound, berm and ditch—but the mound is small and the berm relatively wide. In addition it is surrounded by a bank, almost invariably outside the ditch. A special type of disc barrow, virtually confined to S.E. Dorset and called by Grinsell the 'Dorset type', has a second ditch outside the bank. It has been suggested that this type takes the place of the saucer barrow—a low, broad mound with a ditch and outer bank—of which there are no certain examples in the area. The fifth main type is the pond barrow which consists superficially of a circular depression surrounded by a bank, sometimes with a single gap; such monuments were perhaps not primarily burial places. Bell, disc and possibly pond barrows are associated in particular with the Wessex culture of the Early Bronze Age. There are only a few exceptions to this typology, which by its very comprehensiveness isolates the exceptional. Thus, while the term 'bell-disc barrow' is cumbersome, the fact of its use indicates the unusual nature of the barrow and stresses the presence of features of both bell and disc barrows, producing proportions uncharacteristic of either.

Oval barrows are occasionally found. Sometimes they are clearly related to the true long barrow by reason of their ditch arrangements (e.g. Church Knowle (34)), but others are on account of their size and detail dealt with here as round barrows (e.g. Winterborne St. Martin (28)). Three barrows in square enclosures (Winterbourne Steepleton (24–6)) are seemingly Iron Age or later.

Multiple barrows are also rare. They consist of two or more immediately adjacent mounds apparently forming a single structure, normally surrounded by the same ditch. Double, or twin, bowls are most common (seven examples); Portesham (51) and Winterborne St. Martin (37), both ditchless but with mounds joined by a slight bank, and Tyneham (30), with a common ditch between the mounds, should probably be considered with them. There are four triple bowls, two (Bincombe (44) and West Lulworth (35)) apparently ditchless, and two (Winterbourne Abbas (22) and (24)) with a ditch along one side only. In the same Group (AD), Winterbourne Abbas (26–7) perhaps form a double bell, and Group AJ contains an apparently unique quadruple bell (Winterborne St. Martin (91)). In several other cases (e.g. Winterborne Monkton (9–10)), barrows are conjoined but cut or overlap each other, suggesting they are probably successive and therefore not strictly multiple as defined above.

Although the area covered by this Volume is small it nevertheless contains more barrows than the whole of Somerset or of Gloucestershire and Berkshire added together. (fn. 4) It is of primary importance for any study of barrows as field monuments, and the results of several excavations, ill-recorded though they may be, have a direct bearing on the nature and chronology of the Bronze Age in Southern England. Our aim in this Volume has been especially to consider the siting and distribution of barrows, particularly in groups. A group is defined here as four or more barrows related to one another by proximity, situation or common relationship to some other feature. Three main types of group are distinguished, called respectively compact, linear and scattered. These correspond to the 'nuclear', 'linear' and 'dispersed' types of group recently discussed elsewhere, (fn. 5) though 'nuclear' is there used to indicate a particular type of compact group.

Both compact and linear types of group can be sub-divided. Compact groups, i.e. groups of barrows close together, occasionally appear to be related to one specific focal barrow or 'nucleus', (fn. 6) and have been called 'nuclear' groups. The linear type is sub-divided by the distinction between straight and irregular lines, most straight lines being short whereas irregular lines tend to be strung out over a longer distance. The scattered type is simply a rather loose concentration of barrows. Any group may include elements of group-types other than that under which it is classified. Occasionally other, apparently associated, earthworks occur in group areas perhaps to serve as ritual or mortuary enclosures (see Enclosures, Winterborne Monkton (11), Winterborne Came (49), and Whitcombe (25)).

Particular interest is given to the area covered in this Volume by the consideration of the South Dorset Ridgeway Barrow Group as an entity, consisting of fourteen component groups (R. 1 to 14), with 'satellite' groups, together forming one of the most marked concentrations of round barrows in the British Isles. (fn. 7) Most of the large groups occur on and near the Ridgeway but there are numerous small groups, particularly of linear type, in the rest of the area.

872 round barrows, a few only tentatively identified, are listed in the following Inventory: 195 form the Ridgeway Group together with 38 barrows already listed in Dorset I (the W. end of the Ridgeway Group lies within the parishes of Long Bredy, Kingston Russell and Little Bredy); a further 205 barrows, mostly in groups, lie within the area covered by the Ridgeway map (in pocket), nearly all on spurs projecting northwards from the Ridgeway itself. The concentration of barrows on the Ridgeway and its spurs is emphasised by the fact that half of the barrows listed in the Inventory are grouped there within an area only one-fifth of that covered by the Volume. Allowing for a few unrecognised and some unlocated and destroyed barrows, the total number in the Volume is about half that for the whole of Dorset—some 1800 as estimated by Grinsell (fn. 8) —though the area covered is only one quarter that of the county. Even within the Wessex region, therefore, archaeologically characterised by its large number of Bronze Age funerary monuments, this relatively small area in South Dorset was clearly of special importance.

Outside the Ridgeway area there is no comparable concentration within the Volume boundary, though most of the barrows fall within definitely localised scatters (Fig. opp. p. 634). These occur in five main areas: first on the high chalkland, much of it covered by Chaldon Down, between Poxwell and West Lulworth, with the 'Five Marys' (Chaldon Herring (51–6)) on the northern edge overlooking the heathland dropping N. to the river Frome; secondly on the low ridges to the N. of the river Piddle, spanning the junction of the heath and chalk and continuing N. over the Volume boundary; and thirdly on the heath on the northern side of the watershed between the Poole Harbour basin and the valley of the river Stour. The fourth area, with perhaps the densest scatter, is on the heaths to the W., S.W. and S. of Wareham; though this land is low-lying, most of the barrows are sited on the local ridges between the streams and on the small knolls rising slightly out of the heath. The fifth is on the high ground, partly down and partly heath, forming the ridge between Swanage and Studland and overlooking Poole Harbour to the N. and the sea to the E. Certain areas, however, were clearly avoided. The most obvious of these is the large stretch of heathland along and near the northern shores of Poole Harbour, an area which has not always been one of heath and which may have been partly cultivated in Bronze Age times. (fn. 9) The other area is S. of the Purbeck Hills where the almost entire absence of barrows is particularly striking when compared with the wealth of settlement and burial material of Iron Age and Romano-British times (see Figs. opp. p. 634).

Round barrows sometimes occur in apparently significant relationships to long barrows, mostly in groups (see above, p. 421; and below, p. 426, for discussion of the relationship in the Ridgeway area). Only the Ailwood Down Group, Corfe Castle, on the Purbeck Hills, demonstrates this outside the Ridgeway area. It can be regarded as a compact group of nuclear type, with the long barrow as the focal monument, a relationship emphasised by the absence of round barrows from the ridge for some distance to W. and E.

The Ailwood Down Group is one of the two largest groups in the Volume outside the Ridgeway area. The other, also a compact nuclear group, is on Bloxworth Down with a large bell-disc barrow apparently as the focal monument. (fn. 10) Otherwise most of the groups are on heathland, and a marked characteristic is the frequency with which the whole or part of these groups is based upon straight alignments of barrows. The Five Barrow Hill Group (Tyneham) and the Seven Barrows Group (Wareham St. Martin) are particularly good examples which also demonstrate how advantage was taken of a local ridge in the heathland to place barrows on sites in low-lying terrain so that the mounds are as clearly visible as those on the more obvious skylines of the chalk downs. The use of these low ridges naturally contributes to the linear nature of some of these heathland groups, but the geometric arrangements are apparently deliberate, since other dispositions could have been made. In addition to these short linear cemeteries, there are a few irregular, spaced-out, linear groups, like the Corfe Common Group (Fig. p. 97), as well as a few compact groups; the Rose Lawn Group, Poole, is of this last type. Particularly on the heath, barrows outside groups seldom occur singly, being normally in pairs or in clusters or alignments of three. Really low-lying and isolated barrows, like that near Nottington (Weymouth (434)), are extremely rare.

There is a marked discrepancy in the numbers of special or 'fancy' barrows within and outside the Ridgeway area. Whereas 76 special barrows of all types—bell, disc, pond and multiple bowls—occur in the Ridgeway area (29 of them actually in the Ridgeway Group), there are only 33 others in the rest of the area covered by the Volume, and all of those are bell barrows except for two double bowls (Tyneham (30) and (37)), a possible triple bowl (West Lulworth (35)), and the 'bell-disc' barrow on Bloxworth Down (Bloxworth (39)). The relative proportion of special barrows to others in the Ridgeway area is about 1 in 6 whereas outside this area it is 1 in 14.

The significance of these special barrows is indicated by some further figures. On average, their overall diameter is nearly twice that of bowl barrows, both ditched and un-ditched. The average overall diameter of a bowl barrow within the strict limits of this Volume is 54 ft. and that of a special barrow is just over 100 ft. The difference is further emphasised by the fact that whereas the largest number of special barrows have diameters of about 100 ft., the majority of the bowl barrows fall within a bracket of 30 ft. to 50 ft. A further quarter have diameters of 30 ft. and less, though some apparently ditched bowls (possibly bells) are over 100 ft. in diameter. The comparison might be made by saying that if all the bowl barrows in Dorset II were put side by side in a straight line, they would stretch for about 8 miles. A similar line of special barrows would be just over 1½ miles long. (fn. 11)

So many of the barrows have been damaged, particularly by excavation in the 19th century, and by ploughing in the last 25 years, that there is little to be learnt now by a study of heights and profiles. Off the chalk, however, and particularly on the Reading Beds, it is noticeable that many of the barrows have short, fairly steep sides and broad flat tops, apparently as original features. Only some 140 barrows of all those in the Volume are apparently undamaged. About 220—a quarter of the total—appear to have been excavated, judging from disturbance in the top of the barrow mounds, and known excavation accounts can be attached to specific barrows totalling about a quarter of this figure. Nine barrows have been excavated by modern methods (Arne (29), Bincombe (25, 27), Poole (363–5), Portesham (38), Weymouth (416) and Winterbourne Steepleton (46)); otherwise excavation results are almost entirely derived from 19th-century diggings, mostly into barrow centres. In a number of cases, the mound has been shown to consist partly of a turf stack, and in others, particularly in the East Lulworth area, large stones have formed a significant part of the barrow or grave structure (cf. Dorchester (169), with a boulder weighing nearly 3 tons, and Bincombe (24) and Poxwell (12) with small internal ring-walls). On the chalk, many of the excavations produced burials in graves, and in one case (Church Knowle (40)) the grave was about 10 ft. deep. In only twelve barrows is it probable that the excavators reached primary burials of Wessex culture date or type, the best-known examples—Clandon (Winterborne St. Martin (134)), Ridgeway 7 (Weymouth (403)), Culliford Tree (Whitcombe (9))—being all secondary deposits. With the exception of 'King's Barrow' (Arne (36)) on Stoborough Heath, all are on the higher ground, and most of them on chalk. It is perhaps instructive that three of these barrows occur in two separate compact groups (Five Marys and Bloxworth Down), which on other grounds could be regarded as probably having developed from and around a 'focal' barrow.

There is some evidence from S.E. Dorset for the use of barrows for burials subsequent to the Bronze Age. In Bloxworth and Corfe Castle parishes, for example, are barrows which contained several extended inhumations most likely to be of Romano-British date and almost certainly later than the probably Bronze Age type of secondary cremations, accompanied or otherwise, often inserted into the tops or sides of barrow mounds.

There is ample evidence to show that round barrows were used as markers or fixed points in the laying out of 'Celtic' fields and that in general they were deliberately preserved by the farmers of these fields. In Bere Regis, for example, three barrows in the Roke Down area are at 'Celtic' field angles (see Ancient Field Group (30)).

The Ridgeway Barrow Group (Group R) (fn. 12)

'For sight of barrows, I believe not to be equalled . . .'

Wm. Stukeley, Itinerarium Curiosum (1724), 163.

'. . . notwithstanding the many changes which have taken place since that time [1724], it is certain that every enquiring spectator must be equally struck with this extraordinary district . . . where the adjacent downs or the lofty Ridgeway with its prolonged upland crest, are gracefully undulated with these time-honoured memorials'.

Chas. Warne, Celtic Tumuli of Dorset (1866), 4.

South Dorset, particularly north of Weymouth, is dominated by a sharply-defined, 12-mile-long ridge at the southern edge of the chalk downs. This ridge, continued further to the E. by the Purbeck Hills, consists of Upper Chalk, overlaid at points by Clay-with-Flints and Bagshot Beds, dramatically truncated to the S. by the so-called Ridgeway Fault and the escarpment to the Jurassic Beds.

Topographically, where facing S., the ridge resembles a huge rampart, its domed top clearly defined and in places very narrow, notably on Bronkham Hill. It is best shown on the map by the 400 ft. contour, although at several points it rises to over 600 ft. and in one section (on Black Down by the Hardy Monument) to over 700 ft. above O.D. The general run of the ridge is from N.W. to S.E., the higher end being on Martin's Down in Long Bredy, from which there is a slight slope down along the top of the ridge towards Osmington in the S.E. This fact is indicated on the map by the 600 ft. ring contours N.W. of Black Down, and their absence to the E.

The rather more gradual rise to the ridge-top from the N. contrasts with the abrupt fall to the S. The chalk uplands have been cut into by small streams, now mostly vanished, resulting in a series of spurs jutting N. from the ridge towards the valley of the South Winterborne, a tributary of the River Frome. The spurs, together creating a rolling landscape, are all slightly lower than the top of the ridge and form a prominent part of the view to the N. from it. Geologically, they are part of the same formation, Upper Chalk. The same is true of the only two spurs S. of the ridge—Bincombe Hill and West Hill.

Most of the ridge-top has at some time been ploughed and only the south-east tip of Bincombe Hill, Came Down, and part of the col to the south of Northdown Barn, are in modern pasture. Bronkham Hill, presumably as a result of its acid soil, has not been ploughed, though it is pitted with solution hollows. Came Wood on the ridge-top, and Big Wood on a spur, are the only sizable plantations. For the rest, the ridge, interrupted only by dry-stone walls and barbed-wire fences, remains a windswept stretch of high ground providing splendid views, both along its own length and over much of S. Dorset and the adjacent English Channel.

Along nine of its twelve miles, between Martin's Down on the W. and the S. tip of Broadmayne to the E., the ridge-top is here regarded as the site of a large barrow group. Of the 233 barrows in the group, all but seven fall naturally into fourteen sub-groups which together make up the Ridgeway Group (Group R). A further 129 barrows in fourteen groups and 76 barrows in clusters or singly, occur, mostly above the 400 ft. contour, on the spurs and slopes related to the ridge-top. Group R includes barrows in the parishes of Long Bredy, Kingston Russell, Little Bredy, Winterbourne Steepleton, Portesham, Winterborne St. Martin, Weymouth, Bincombe, Winterborne Came, Whitcombe, Broadmayne, Poxwell and Osmington. Most of these parishes, plus Winterbourne Abbas, Winterborne Monkton and Winterborne Herringston, also contain barrow groups and barrows related to Group R in ways discussed below. Lack of barrows on the remainder of the ridge suggests that the concentration is due to something more than the existence of any contemporary route that may have run along the top.

The main reason for describing the 233 barrows as part of one huge group is that they constitute a most unusually high concentration clearly associated with the ridge. They are, moreover, contained on one particular stretch between two bank barrows, markedly similar monuments without close parallel in England. Although the ridge itself continues to the S.E. of the Broadmayne bank barrow, apart from the East Hill Group (R. 14) ¾ mile from it, there are only two further barrows before the Poxwell Gap and only a scatter of barrows beyond. It might be suggested that the Broadmayne bank barrow is placed where it is, most delicately sited on the exact crest of the ridge, because it is clearly visible from the unusual spurs to the S. as well as from the lower ground to the N. The Martin's Down bank barrow at the W. end is even more dramatically sited at a natural break in the ridge. It can be suggested that the two bank barrows define the ends of a length of ridge-top which was of significance before it became studded with round barrows: its length suggests comparison with the function of the probably contemporary cursuses. There may in fact be a conceptual connection between the linear aspect of long barrows and bank barrows (apparently a local and abnormal development (fn. 13) ) and of cursuses and more particularly perhaps, between the ridgeway itself and the ridge-like appearance of the bank barrows. Of the six long barrows in Group R, five are towards the ends of the group as defined by the bank barrows. It may be significant that no henge monuments are known on the Ridgeway. The fact that a large number of round barrows was then apparently related to the demarcated stretch of ridge-top is strongly suggestive of continuity between the Neolithic and Bronze Age periods.

The Ridgeway Group is essentially an irregular linear cemetery because of the nature of the topographical feature on which it is sited; and eight of the groups within it are predominantly linear groups. Six groups —e.g. R.7 on Ridge Hill—contain straight alignments and one group just off the ridge-top, that on West Hill (R.12) on a southern spur, consists entirely of a straight line of nine barrows. Other groups are compact, or clustered round a 'focal' barrow—e.g. the greater (western) part of the East Hill Group (R.14) or the Martin's Down Group (R.1). Only two groups—R.4 and R.5—are just a scatter with no obvious coherence. The numbers of barrows in groups within Group R varies between five and thirty-eight. The largest group related to the ridge is that on Winterbourne Abbas Poor Lot (Group AD) with forty-four barrows down in a valley (although still over 400 ft. above O.D.).

All of the groups, and notably the last-mentioned, both on and off the ridge-top, appear to be deliberately sited so that a complex system of intervisibility is created. Clearly the height, and at times the narrow top, of the ridge make it inevitable that some barrows should be landmarks and intervisible. But many individual barrows and barrow groups are so sited in exactly the right place to achieve a striking effect, to appear on the sky-line when viewed from certain points, and to be seen easily from other barrows and barrow groups, that a considerable degree of control and deliberation must be postulated. The great Ridgeway bell-disc barrow (Winterborne St. Martin (67)), for example, though sited off the ridgecrest and in a dip of the ridge-top so that it is not readily visible when looking along the Ridgeway, is strikingly situated on the sky-line when viewed from the long mound in Maiden Castle, towards which it is tilted.

The topographical advantages of the ridge-top are demonstrated by the length of time it has been used as a burial site. The six long barrows show that the ridge was being used in the Neolithic period, and round barrows were built on the ridge-top between the bank barrows and on related sites by Beaker people. Most of the round barrows were probably erected during the Early and Middle Bronze Ages, but burials continued to be made here intermittently for another thousand years, at least until the Iron Age A burials on Ridgeway Hill (R.8) (Bincombe (24)). The ridge's most recent use for the commemoration of a dead personage is represented by the Hardy Monument (Portesham (3)) on its highest point.

There are only two other large areas of comparable barrow density in England and these are both in Wiltshire, around Avebury and around Stonehenge. In the 9 square miles around Avebury, there are about ten barrows per square mile, and in the 12 square miles around Stonehenge, the figure is about twenty-five. Even the whole area shown on the Ridgeway map—some 45 square miles—has an average density of ten barrows per square mile, while an area 1 mile wide based on the Ridgeway and the nearby groups to the N. gives a density of about forty barrows per square mile. Alternatively, the concentration can be emphasised by taking, for example, the arbitrary area within the parish boundary of Winterborne St. Martin which still gives a figure of about twenty-four barrows per square mile. In utter contrast, the Berkshire Downs have two barrows per square mile and the 14 mile long chalk ridge between the Rivers Nadder and Ebble in S. Wiltshire one to two barrows per square mile. (fn. 14)

It might be thought that such a relatively high number of barrows in a limited area involved large numbers of people. If, however, most of the barrows were built in the 2nd millennium B.C., as seems likely, then on average only one was built on the Ridgeway every 4 to 5 years. In the whole Ridgeway area, one barrow was built on average every 2 years, and, even if it is accepted that practically all the round barrows were built within the 500 years around 1500 B.C., on average only one barrow was built every year. People were almost certainly attracted from some distance to bury their dead, so even allowing for the probability that only the more important were given barrows and that many barrows held several secondary burials the figures hardly indicate a large population in the Bronze Age.

The main point which these figures suggest is that there was continual sepulchral activity in the Ridgeway area for a very long time, resulting in the relatively coherent pattern we now see. The careful siting within groups of individual barrows, for example, implies direction of labour by persons with a perceptive eye for the lie of the land. This is particularly well illustrated at either end of Group R in the siting both of the bank barrows and of the groups topographically and visually related to them.

It is, however, difficult to trace the development of Group R, partly because there are few definite earthwork relationships affording relative dates, partly because the excavation records are so defective or non-existent (see below).

Six groups on the Ridgeway contain only bowl barrows, which are by far the most frequent type, 205 being recorded. It is likely that excavation would show some of these to be bell barrows, and some of the barrows known only from crop or soil-marks on air photographs, and classified as bowl barrows for lack of other evidence, may also have been bell barrows. The same qualifications must be made for the figure of 153 bowl barrows in the associated area. Such qualifications must not, however, detract from the validity of the total number of barrows in and associated with Group R.

Bell barrows are the next most frequent type, although small in number compared with bowl barrows. There are seventeen in Group R and seventeen (three of which are doubtful) on associated spurs. They occur in only six of the Ridgeway groups, being most numerous on Bronkham Hill (R.6), where there are four in a group of thirty, and on Ridge Hill (R.7) where there are five in a group of thirty-eight. Three 'bell-disc' barrows also occur, two small examples (Winterbourne Abbas (14–15)) in the Poor Lot Group (AD), and the other, with the largest diameter of any in Group R, at the eastern end of the Ridge Hill Group. This barrow (Winterborne St. Martin (67)) is, however, quite exceptional and it is probably misleading to classify it.

Disc and pond barrows also occur, though infrequently, in Group R and on associated spurs. Both disc barrows in Group R are on Black Down (R.2), while of the possible eleven such barrows related to Group R (two on the Bincombe-Winterborne Herringston boundary being doubtful), six occur in the Poor Lot Group and another, a well-preserved example, with an outer but no inner ditch, is in the Came Down Group (AG). Pond barrows are slightly more numerous, there being eighteen altogether: seven in Group R, ten in related groups and one, Winterbourne Abbas (32), in the valley bottom near the Broad Stone. Five in Group R are in the Culliford Tree Group (R.11), and there are also five in the Poor Lot Group. All but two of the others occur singly in different groups. The distribution of both disc and pond barrows shows them to be near either end of Group R, which fact perhaps further suggests that the limits to the group were recognised and thought to be of special significance. Certainly pond barrows seem to be especially related to the Ridgeway Group: there are no other certain examples in the whole of Dorset.

Although about 100 barrows in and related to Group R have been dug into, only seven (three of them pond barrows) (fn. 15) have been excavated by modern methods, and only two of those fully published. The value of the bulk of the excavated evidence is limited. Many of the barrows excavated in the 19th century for which excavation records exist have now been identified on the ground, but there remain on the one hand accounts which cannot be related to any one barrow, and on the other barrows which have clearly been dug into but for which there are no recognised records. (fn. 16) The latter are noted in the following Inventory, while the former are included in Grinsell's lists and used, for example, in the analysis of barrow structure (Dorset Barrows, 46–9), the details of which are not repeated here.

In only a few barrows can it be certain that the primary burial was excavated, and in the majority of excavations it is probable that secondary burials were missed. However, even though dug into by the central hole or trench methods, many excavated barrows have produced more than one secondary burial, and it is quite clear, despite the inadequately recorded evidence, that most of the Ridgeway and related barrows were used many times and over a long period. On the other hand, it is difficult to be certain about the sequence in any given barrow, since stratigraphy was seldom noted, and it is frequently not clear whether the lowest burial found was in fact the primary one. Further, the discovery of burials in the unexcavated, often greater, part of the barrow might alter the interpretation of the excavated evidence at present available. However, Group R and related groups contain some rich deposits, most notable being those from two barrows in Winterborne St. Martin ((134), the 'Clandon' barrow, and (82) in the Eweleaze Barn Group (AJ)); from Weymouth (403), better known as Cunnington's 'Ridgeway 7'; and from Whitcombe (9), the 'Culliford Tree' barrow. The finds as a whole show the group to have formed during the Beaker period and Early Bronze Age, and to have developed, probably nearly to its full extent, during the Middle Bronze Age.

There is very little evidence for contemporary settlement associated with or adjacent to Group R and its related groups. (fn. 17) Earthwork associations occur in several instances between barrows and 'Celtic' fields. The triple bowl barrow on the N. of the Poor Lot Group, Winterbourne Abbas (24), is perhaps on top of a 'Celtic' field lynchet; and, within the 'Celtic' field immediately to the E., a small mound (Winterbourne Abbas (25)) is possibly a ploughed-out barrow. 'Celtic' fields probably also impinge on the S.W. edge of the same group. Barrows Winterborne St. Martin (79–80), N. of the W. end of the Ridge Hill Group, stand among ancient fields (Ancient Field Group (6)); and 'Celtic' field lynchets butt against the southern barrows on Bincombe Hill (Ancient Field Group (8)), and against two of the barrows, still preserved in pasture, in the centre of the Northdown Barn Group (Weymouth (419–20), Ancient Field Group (10)). Numerous barrows, particularly high ones, were used as fixed points for parish boundaries which crossed them, though the continuous line of parish boundaries running E. from the Hardy Monument to Broadmayne (broken now by the reconstituted Bincombe) runs immediately S. of most of the barrows, ignoring them. On the other hand, an unusual small bank and ditch, possibly a boundary line, cuts across the top of barrow Winterborne St. Martin (47) towards the E. end of Bronkham Hill (Fig. p. 519, Plate 209).

REGISTER OF BARROW GROUPS (fn. 18)

Abbreviations used in Tables I and II:

Group types (see p. 423) C=compact

FC = compact with focal monument

L = linear, geometric

IL = linear, irregular

S = scattered

Uncommon barrow types db = double bowl

tb = triple bowl

qb = quadruple bell

bd = bell-disc

bb = bank barrow

Table I: Barrow Groups in Group R

Table II: Barrow Groups other than Group R