Colin has excavated a standing-stone quarry at Vestra Fjold in Orkney so he knows a fair bit about what a Neolithic megalith quarry should look like.52 He has discovered that Neolithic quarry workers did not need stone mauls and pounders, nor did they leave much in the way of stone waste. Instead, they split natural pillars away from the bedrock by following existing fault lines, presumably using wooden wedges that were hammered into cracks and left in the rain to swell, thereby detaching the monoliths from the living rock. Neolithic quarry workers would then dig a rough trench at the base of the outcrop, within which the monoliths could be propped on stones before being set on to wooden cradles and hauled away.
On our first visit to Carn Goedog, we were joined by Louise Austin, a senior archaeologist at Dyfed Archaeological Trust. Her colleagues had recently carried out a survey of the area, and she pointed out prehistoric field boundaries and settlements; this rough moorland was once farmland, before the development of the blanket bog that now envelops Preseli’s lower slopes. Carn Goedog is an impressive outcrop of spotted dolerite, clinging to the side of the hill rather than sitting on the true summit. Broken boulders from the outcrop litter the hillside but, on its uphill side, the rock forms pillars reaching skyward. Heaps of broken stones here are evidence of more recent quarrying, but Colin recognized a narrow depression, about five meters wide, running close to the edge of the outcrop along this upper side. It would have been easy to detach the pillars, and to set them up on props within this gully. There are even three long stones lying within the depression, far enough away from the rock face to be something more than chance rockfalls. Might these be monoliths abandoned within the quarry? There is even an excellent natural ramp, completely free of stones or boulders, running from here down the side of the hill toward the low ground of the valley in which Craig Rhosyfelin sits.
The maximum extent of the British–Irish Ice Sheet c. 27,000 years ago. The glaciers from this and previous glaciations never reached Salisbury Plain.
The biggest surprise awaited us just below Carn Goedog. Set on two natural terraces are fifteen prehistoric houses.53 They were first identified in the 1970s by Peter Drewett, coincidentally one of Geoff’s supervisors at Durrington Walls in 1967. Their official description is “Bronze Age hut circles” but, when we looked closer, we could see that nine are actually rectangular in plan. Such architecture is not found in the Bronze Age or even in the Iron Age, when houses were uniformly round throughout Britain. These houses could well be Neolithic in date, and are slightly smaller than the houses at Durrington Walls. Might these be the quarry workers’ houses? If so, could they give us a date for when the spotted dolerite bluestones were being quarried? Once again, there’s only one way to find out, and that is by archaeological excavation. Perhaps one day archaeologists will be allowed to carry out a dig here within the National Park.
Traces of Neolithic quarries are impossible to verify without digging, so in 2011 our Stonehenge team moved to Preseli. Four days before we left for west Wales, Richard Bevins contacted me to say that he and Rob Ixer had pinpointed the geological match for one of the Stonehenge rhyolites. Thanks to its unique geology—called “Jovian” by Rob because under the microscope its swirls and blobs look like the weather system on Jupiter—it could be located to a precise part of the Craig Rhosyfelin outcrop, the very same spot we reckoned was a Neolithic quarry. We carried out geophysical surveys on the ground next to the outcrop, but the results were inconclusive. Only digging would provide an answer. The farmer and his neighbors were enthusiastic about our plans, and soon we had the use of a mechanical mini-digger to open up a good-sized area next to the outcrop. It wasn’t long before we discovered that the ancient ground surface had been protected under layers of soil washed down from higher up the valley. When we started finding hammerstones on that ground surface, we realized that we had not just a prehistoric quarry but a perfectly preserved one—the Pompeii of prehistoric stone quarries. If that wasn’t enough, Josh’s expert use of the mini-digger unearthed a long slab of rhyolite lying on the quarry floor. At 13 feet long and weighing about four tons, it is the same size as the larger bluestones at Stonehenge. Someone had left behind a monolith when the quarrying had ended. We could hardly believe our luck. This was a smoking gun; the game was up for anyone still trying to argue that the bluestones were not quarried in Preseli during the Neolithic, and then taken to Wiltshire.
Our small trench also yielded tantalizing glimpses of what we may learn in future years. I’d always assumed that archaeology could never answer questions about how the stones were moved, but here is a monolith abandoned in transit. It lies at the end of what appear to be three parallel rails of stones set on edge, leading to that part of the outcrop, about 15 meters away, from which it has been detached. These rails could have provided a fulcrum for moving the stones with wooden levers in just the way that Gordon Pipes predicted. The monolith’s position at the end of the rails and just in front of a drop in the ground level also suggests that it was about to be transferred to a new means of transport. Was this the point where monoliths were lifted onto a wooden sledge and rollers? Or might they have been carried on wheeled sledges? Future excavation of the prehistoric track leading out of the quarry will tell.
Richard and Rob had been busy with geological identifications. As well as Carn Goedog, they had identified other Stonehenge bluestones as coming from the northern edge of Preseli: spotted dolerite from Carn Breseb and plain dolerite from Carn Ddafad-las. It was beginning to look as if the bluestones had been quarried from outcrops of the upper Nevern valley, and not from the top of Preseli or its southern flanks. On our last day of fieldwork in 2011, Colin and I went in search of other quarries where the River Nevern forms a deep gorge with rapids and a waterfall. Rob and Richard suspect that the other types of igneous rock from Stonehenge might be found here, and a local farmer pointed us to the gorge’s dramatic outcrops. Wading through torrents, hacking through brambles and scaling steep cliffs, Colin and I not only had a rare opportunity to behave like Indiana Jones but also found two likely quarry sites beside the cacophonous waters.
If the bluestone sources were concentrated in the valleys of the upper Nevern, then not only have we all been looking in the wrong places on top of Preseli but we need to see if there is any evidence of Neolithic occupation that might provide a social context for the bluestones. Apart from the local portal dolmens, such as Pentre Ifan, very little is known about Neolithic settlement remains in this area. My eye is drawn to Castell Mawr, one of the many Iron Age hillforts in the area, because its geometrically oval plan looks remarkably like a henge. With its large outer bank seemingly without an outer ditch, and its small inner bank next to a wide inner ditch, Castell Mawr looks like a Neolithic henge whose outer bank and inner ditch have later been modified by Iron Age people into a hilltop fort. The same conclusion was reached in the 1980s by archaeologists carrying out a partial geophysical survey.54 Future investigation should allow us to find out if our suspicion is right. If so, this would be the largest henge in Wales, a fitting social and political center for the people of the bluestones.
Our geophysical survey of Waun Mawn failed to reveal any evidence that its four stones were once part of a larger stone circle. If the bluestones were set up as one or more circles, and then dismantled and moved to Stonehenge, a more likely proposition is that any such dismantled circle would have been located in or around the potential henge at Castell Mawr. With two likely quarries in the Nevern gorge directly below Castell Mawr, there is a good chance that the bluestone sources were all providing stones for a local ritual center. From Craig Rhosyfelin, for example, stone movers were confined to the steep-sided valley floor until reaching the gentle incline that leads out of the Nevern valley to Castell Mawr a mile away.
Our fieldwork in 2011 opened up an entirely new hypothesis that can be followed up in future years through geological sampling and archaeological excavation. If we can demonstrate that the bluestone quarries were on the north side
of Preseli, within the Nevern valley and focused on a large henge, then the healing theory, which relies on the supporting evidence of springhead sites on top of Preseli and its southern flanks, can be rejected. Our find at Craig Rhosyfelin also raises the likelihood that the link between Preseli and Stonehenge was not to do with a magic mountain or any intrinsic properties of its bluestones, but involved a powerful polity within the Nevern valley, a people whose earliest Neolithic ancestors had brought traditions of megalith construction to this part of Britain. Perhaps they had celebrated their power and their ancestry by erecting one or more stone circles with monoliths taken from a range of nearby quarries. Then later, for one reason or another, the momentous decision was taken to dismantle these circles and move them over 180 miles to Stonehenge. If these now-vanished stone circles are there, one day archaeologists will find the emptied holes where bluestones once stood.
The Neolithic people of the Nevern had potentially excellent transport links. The glaciated valleys of Preseli have U-shaped profiles, with wide, flat, stone-free bottoms. There is plenty of room for moving megaliths along a valley bottom without having to negotiate its stream. Even where water may have to be crossed, the streambeds are shallow and firm, so are no great impediment.
North of Carn Goedog and Craig Rhosyfelin, the Brynberian valley joins the much larger valley of the River Nevern (Nyfer in Welsh), which flows westward to meet the sea at Newport. This is one of the hotspots of Britain’s Early Neolithic. The Nevern valley is dotted with portal dolmens, both on its slopes, like Pentre Ifan, and in its valley bottom at Carreg Coetan. Little has yet been found of these earliest farmers’ settlements, but they probably lived along the valley sides. The valley’s slopes would have been densely forested, so the open and lightly wooded moorland pastures of the Preseli Hills would have been reached by following the Nevern valley’s tributaries, such as the Brynberian valley. These wide-bottomed valleys would have formed ideal droveways for taking the cattle on to the high pastures, past Craig Rhosyfelin and on to Waun Mawn and Carn Goedog. These two bluestone outcrops were probably well-known landmarks to generations of Neolithic farmers long before Stonehenge was built.
As the team explored the ancient landscape of Pembrokeshire, Andrew Chamberlain and I were convinced that Newport, where the River Nevern meets the sea, was the embarkation point for the bluestones’ sea voyage along the Welsh coast, but Colin was thinking otherwise. He reminded us of the pageantry and display involved in megalith-moving around the world today; these Neolithic stone-moving events were surely great social spectacles involving thousands of people, the sacrifice of animals, and plenty of feasting, not simply laboring tasks to be performed as quickly as possible. Taking stones by sea would have limited the renown of such feats; dragging the stones in long-distance relays overland, handing them on from one community to the next, was surely a more likely proposition.
What has persuaded me to Colin’s point of view is the distribution of Neolithic polished-stone axes in south Wales. Somewhere not yet located within west Wales, probably not far from St. David’s Head, lies a source of rhyolite that was used to make stone axes in the fourth millennium BC, during the centuries before Stonehenge was built. These axes are of a different type of rhyolite to the four types of rhyolite identified at Stonehenge; geologists can identify this specific rock under the microscope. The St. David’s source is known as Group VIII, and its products are distributed all over south Wales.55 Axes were also imported from Cornwall into Wales, but these all have a coastal distribution, whereas the Group VIII axes are found mostly at inland locations east of Preseli. In other words, unlike the Cornish imports, these objects were not being moved by sea but were traded overland. Long before the bluestones were dragged off the hills, there had been an established network of routeways leading eastward from Preseli.
Recent archaeological investigations in advance of new pipelines have found evidence of many Neolithic sites in south Wales valleys.56 Neolithic traders would have used these glaciated valleys not only to avoid the thickly wooded hillsides but also to pass through the many settlements. The principal routeways would have followed the valleys of rivers such as the Taf, the Towey, and the Usk. These flat-bottomed valleys were the Neolithic equivalent of freeways, cleared of forest by the earliest Neolithic farmers and facilitating long-distance movement of people and their goods. For movers of bluestones, the route was relatively straightforward through south Wales—eastward along the Nevern valley, across its low watershed into the Taf valley, following that valley southward to what is now Carmarthen, then joining the Towey valley heading northeast into the Brecon Beacons. After the gentle incline of the pass leading into the Usk valley, the route was downhill past Brecon as far as the deepening waters of the River Usk. Today most of this route is followed by the A40, and the many ancient sites—Roman camps and Medieval castles—along the way are testament to its enduring importance for access in and out of Wales.
On reaching the lower Usk valley, travel by water would have been difficult to avoid, but the calm waters of the Severn estuary are at least easier to navigate than the ocean swells of the Irish Sea. If the stones were shipped by boat from somewhere around what is now the town of Usk, they could have been taken across the estuary to Avonmouth, where the Somerset Avon reaches the sea after winding its way from the edge of Salisbury Plain. Somewhere around Bradford-on-Avon, the bluestones could have been landed and pulled the last twenty miles to Stonehenge.
I doubt whether archaeologists will ever be able to prove beyond doubt that this was the route taken by the bluestone movers. If the source of the Altar Stone and the other sandstone monoliths at Stonehenge can be found (these could be from the Brecon Beacons or from west Wales), it may just help to confirm or negate this theory. Finding the sources of all the other types of bluestone—rhyolitic ignimbrite (tuff), calcareous ash, volcanic ash, and altered volcanic ash, unspotted dolerite, and one other type of rhyolite—will also help to identify whether all or just some of the stones were taken from the northern edges of Preseli, and which route was taken. There is much for archaeologists and geologists to look for in the years to come.
17
ORIGINS OF THE SARSENS
__________
The Stonehenge sarsens never attract as much interest as the bluestones, and discussions about their origin have been much less heated. The sarsens really ought to be the center of attention because they’re so big. The extraction, transport, and shaping of these giants would have been an engineering project on a scale never before seen in prehistoric Britain. Sarsen occurs naturally in Kent as well as in central England, but it is best known from the deposits on the Marlborough Downs in north Wiltshire. Even today, fields around Fyfield Down and Piggledean are littered with large gray lumps of sarsen, and it’s easy to see how they acquired the name of graywethers, or sheep. They are scattered in stony flocks across the grasslands.
Since John Aubrey’s time, almost four hundred years ago, these downs have been thought to be the source of the Stonehenge sarsens. Yet none of the stones lying around today is anywhere near as big as those used for the Stonehenge uprights. The largest uprights are those of the great trilithon; the stone still standing (Stone 56) is more than nine meters tall and the fallen one (Stone 55) is ten meters long. Each of these must weigh around 35 tons and, together with their lintel, once formed the largest single megalithic structure in Britain.
The uprights of the sarsen circle are generally around five meters tall and each weighs around 20 tons. The squared-off appearance of the stones above ground (they look to me like a set of dominoes) belies their irregular bases: The buried parts of the stones, as Hawley discovered, are often irregular in shape.1 The builders were keen to present an image of symmetry and rectangularity above ground, so any awkwardly shaped ends are underground. The problem with this strategy is that it compromised the stability of many of the uprights, and is why so many have fallen down or have had to be restored in the twentieth century. During our last tr
ip to Stonehenge, Ramilisonina noticed immediately—drawing on his own knowledge of stone-erecting—that many of the fallen sarsens have toppled because of their inadequate bottom ends.
The lintels are much smaller than the uprights. All five of the trilithon lintels survive, three of them still in position on the tall uprights and two on the ground nearby. The lintel that once lay across the tops of Stones 59 and 60 (the north trilithon) has been broken into three lumps that no longer fit together; it has evidently been chipped and quarried in more recent centuries. The trilithon lintels are five meters to six meters long, and most of the still-complete examples of the surviving seven lintels in the outer circle of sarsens are each just over three meters in length.
In all, the builders of Stonehenge had to look for sarsen stone to create ten trilithon uprights, five trilithon lintels, thirty circle uprights, thirty circle lintels, four Station Stones and three Slaughter Stones, a total of eighty-two stones. The Heel Stone may be a naturally occurring sarsen, and therefore already present when building work began, and the three empty stoneholes between it and the Slaughter Stone could have once been filled by an earlier arrangement of the Heel Stone and the two accompaniments to the Slaughter Stone. If two of the three Slaughter Stones were already on site when the major building work began, in an earlier alignment with the Heel Stone, that would have made eighty sarsens to go out and find—about the same number as the bluestones.
Weathered natural sarsens develop a thick, hard crust that makes this rock extremely difficult to shape. Some sarsens have an iron-hard quartzitic surface that is resistant to stone tools, even in the hands of skilled masons. This quartzitic sarsen was only of interest to Stonehenge’s builders in its pebble and boulder form because these tough rocks could be used as hammerstones and mauls. Mauls were boulders that could be used to pound by hand, but were probably more efficient if suspended from a rope and swung at the sarsen block. By hitting it at the right angle, the stone worker could detach a large flake, thereby gradually shaping the block into the desired form. The softest sarsen has a sugary, crystalline texture, though there is much variation in texture and hardness.2
Stonehenge—A New Understanding: Solving the Mysteries of the Greatest Stone Age Monument Page 30