The date of the Greater Stonehenge Cursus.

Thomas, Julian ; Marshall, Peter ; Pearson, Mike Parker 等

Introduction

The Greater Stonehenge Cursus was first identified in 1723 by William Stukeley, who famously supposed it to have been a Roman chariot-racing track (Stukeley 1740: 41). As well as the first cursus monument to have been recognised, it is also one of the largest. At nearly 3km in length, it is only eclipsed by the Dorset and Stanwell cursuses, and it remains the largest prehistoric structure in the Stonehenge and Avebury World Heritage Site (Figure 1). Despite this, it has never been satisfactorily dated, and consequentially its place in the development of the Stonehenge landscape has been obscure. While pit and post cursuses in the Scottish lowlands probably date to the centuries between 4000 and 3600 BC, cursus monuments defined by banks and ditches are more likely to have been constructed between 3600 and 3000 BC (Barclay & Bayliss 1999: 29; Thomas 2006). However, the only existing radiocarbon determination from the Greater Stonehenge Cursus, taken on deer antler recovered from the ditch in 1947 by J.F.S. Stone, falls in the earlier third millennium cal BC (see below). This might either mean that the monument is very late for its kind, or that the date is not primary, and relates to intrusive material.

In this paper members of the Stonehenge Riverside Project describe the new (2007) investigations and argue for a revised date and context for the monument.

[FIGURE 1 OMITTED]

Shape and setting

Morphologically, the Greater Cursus is unusual in that its side ditches do not run parallel with each other, for the southern ditch has a pronounced 'kink' outwards in its western portion, so that the width of the enclosure varies between 100 and 150m along its length (Darvill 2007: 89). The northern ditch is generally straighter, but aerial photographs reveal that at a smaller scale it follows a comparatively uneven and meandering course. The cursus runs roughly east-west linking two areas of higher ground, at Fargo Plantation and the King Barrow Ridge respectively, and dips into Stonehenge Bottom between them. In this respect it conforms to the common pattern amongst cursus monuments of incorporating seasonally wet ground, or even watercourses, into their fabric (Brophy 2000).

At the western end it is notable that the cursus spans the top of the Fargo ridge, and the terminal faces westward towards Airman's Corner and Winterbourne Stoke Down. From the terminal, much of the rest of the cursus is invisible, and Beacon Hill rises up above the near eastern horizon. The Lesser Cursus crests the northern skyline, with the Robin Hood's Ball causewayed enclosure beyond it. Perhaps intentionally, then, the western end of the cursus seems to relate to a quite different set of landscape referents from the rest of the monument. At the eastern end of the cursus the situation is quite different, for the terminal ditch runs parallel with the long barrow Amesbury 42, which occupies the crest of King Barrow Ridge. This long mound was excavated by Thumam, who describes encountering an ox skull in a primary position, but only secondary human burials (Thumam 1869: 180). Further investigations by Julian Richards (1990: 98) demonstrated that the mound had been flanked by two successive sets of side ditches, the later and outermost of which were considerably wider and deeper. The implication of this is that at some point in its history the barrow had been enhanced, to form a massive structure dominating the ridge-top. At this point the cursus terminal falls short of the barrow by around 30m. The character of the bank and ditch has never been investigated archaeologically, but it is possible that the elaboration of the long mound was related to the construction of the cursus, forming a definitive terminal point for the complex as a whole.

Eastwards from the King Barrow Ridge, the axis of the cursus runs through Woodhenge and the Cuckoo Stone. Burl (2006: 92) argues that the latter was originally a massive monolith, standing 5m high, and that the cursus was oriented upon it. Excavations conducted by Colin Richards in 2007 indicate that this is unlikely, for the Cuckoo Stone appears to have always been relatively diminutive, and to have been quarried from a pit adjacent to its present resting place. However, a Neolithic pit (containing Earlier Neolithic pottery) in its immediate vicinity demonstrates that the stone was a significant cultural landmark around the time that the cursus was constructed, as Burl has recently speculated (2006: 85, Figure 15). Joshua Pollard's 2006 excavations at Woodhenge (also on the axis of the cursus) revealed an Early Neolithic feature in the form of a tree-throw hole filled with Early Neolithic Carinated Bowl pottery. It is therefore arguable that the east-west alignment of which the cursus forms a portion was established as significant early in the Neolithic period.

Stukeley's 1740 drawing of the cursus shows a further ditch and bank cutting off the western end of the enclosure, and this feature was also recognised by Colt Hoare (1810: 159). This cross-ditch is still recognisable today, running at an angle roughly NNW-SSE between the northern and southern ditches. A bank had originally stood to the west of the ditch, so that the western end of the cursus effectively represented a distinct D-shaped enclosure. In this respect it bears comparison with the south-eastern end of the Dorchester-on-Thames cursus, where the enclosure was evidently a primary feature onto which the cursus was appended (Bradley & Chambers 1988; Whittle et. al. 1992). However, it has been suggested that the cross-ditch may have cut the cursus bank, and was therefore a later feature (RCHM 1979: 14).

Previous investigations of the cursus

The first archaeological intervention at the Greater Cursus was undertaken by Farrer (1917), who dug a cutting across the northern bank and ditch toward the eastern end. He described a flat-bottomed ditch cut, and a bank composed of turf with little chalk content. Interestingly, he pointed to the comparatively limited quantity of chalk rubble in the ditch, and argued that it may have been deliberately kept clean until it was backfilled in a single episode. Further excavations on the southern ditch, a little to the east of Fargo Plantation, were conducted by Stone (1947). His small cutting revealed a causeway, constituted by one rounded and one square ditch terminal, too narrow to represent an entrance. These circumstances strongly suggest that the causeway represented the meeting between sections of the cursus ditch that had been dug by separate work gangs (Richards 1990: 93). Stone's ditch section was puzzling, for it portrayed only a very little chalk silt on either side of the base, and a homogeneous reddish soil containing little or no chalk filling much of the rest of the profile (Stone 1947: 14). He conjectured that the chalk of the bank may not have returned quickly into the ditch as it had been contained within a turf revetment. Following Farrer, he also pointed to the lack of any turfline within the ditch fill, implying that a deliberate backfilling had taken place.

Stone's excavations were especially notable for the recovery of flakes of both sarsen and Stonehenge bluestone from the cursus ditch. The latter has recently been identified as a fragment of Palaeozoic litharenite, that is, a sandstone similar but not identical to the Altar Stone. Neither of these stones derive from the more recent Devonian strata at Milford Haven and are likely to come from the Brecon Beacons (Rob Ixer & Richard Thomas pers. comm.; Ixer & Turner 2006). The upper part of the ditch in Stone's trench also produced fragments of Later Bronze Age pottery. On the southern side of the ditch, Stone identified an 'embayment' or recess, cut back into the ditch edge, which contained an antler crown with two tines. It was this that later provided a radiocarbon date of 2890-2460 cal BC (OxA-1403; 4000 [+ or -] 120 BP). Richards (1990: 96) pointed out that the feature was probably intrusive, and that the radiocarbon date was likely to be considerably later than the construction of the cursus. However, the Later Neolithic attribution has continued to exercise some influence in the literature. Finally, Stone identified two small flint knapping clusters on the ditch base.

More extensive excavations at the western end of the cursus were carried out by Patricia Christie in 1959 (Christie 1963). At the terminal, Christie observed that the ditch was considerably deeper than along the sides of the cursus, providing quarried material for a much more substantial bank (ibid.: 370). As she noted, the massive terminal bank would have resembled a long barrow, a feature that is echoed by the Thickthorn Terminal of the Dorset Cursus, where the cursus bank has two long barrows aligned upon it, forming a continuous line of mounds (Barrett et al. 1991: 50). Passing beside these structures, it is difficult to tell which is bank and which is barrow. Furthermore, when Cannon Greenwell excavated the terminal of Rudston Cursus A, he fully believed he was digging a long barrow (Greenwell 1877: 253-7), so the affinity between the two kinds of monuments may have been quite intentional.

In the terminal ditch, Christie found a primary filling of coarse chalk rubble and fine rain-washed silt, with dark grey bands which she identified as collapses of turf from the bank. Above this was a fine grey silt, which she considered to have been a windblown deposit (Christie 1963: 372). The primary rubble at the terminal contained a well-defined flint flaking deposit, but similar deposits were absent from the side ditches. The technology of this assemblage has been the subject of some debate: the flakes do not suggest parallel blades struck from a prepared core, but they appear rather fine for a Later Neolithic flake-based assemblage (Richards 1990: 96; Darvill 2007: 87). Christie (1963: 374) noted the marked differences in the character of the filling of the terminal and side ditches. Her explanation was that they had probably been dug as entirely separate episodes.

Following the felling of a portion of Fargo Plantation in 1983, Julian Richards was able to excavate two small trenches in the side ditches of the cursus. The first of these, W56A, was on the southern ditch, immediately inside the eastern part of Fargo Plantation (Richards 1990: 93). This was less than 100m from Stone's cutting, and the results were very similar. The basal chalky fill, context 17, was very limited in extent, and was succeeded by context 16, a red-brown decalcified fill, similar to that described by Stone. The interstice between the two layers on Richards' section appears to be marked by a V-shaped arrangement of stones, which may represent a cut, truncating context 17 (Richards 1990: Figure 62). Context 13, a run of stones across the ditch, suggests that a stabilisation may have occurred between the deposition of context 16 and the similar material of context 11. The second trench, W56B, cut the southern ditch adjacent to the Larkhill byway (roughly in the middle of the cursus). Here, the ditch fills were less decalcified, and the section lacked the indication of a V-shaped recut (Richards 1990: 95). Evidently some significant change overcame the ditch deposits as the cursus descended into Stonehenge Bottom, but it was not clear whether this was gradual or abrupt in character.

Two round barrows of Early Bronze Age date are enclosed within the western terminal of the cursus. The easternmost of the two, Amesbury 56, contained a Beaker inhumation and a child burial, as well as a cist covering a further inhumation with a knife-dagger (Stone 1947: 9; Grinsell 1978: 26). The other mound, Winterbourne Stoke G.30 was excavated by Colt Hoare, and contained a cremation deposit with a central cremation pit, which had been discoloured by heat. Re-excavation by Christie revealed another small pit, which pre-dated the barrow and produced a quantity of pine charcoal (Christie 1963: 377). This suggests (but does not prove) a Mesolithic date for the pit, and demonstrates the possibility that the cursus had been constructed in an area that had a long history of use and significance.

The 2007 excavations

New fieldwork at the Greater Cursus was conducted in August and September of 2007 in pursuit of one of the principal research objectives of the Stonehenge Riverside Project (see Parker Pearson et al. 2006). The stratigraphic positions of both the dated antler and the bluestone fragment in Stone's cutting were ambiguous, and the variation in the filling of the cursus ditch demanded clarification. It was also considered important to examine sections in different parts of the structure simultaneously as a basis for comparison, and to provide the opportunity for samples to be taken for soil micromorphological analysis. The presence of a relatively early pit in association with Winterbourne Stoke G.30 and the discovery of cut features immediately inside the terminals of other cursus monuments (Thomas 2007: 166-97) indicated the importance of investigating areas of the cursus interior. Similarly, the long-standing question of whether some kind of megalithic structure had existed in or around the cursus raised the possibility that features which post-dated as well as pre-dated the monument might be encountered. Finally, the investigation of structural sequence within the cursus was defined as a priority, and in particular the relationship between the cross-ditch (which had never previously been excavated) and the main perimeter ditch was a pressing question.

In order to address these issues, an extensive resistivity and magnetometer survey across the cursus interior was conducted by Neil Linford for English Heritage and by Kate Welham under the aegis of the project, but this revealed few anomalies. Five trenches were then laid out, all at the western end of the cursus, and numbered according to the overall sequence for the project as a whole (see Figure 2). Trenches 36 and 38 were positioned to test two of the more promising geophysical anomalies, but the results were largely negative. In combination with the extreme paucity of cultural material from the ditch cuttings, the absence of internal structures suggests that the use of the monument did not involve protracted gathering and food consumption.

Trench 26 was set at the southern end of the terminal ditch, where the geophysical survey had suggested that it entered a gentle curve to meet the side ditch. The unusual shape of the trench was designed to accommodate this curve, and provide perpendicular sections at either end of the cutting. In practice, the ditch ran straight across the trench, indicating that the western terminal was much more rectilinear than anticipated. The ditch sequence was very similar to that described by Christie (Figures 3 and 4). The uppermost part of the profile had been heavily disturbed by rabbit burrowing, bulldozing and military activity on the site, but below this were a series of secondary silts, containing dark brown lenses, which may indicate seasonal variations in the natural depositional process. In the base of the ditch was a thick mass of primary chalk rubble, evidently derived from the ditch edges. This material was angular and varied in size, and was contained within a loose matrix of light yellow-brown chalky silt. Dispersed within the rubble was a group of flint cores, and a series of discrete knapping clusters, each no more than 70cm in diameter. It was notable that these clusters were found throughout the primary chalk rubble, from top to bottom. This suggests that they were not all precisely contemporary, but that they were being generated throughout the period whilst the initial weathering-back of the ditch edge was taking place. Indeed, as this process of weathering would have resulted in the periodic exposure of flint nodules in the ditch edge, it is possible that knapping took place as these came to the surface. At the base of the rubble, on the floor of the ditch, the most significant find of the excavation was discovered. This was a battered frontal tine from a red deer antler pick; it represents an ideally-placed radiocarbon sample to date the digging of the cursus ditch (Figure 5).

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Trench 27 was located over the northern side ditch of the cursus, at the point where geophysical survey results indicated a junction with the cross-ditch--although in practice the latter terminated in a butt-end some 2m short of the cursus ditch. The ditch in Trench 27 was much shallower than that at the terminal, but its filling was broadly comparable. One especially significant aspect of this cutting was the presence of a small intrusive feature (context 030/081) cut through the ditch fill and into the side of the ditch. This was comparable with the 'embayments' in Trench 28 (see below), and indicates that the recutting that these represent may have taken place throughout a large portion of the cursus ditch. The cross-ditch, by contrast, proved to contain a palisade slot, from which considerable quantities of animal bones and Late Bronze Age pottery were recovered. It is not entirely clear whether the cross-ditch was a Bronze Age feature, or an earlier silted ditch into which a Bronze Age palisade had been inserted.

[FIGURE 5 OMITTED]

Trench 28 was designed to test the results of Stone's 1947 cutting, sampling the ditch sequence in this area and attempting to recover further bluestone flakes from the ditch (an objective not achieved). The uppermost filling of the cursus ditch here was a striking orange-brown fine silty clay with no trace of chalk content, containing Beaker period artefacts. It was contained within a V-shaped recut, which descended through the earlier ditch fills to form a furrow in the ditch bottom, strongly signalling its intrusive character (Figure 6). The removal of the recut fill revealed a shallow, bowl-shaped feature (context 050) on the northern side of the ditch, securely stratified between the cutting and silting of the cursus ditch and the v-shaped recut. This feature was similar in location and morphology to the 'embayment' described by Stone, and it is extremely likely that they form elements of a series of contemporary features, also including 030/081, described above.

The radiocarbon dates

Two radiocarbon determinations were acquired from the antler from the terminal ditch (Table 1). The two measurements (OxA-17953 and OxA-17954) are statistically consistent (T'=0.2; v =1; T'(5%)=3.8; Ward & Wilson 1978) and allow a weighted mean (4706 [+ or -] 25 BP) to be calculate which provides a calibrated date of 3630-3370 cal BC. In the local context, the dates from the Greater Cursus are indistinguishable from those for the Lesser Cursus (OxA-1404; 4550 [+ or -] 120 BP and OxA-1405; 4640 [+ or -] 100 BP). The latter have broad standard deviations, and could now be improved using more accurate methods. It should be possible to establish whether the Lesser Cursus was built substantially later (i.e. in the later thirty-fourth to the thirtieth centuries BC) but it is unlikely to identify which was earlier if the two monuments were built within a century or two of each other. However, it is now clear that the Greater Cursus was constructed long before the first phase of Stonehenge, at 3015-2935 cal BC (Bayliss et al. 1997: 46). In realigning the landscape in a fundamental way, the cursus can be said to have established the conditions within which Stonehenge was created.

[FIGURE 6 OMITTED]

Discussion

The 2007 excavations have served to clarify the history and character of the Greater Stonehenge Cursus in a series of ways. Most important is the radiocarbon evidence, which places the monument into both a regional and a broader chronological sequence. The date of 3630-3370 cal BC is consistent with other determinations on reliable material from secure primary contexts in ditched cursus monuments elsewhere in southern Britain (Figure 7). The Dorset and Drayton cursuses apparently fit into the same interval, while the comparatively early date for Sarn-y-bryn-caled may be attributable to old wood effect (Barclay & Bayliss 1999: 18), leaving only the Dorchester-on-Thames monument looking appreciably later. This harmonises with recent indications that particular monumental traditions (such as long mounds and causewayed enclosures) may have been operative over comparatively short periods, rather than representing characteristic elements of social formations that remained stable over many centuries (Whittle et al. 2007: 138). In turn, the notion of a Neolithic period distinguished by continuous historical change is strengthened. In a wider European context, the identification of the thirty-sixth to thirty-fourth centuries BC as a broad chronological horizon for the construction of the large ditched cursus monuments of southern Britain is highly significant. As Andrew Sherratt long ago pointed out, Neolithic Britain saw an escalation in the scale and complexity of monuments that was really only matched in Brittany (Sherratt 1984: 130). The large cursuses marked the point at which this distinctive trajectory toward the colossal first emerged.

[FIGURE 7 OMITTED]

A significant observation made in the course of fieldwork was that, as one walks toward Trench 28 from the west, along the line of the western portion of the southern cursus ditch, Beacon Hill looms on the horizon immediately above the cutting. This prompted the speculation that this portion of the ditch might have been aligned on Beacon Hill. More specifically, the westernmost 600m or so of the southern ditch points directly at a notch on the northern side of Beacon Hill, formed by the intersection of the chalk with the Reading Beds, which is especially prominent when viewed from the west. Descending into Stonehenge Bottom, the ditch kinks slightly to the north, toward Amesbury 42, as we have already noted. At the western terminal, the King Barrow Ridge and Amesbury 42 are barely visible, for the Fargo ridge rises up to form a near horizon. It is conceivable, then, that the western end of the cursus was laid out in relation to Beacon Hill, with Amesbury 42 providing the referent for the eastern portion. This implies that the southern side of the cursus was laid out first, with the northern ditch added by a series of offsets, which might explain its more sinuous course. A further implication of this argument is that Amesbury 42, at least in its primary form prior to the digging of the second set of flanking ditches, must pre-date the cursus. However, the likelihood that the construction of the monument proceeded from west to east has no necessary bearing on the way that it was to be used.

The distinction between the terminal ditch and the side ditches was evident both in the extent of the cut feature and in the comparative density of knapping clusters in Trench 26 by contrast with Trenches 27 and 28. The presence of debitage knapped in situ in the ditches of long barrows and cursus monuments has been remarked upon before (Thomas 1999: 78; Whittle et al. 1993: 210), and it has been conjectured that the manufacture of any artefact may have been less important than the practice of flintworking in itself. This seems to be supported by the material from the western terminal ditch which, technologically, appears atypical in a fourth millennium BC context, although not lacking in skill.

While the contrast between terminal and side ditches is suggestive, that between Trench 28 and the other cuttings was startling. The orange-brown fine clayey recut fill is perplexing, as neither its explanation as a deliberate backfill nor as a wind-blown deposit derived from the ploughing of an area of clay with flints appears entirely satisfactory. Undoubtedly, this part of the cursus ditch was treated in an entirely different way from the rest of the monument, first having had its chalk rubble fill cleaned out, and later having been recut, during a time in which Beaker pottery was in use at the earliest. Despite the line of stones indicating a standstill horizon during its deposition, the fill was remarkably homogeneous, and this argues for a shorter rather than longer time over which it was laid down. A very similar sequence was present in Richards' W56A cutting, as well as in Stone's trench, but not in any other excavation that has been conducted on the cursus ditch. It may not be coincidental that the evidence for ditch-cleaning, recutting and orange-brown decalcified deposits are restricted to the westernmost part of the southern side ditch, precisely the portion of the ditch that is aligned on Beacon Hill, and which we have suggested may have been the first part of the cursus to have been laid out. Is it possible, then, that this part of the monument possessed a certain primacy, which was recognised and remembered over a very long period, resulting in its refurbishment or recreation on a number of separate occasions?

On the other hand, there is other evidence for the cursus having been re-established at other times. One of the most important pieces of evidence from the 2007 excavation is the recognition that Stone's 'embayment' was not an isolated feature, but part of a series of intrusive pits which might conceivably extend along the entire length of the monument. The original 'recess' was matched within Trench 28 by cut 050, and in Trench 27 by cut 030/081. The latter was tightly positioned stratigraphically, between the silting of the (Early Neolithic) cursus ditch and the V-shaped (Early Bronze Age) recut. This means that the Later Neolithic date from the antler that Stone recovered from his embayment is entirely comprehensible, but all the more intriguing. For it indicates that the architecture of the Greater Cursus was reinstated in the form of a discontinuous series of pits at much the same time as the construction of the sarsen settings at Stonehenge and the Southern Circle at Durrington Walls (Parker Pearson et al. 2007: 626-31). In other words, this was a time at which the greater Stonehenge landscape as a whole was being extensively reconfigured.

The lack of internal features in the cursus is perhaps disappointing, yet it confirms the contrast between this massive enclosed space and monuments like Durrington Walls, with their extensive evidence for occupation and deposition. The cursus was a conspicuously 'clean' place, and this may reflect its status as an area that was important, and yet had been set aside as either sanctified or cursed (see Johnston 1999). Given this comparative cleanness, the recovery of the antler from the base of the ditch in Trench 26 appears particularly fortuitous. The dates which it has provided have confirmed the early date of the cursus, while the identification of the probable cleaning-out of the ditch, of two phases of recutting and of a Bronze Age palisade in the cross-ditch indicate its longevity. Over a period of more than a millennium, the cursus represented a significant element of the Stonehenge region. Initially laid out in such a way as to integrate references to the local topography and to existing monuments, the cursus was repeatedly reworked and reinstated, so that it could be incorporated into new configurations of the landscape.

Acknowledgements

We would like to thank the National Trust for permission to excavate at the Greater Cursus, and the Arts and Humanities Research Council and the National Geographic Society for financial support. We also thank our colleagues on the Stonehenge Riverside Project, Umberto Albarella, Mike Allen, Mark Dover, Charley French and Karen Godden, and the project staff who worked in the field: Marcus Brittain, Ben Chan, Jolene Debert, Irene Garcia, Ian Heath, C.J. Hyde, Neil Morris and Phil Wilson. Amanda Chadburn, Chris Gingell, Richard Osgood, Mike Pitts and Julian Richards provided invaluable advice and assistance.

Received: 21 January 2008; Accepted: 17 June 2008; Revised: 11 July 2008

References

BARCLAY, A. & A.L. BAYLISS. 1999. Cursus monuments and the radiocarbon problem, in A. Barclay & J. Harding (ed.) Pathways and ceremonies: the cursus monuments of Britain and Ireland: 11-29. Oxford: Oxbow.

BARRETT, J.C., R.J. BRADLEY & M. GREEN. 1991. Landscape, monuments and society: the prehistory of Cranborne Chase. Cambridge: Cambridge University Press.

BAYLISS, A.L., C. BRONK RAMSAY & F.G. McCORMAC. 1997. Dating Stonehenge, in B. Cunliffe & C. Renfrew (ed.) Science and Stonehenge: 39-60. Oxford: Oxford University Press.

BRADLEY, R.J. & R. CHAMBERS. 1988. A new study of the cursus complex at Dorchester on Thames. Oxford Journal of Archaeology 7: 271-89.

BROPHY, K. 2000. Water coincidence? Cursus monuments and rivers, in A. Ritchie (ed.) Neolithic Orkney in its European context: 59-70. Cambridge: McDonald Institute for Archaeological Research.

BURL, H.A.W. 2006. Stonehenge: a new history of the world's greatest stone circle. London: Constable.

CHRISTIE, P. 1963. The Stonehenge Cursus. Wiltshire Archaeological Magazine 58: 370-82.

DARVILL, T.C. 2007. Stonehenge: the biography of a landscape. Stroud: Tempus.

FAR,R, R. 1917. Excavations in 'The Cursus', July 1917. Unpublished manuscript, Devizes Museum.

GREENWELL, W. 1877. British barrows: a record of the examination of sepulchral mounds in various parts of England. Oxford: Clarendon Press.

GRINSELL, L.V. 1978. The Stonehenge barrow groups. Salisbury: Salisbury & South Wiltshire Museum.

HOARE, R.C. 1810. The ancient history of Wiltshire. London: William Miller.

IXER, R. & P. TURNER. 2006. A detailed re-examination of the petrography of the Altar Stone and other non-sarsen sandstones from Stonehenge as a guide to their provenance. Wiltshire Archaeological and Natural History Magazine 99:1-9.

JOHNSTON, R. 1999. An empty path? Processions, memories and the Dorset Cursus, in A. Barclay & J. Harding (ed.) Pathways and ceremonies: the cursus monuments of Neolithic Britain and Ireland: 39-48. Oxford: Oxbow.

PARKER PEARSON, M., C. RICHARDS, M. ALLEN, A. PAYNE & K. WELHAM. 2004. The Stonehenge Riverside Project: research design and initial results. Journal of Nordic Archaeological Science 14: 45-60.

PARKER PEARSON, M., R. CLEAL, P. MARSHALL, S. NEEDHAM, J. POLLARD, C. RICHARDS, C. RUGGLES, A. SHERIDAN, J. THOMAS, C. TILLEY, K. WELHAM, A. CHAMBERLAIN, C. CHENERY, J. EVANS, C. KNUSEL, N. LINFORD, L. MARTIN, J. MONTGOMERY, A. PAYNE & M. RICHARDS. 2007. The age of Stonehenge. Antiquity 81: 617-39.

RICHARDS, J. 1990. The Stonehenge Environs Project (Historical Buildings & Monuments Commission for England Archaeological Report 16). London: Historical Buildings & Monuments Commission for England.

Royal Commission on the Historical Monuments of England. 1979. Stonehenge and its environs: monuments and land use. Edinburgh: Edinburgh University Press.

SHERRATT, A. 1984. Social evolution: Europe in the Later Neolithic and Copper Ages, in J.L. Bintliff (ed.) European social evolution: archaeological perspectives: 123-34. Bradford: University of Bradford.

STONE, J.F.S. 1947. The Stonehenge Cursus and its affinities. Archaeological Journal 104: 7-19.

STUKELEY, W. 1740. Stonehenge, a temple restor'd to the British druids. London: Printed for W. Innys & R. Manby.

THOMAS, J.S. 1999. Understanding the Neolithic. London: Routledge.

--2006. On the origins and development of cursus monuments in Britain. Proceedings of the Prehistoric Society 72: 229-42.

--(ed.) 2007. Place and memory: excavations at the Pict's Knowe, Holywood and Holm Farm, Dumfries and Galloway, 1994-8. Oxford: Oxbow.

THURNAM, J. 1869. On ancient British barrows, especially those of Wiltshire and the adjoining counties (part I--long barrows). Archaeologia 42: 161-244.

WARD, G.K. & S.R. WILSON. 1978. Procedures for comparing and combining radiocarbon age determinations: a critique. Archaeometry 20: 19-31.

WHITTLE, A., R.J.C. ATKINSON, R. CHAMBERS & N. THOMAS. 1992. Excavations in the Neolithic and Bronze Age complex at Dorchester-on-Thames, Oxfordshire, 1947-1952 and 1981. Proceedings of the Prehistoric Society 58: 143-201.

WHITTLE, A., A.J. ROUSE & J.G. EVANS. 1993. A Neolithic downland monument in its environment: excavations at the Easton Down long barrow, Bishop's Cannings, north Wiltshire. Proceedings of the Prehistoric Society 59: 197-240.

WHITTLE, A., A. BARCLAY, A. BAYLISS, L. MCFAYDEN, R. SCHULTING & M. WYSOCKI. 2007. Building for the dead: events, processes and changing worldviews from the thirty-eighth to the thirty-fourth centuries cal BC in southern Britain. Cambridge Archaeological Journal 17 (supplement): 123-47.

Julian Thomas (1), Peter Marshall (2), Mike Parker Pearson (3), Joshua Pollard (4), Colin Richards (1), Chris Tilley (5) & Kate Welham (6)

(1) School of Arts, Histories and Cultures, University of Manchester, Manchester, UK (Email: julian.thomas@ manchester.ac.uk)

(2) ARCUS, Sheaf Bank Business Park, Sheffield, UK

(3) Department of Archaeology, University of Sheffield, Sheffield, UK

(4) Department of Archaeology and Anthropology, University of Bristol, Bristol, UK

(5) Department of Anthropology, University College London, London, UK

(6) School of Conservation Sciences, University of Bournemouth, Bournemouth, UK

Table 1. Radiocarbon dates from the Greater Stonehenge Cursus, 2007.

Lab code Material and context Date BP

OxA-17953 Antler, base of cursus 4716 [+ or -] 34
OxA-17954 ditch, Tr 26 4695 [+ or -] 34

 Date cal BC at
Lab code 95.4% confidence level [sup.13]C relative to VPDB

OxA-17953 3632-3375 -21.70
OxA-17954 3630-3370 -21.59