The voice of archaeology in Britain and beyond

Cover of British Archaeology 110

Issue 110

Jan / Feb 2010



Burnt mound theory tested to perfection

Dig find proves flowers placed in bronze age graves

UK's first complete Roman lantern found in Suffolk

Research continues as Saxon hoard is valued at £3.3m

in the press

in brief & phase 2


Newhenge: Latest discoveries and interpretations from the Stonehenge Riverside Project team

Dig the beat: Exploring pop music from an archaeological perspective, including additional online content

THE BIG DIG Mellor: A hillfort in the garden: This long-running research excavation near Stockport, Greater Manchester, is now ready for publication

The Peat Men from Clonycavan and Oldcroghan: Findings of the Bog Bodies Research Project at the National Museum of Ireland, with Bibliography


your views and responses

on the web

Caroline Wickham-Jones looks at archaeological gifts

Dan Pett summarises the website set-up and technologies for the Staffordshire Hoard


faux pas


Sebastian Payne asks what cremation burials can tell us

in view

Greg Bailey is impressed by Open University broadcasting

CBA Correspondent

Lynne Walker and Sue Morecroft look at the past year of listed building casework

my archaeology

David Attenborough remembers the early days of television


ISSN 1357-4442

Editor Mike Pitts



Last summer Mike Parker Pearson, Joshua Pollard, Julian Thomas and Kate Welham found evidence for a previously unknown stone circle near Stonehenge. This is their exclusive report on what they call "Bluestonehenge".

2009 was a perfect summer. The dig site was in a secluded meadow on the bank of the river Avon, and the sun shone most days. After work, we cooled down in a nearby swimming hole before heading back to our campsite in the garden of a riverside pub. After the miseries of the so-called summer of 2008, when the campsite turned into a sea of mud in the unceasing rain, this season's delights wiped away those soggy memories.

The year had started well – many of the Stonehenge Riverside Project team had previously met up on Easter Island, where SRP directors Colin Richards and Kate Welham are investigating the statues and their quarries. Radiocarbon dates had also come back from last summer's findings. The most exciting was the date of an apparently insignificant ditch found in the trench farthest away from Stonehenge, where we were now about to start digging again.

Bluestones at Stonehenge

In 2008 we dug within Stonehenge itself, recovering 60 cremation burials from a pit, one of the circle of 56 Aubrey Holes that are concentric with the ditch and bank (see News, Nov/Dec 2008). These had originally been excavated by William Hawley in the 1920s when he dug 32 of the Aubrey Holes. Because the cremated bones were considered at that time to be of little scientific value, they were dumped in a mixed-up heap in Aubrey Hole 7 in 1935.

Our greatest discovery in that small hole, however, was that Hawley's workmen had not fully dug it out. In the bottom sat the undisturbed residue of a layer of chalk packing and a patch of crushed chalk caused by the weight of a standing stone. Checking Hawley's diary, we found that he had initially decided that the Aubrey Holes once contained small standing stones that were later pulled out. Sadly, Hawley did not have the courage of his convictions and, when confronted with the huge postholes of Woodhenge at Maud Cunnington's excavations in 1926–27, he changed his mind.

We realised that the small stones that once stood in the Aubrey Holes had to have been bluestones, the monoliths from the Preseli Hills and other parts of south Wales. By radiocarbon dating a cremation burial found in the chalk packing of Aubrey Hole 32, dug by Richard Atkinson in 1950, we had evidence that this stone circle was likely to have been put up around the time that the ditch and bank were dug in 3015–2935BC. This and other radiocarbon dates at Stonehenge have been refined by Bayesian statistical modelling of its stratigraphic sequence.

This meant the conventional threefold scheme of Stonehenge as an earthen monument, succeeded first by wooden posts and then by stone uprights could not be correct. Stonehenge was a stone monument from the beginning, and five of our new radiocarbon dates on cremated and unburnt human bone showed that it had also been a place of burial from this moment until at least 2400BC (statistically modelled as 2470–2300BC) – certainly during the erection of megaliths, and probably after as well.

The Stonehenge Avenue

We made another dramatic discovery in 2008, immediately outside Stonehenge's north-east entrance. Here the Stonehenge Avenue leads towards the direction of summer solstice sunrise for about 500m, before turning sharply and heading towards the river Avon, a total distance of 2.8km (1.75 miles). The Avenue consists of two parallel ditches, about 20m apart, with banks on both sides. While confirming that these ditches were dug around 2400BC, after the sarsens were put up at Stonehenge (estimated to be between 2580 and 2470BC), we noticed something unusual. Or rather our environmental specialists, Mike Allen and Charly French, did.

The avenue ditches were dug alongside a pair of natural chalk ridges whose orientation was coincidentally on the midsummer sunrise in one direction and the midwinter sunset in the other. In 2009, GT Frontline, a Dutch ground-penetrating radar team, confirmed the presence of the two parallel ridges and mapped a third, equally-spaced parallel ridge to the east. It now looked as if the builders of the first Stonehenge had added their stone circle and cremation enclosure onto the end of a remarkable natural phenomenon, where the sun's solstitial extremes were apparently marked by the contours of the land.

We had found out why the Stonehenge Avenue's first stretch runs towards midsummer sunrise. But why does it then change direction towards the river? Why was it so important for Stonehenge to be linked via an avenue to the river Avon?

Among the other 15 trenches dug by the SRP in 2008, one was where the projected line of the Avenue met the riverbank. In 1973, George Smith, now with Gwynedd Archaeological Trust, had traced the Avenue's northern ditch as close as 150m from the river, but did it actually go any further? Within the meadow on the river's edge, we used every form of non-invasive survey in an attempt to answer that question – topographic, magnetometry, resistivity and radar. But there was no sign of any Avenue ditches.

An auger survey produced more favourable results: we were convinced that it had located a pair of parallel ditches near the river, but only digging would prove us right or wrong. When we opened up the trench, we foundthat one of these ditches was medieval. The other was not even parallel with it. Disappointment turned to puzzlement when we realised that this second ditch was curved. We had hit the northern edge of a ring ditch. Was this what the Avenue had been heading for, or was it a bronze age barrow built long after the Avenue? We desperately needed dating evidence.

On the last day of the dig in 2008, we were about to backfill. Reg Jury was already revving the engine of his mechanical digger when trench supervisor Jim Rylatt, removing the very last bit of clay from the ditch bottom, pulled out the broken-off tip of an antler pick. This would give us a secure radiocarbon date for the ring ditch's construction.

During the winter months, while we waited for results from the radiocarbon lab, we mulled over the implications of this curving ditch. It was likely to be a henge because its eroding bank had filled the ditch from the outside – such an arrangement being the key definition of a henge. The interior appeared to have a cairn or platform of flint nodules. Scrutinising the two resistivity plots of the area, we realised in retrospect that the ring ditch could be seen. One of the resistivity surveys also showed four areas of high resistance in a square arrangement within the circular ditch. Were these holes for standing stones, filled with the smashed-up debris of sarsens broken up in medieval times? But analysis of the finds in the ditch yielded no such debris.

When it came, the radiocarbon date of 2470–2280BC for the antler pick was too early for a bronze age barrow, and confirmed that we had indeed found a henge. The Avenue and the ring ditch could well have been built together. We now needed to see if the Avenue led up to it, and to find out whether it might enclose a setting of sarsens.

The Avenue at the river

Four trenches were planned in 2009: a long trench in the middle of the meadow to try again to locate the Avenue's two ditches, a shorter one to see if the Avenue's eastern ditch came as far as the henge, a trench in the north-east quadrant of the henge interior, and a small trench across the west side of the henge. Trench sizes were kept to a minimum to ensure that we could answer our research questions and still leave as much as possible undisturbed for future researchers.

The two trenches looking for the Avenue ditches were full of medieval features, mostly ditches and pits with 13th century pottery from the nearby Laverstock kilns. This was the time when the small suburb of West Amesbury was expanding along the road leading out of Amesbury. Despite high densities of worked flints from the topsoil, there were only three prehistoric features. One was a curving ditch with a neolithic arrowhead amongst its flints. The other two were parallel V-shaped ditches spaced 18m apart. They were packed with worked flint, and the eastern ditch contained a ripple-flaked oblique arrowhead, dating to around 2500–2200BC. We had found the Avenue.

Small postholes in the bottoms of the ditches showed that they had held palisades. The Avenue had run along a ridge of decayed chalk which sloped gently towards the water's edge. We did not find the end of it, but its eastern ditch reached as far as the edge of the henge bank. If henge and Avenue were contemporary, there was no room for a grand entrance to the Avenue at its riverside end.


Inside the henge, our expectations of a square setting of stone-filled pits were dashed when we realised that the high-resistance anomalies were produced by a carpet of naturally-deposited flint nodules. Then we spotted features near the henge's centre.

Among these were what at first looked like three large pits. These were soon revealed to be six interlinked pits in the main trench, forming a curving or circular cutting with three other pits in the western trench. Each pit had its own ramp leading in from the outside of the circle. By cutting our sections across the holes and along the axis of the circle of holes, we could see that their upper fills were disturbed by the removal of uprights that had once stood within them. The holes were wider and shallower than any dug for neolithic posts from nearby Woodhenge and Durrington Walls. These pits had once held uprights of stone, too small to have been slabshaped sarsens.

The standing stones had been placed on individually-tailored cushions of river clay and pads of packed flint nodules. Each pit base was different, suggesting that the co-operative activity of erecting a stone circle had been carried out by separate teams assigned to each stone. One stone had sat on a carefully constructed nest of nodules. Others had merely a thin cushion of clay between their bases and the chalk. Another sat on a rock-solid pad of nodules and rammed clay.

The bottoms of five of the holes contained imprints of their stones, pressed through the clay cushions into the soft chalk beneath. Whilst the profiles of the robbed-out holes had already indicated that they had contained neither posts nor sarsens, it now became patently clear from the imprints in their bases that they had held stones whose shapes closely matched the Stonehenge bluestones, with their variously curved, indented and straight edges. Assuming the nine excavated stoneholes and the curvature of their plan are representative of the complete arrangement, we can propose that it would have consisted of 25 monoliths in a 10m diameter circle.

When had these stones been taken away and where did they go? The circle's construction can be dated by two flint chisel arrowheads (a style used around 3400–2500BC) from the packing of the stoneholes. There are at least ten of this arrowhead type from the site, distinguishing it from Durrington Walls, for example, where the many arrowheads are mostly of the type from the Avenue ditch, or from Stonehenge itself where few arrowheads of any kind have been found. Similar numbers of chisel arrowheads to those from Bluestonehenge were found during excavations in the 1980s at Coneybury henge by Julian Richards, where a pit was dated to 3350–2765BC. Interestingly, this pit also had the typical appearance of a bluestone pit or Aubrey Hole at Stonehenge, and perhaps too held a monolith.

The Bluestonehenge builders had left an antler pick at the bottom of one of the stoneholes. Hopefully this will allow radiocarbon analysis to date its creation more precisely, but sadly the first attempt has failed because of inadequate collagen in the antler. We can, however, date the circle's dismantling. A second antler pick had been discarded on the ramp of one of the stoneholes after the stone had been removed: this has just been dated to 2469–2286BC (see end note). It is not possible to be sure whether the pick was used to remove the stone, or slightly later in digging a shallow ditch into the top of the circle of robbed-out stoneholes. This radiocarbon date does, however, demonstrate that one stone, and probably the whole circle, was removed around or before that date – at about the time of the construction of the henge.

The henge

As we excavated the primary fill of the henge ditch, we realised that it narrowed to a terminal on its east side, forming an east-facing entrance into the henge. The bottom of the ditch terminal was covered with a spread of artefacts – antler, stone and bone tools probably for flintknapping, an antler pick, knapped flints and the carbonised remains of an organic bowl or basket. This is a classic example of a special deposit of placed artefacts, a common feature of neolithic ditch ends.

At the centre of the henge, other features had been dug into the stone pits. These turned out to be a small penannular ditch and four postholes – whose traces were very different from those of the wider and shallower neolithic pits. The few sherds of pottery from these features were flint-gritted post-Deverel-Rimbury ware, dating to the late bronze age (1150–800BC). Someone had re-used the henge over a thousand years after its construction. The north-east sector of the henge ditch had also been recut, possibly at this time or perhaps earlier; this later backfill contained tiny sherds of Beaker pottery. In its north-west sector, the filling of the henge ditch had been punctuated on two occasions during the bronze age by the laying of flint cobble surfaces on its base. Were these where people gathered to watch events inside the henge?

Route to the afterlife?

Stonehenge was massively rebuilt around 2500BC (when the sarsen circle and trilithons were erected), and renovated again around 2200BC. The 56 Aubrey Hole bluestones were rearranged in two settings within the sarsens, together with an estimated further 23 to 33 bluestones from somewhere else. Were the monoliths from West Amesbury – Bluestonehenge – the additional stones? If so, did the Avenue commemorate the route they were dragged along to join the others at Stonehenge? Perhaps the Stonehenge Avenue was marked by its ditches some time after its actual period of use.

The Avenue is not the most direct route from Bluestonehenge to Stonehenge but its gentle contours provide the best incline to pull big stones up from the river. The shortest route on foot, however, climbs the valley which leads to Coneybury henge and thence follows the dry valleys to Stonehenge, to approach either Stonehenge's south entrance or its more grandiose north-east entrance.

What was the purpose of Bluestonehenge? Topsoil and turf collapsing from the uppermost edges of the stoneholes filled the voids left by their extracted stones. As well as providing evidence for the riverside environment, this soil contains further clues about the activities performed on the ground surface among the stones.

Analysis of these soil samples is only just underway, but the results so far are very exciting. One stonehole contained a chip of micaceous Devonian sandstone, which may be a match for Stonehenge's Welsh sandstone monoliths including the Altar Stone. There is also a lot of wood charcoal from the former ground surface, together with a fragment of human bone. Were bodies brought here for cremation and excarnation, prior to their bones being taken for burial at Stonehenge? If so, then this is another piece of the jigsaw which puts Stonehenge as a final destination for the ancestors, reached from the domain of the living at Durrington Walls via their riverine route to the afterworld.

The Stonehenge Riverside Project is run by a consortium of university teams directed by Mike Parker Pearson (University of Sheffield), with codirectors Josh Pollard (University of Bristol), Julian Thomas (University of Manchester), Kate Welham (University of Bournemouth), Colin Richards (University of Manchester) and Chris Tilley (UCL). The 2009 excavation was funded by the National Geographic Society, Google, the Society of Antiquaries of London, and the Society of Northern Antiquaries (Denmark). The wider project is funded by the Arts and Humanities Research Council and the Royal Archaeological Institute. The new date for Bluestonehenge is OXA–21278, 3884±30BP.

index | next >>

CBA web:

British Archaeology

Jan/Feb 2005
Mar/Apr 2005
May/Jun 2005
Jul/Aug 2005
Sep/Oct 2005
Nov/Dec 2005
Jan/Feb 2006
Mar/Apr 2006
May/Jun 2006
Jul/Aug 2006
Sep/Oct 2006
Nov/Dec 2006
Jan/Feb 2007
Mar/Apr 2007
May/Jun 2007
Jul/Aug 2007
Sep/Oct 2007
Nov/Dec 2007
Jan/Feb 2008
Mar/Apr 2008
May/Jun 2008
Jul/Aug 2008
Sep/Oct 2008
Nov/Dec 2008
Jan/Feb 2009
Mar/Apr 2009
May/Jun 2009
Jul/Aug 2009
Sep/Oct 2009
Nov/Dec 2009
Jan/Feb 2010
Mar/Apr 2010
May/Jun 2010
Jul/Aug 2010
Sep/Oct 2010
Nov/Dec 2010
Jan/Feb 2011
Mar/Apr 2011
May/Jun 2011
Jul/Aug 2011
Sep/Oct 2011
Nov/Dec 2011
Jan/Feb 2012
Mar/Apr 2012

CBA Briefing

Courses & lectures
CBA Network
Grants & awards

CBA homepage