Bulford: Pits, Double Henges and a Claimed Solstice Alignment — Some Sceptical Notes

18 June 2026

News reports this week described two large post pits at Bulford, roughly 120 m apart and aligned on the midsummer sunrise and midwinter sunset around 2950 BC. A finely worked disc-shaped flint knife, placed upright in a smaller pit on the same line, was presented as possible supporting evidence for deliberate solar symbolism.

The discovery has been described as a possible early “prototype” for solar marking in the Stonehenge landscape. Before accepting that interpretation at face value, however, it is worth pausing to consider some basic questions about context, numbers, and coincidence.

The clearest published plan of the surrounding archaeology is Figure 10 below. It shows a noticeable scatter of Late Neolithic pits across the hilltop, with the densest concentration lying southwest of two conjoining segmented ring ditches (the double henges). Some pits occur inside or immediately adjacent to the henges. A smaller, more isolated group appears in the inset.

Figure 10: Woodlands pits at Bulford.

Reproduced from Leivers, M. 2022. The Army Basing Programme, Stonehenge and the Emergence of the Sacred Landscape. Internet Archaeology 56. https://intarch.ac.uk/journal/issue56/2/full-text.html (Figure 10). Used under the journal’s licensing terms.

What the Plan Actually Shows

Dozens of pits are visible on this single plan (easily 30–50+ black dots across the main area and insets). The whole site is not randomly scattered; it occupies a ridge with particular views and topography. The double henges have northern entrances. In such a setting, features aligned roughly on the solstices are more likely to occur than they would on completely flat or randomly oriented ground.

The Structural Reality: Credit to the Excavators

Phil Harding did not simply draw a line across a map and hunt for pits to match it. That is bad archaeology, and Harding is far too experienced for that. What actually happened is much more structurally sound: out of 48 pits excavated across the site, Harding noticed that two were morphologically and functionally exceptional. While 46 were typical shallow scrape-holes, these two measured roughly a metre across and 30 inches deep, with steep sides and environmental traces of ash wood. They were engineered structural postholes dug to support massive timber uprights. Because they are the only two massive structural post pits on the 30-acre site, drawing a line between them isn't arbitrary. Harding identified the structural anomaly first, and then checked the alignment. He noted it sat about 50 degrees off direct north—matching the midsummer sunrise.

The Upright Knife and Structured Deposition

Work on the site has produced several pits with clear structured deposits. At least three contain discoidal flint knives:

• Pit 5008 — discoidal knife in a rich flint assemblage with Grooved Ware and animal bone (~3100–2900 cal BC).
• Pit 8050 — polished discoidal knife with paired antlers, a small spherical chalk ball, Grooved Ware and other finds (~3020–2900 cal BC).
• Pit 8331 — discoidal knife with ball flints, greenstone axe fragment, worked chalk and tools (~3020–2890 cal BC).

These are genuine special deposits. The question is whether the upright “star find” knife highlighted in the news comes from one of these known pits, from another pit in the same cluster, or from a feature more directly associated with the two large post pits. Until the exact context number and its position on a plan like Figure 10 are published, we cannot tell whether the knife strengthens the case for deliberate solar symbolism or simply adds to the general pattern of careful deposition already visible across many pits.

What We Can Reasonably Say

Bulford contains a genuine concentration of Late Neolithic pits with structured deposits, including multiple discoidal knives, set around and near a pair of segmented ring ditches. The 2026 reports add two large post pits and an upright knife on a claimed solstitial line. That combination is interesting and worth serious attention.

However, with dozens of pits visible on the published plan, a general topographic grain that favours certain orientations, and no detailed statistical assessment or precise overlay of the new features yet available, it remains possible that the alignment is at least partly coincidental. Distinguishing deliberate solar marking from the background pattern of pit digging and deposition will require the original context plans, the exact numbers and positions of all features, and a transparent statistical approach.

Figure 10 remains the best publicly available overview of the pit-and-henge complex. It provides the essential framework for asking these questions, even if it cannot yet answer them definitively.

Further detailed plans and analysis from the excavation archive would allow a more robust evaluation. Until then, a degree of scepticism is warranted alongside interest in the new evidence.

Example pit alignment

On the published plan of the Bulford pits (Figure 10), one possible northeast–southwest line can be drawn that passes through several black dots representing pits, including what appears to be one of the larger features. At least three, and possibly more, pits lie on or close to this line, with spacing between some of them roughly consistent with the 120 m distance given in the news reports for the two large post pits. The news coverage states that the line through the two post pits has an azimuth of 48°. This raises the possibility that the two post pits may form part of a longer linear arrangement rather than standing as an isolated pair. This is only one visual example and remains a guess until the exact positions of the reported post pits are plotted on the plan.

---

References

• Greaney, S.E. 2022. The Archaeology of Power: Understanding the Emergence and Development of Neolithic Monument Complexes in Britain and Ireland. Volume 2: Appendices. PhD thesis, Cardiff University / University of Southampton.
• Leivers, M. 2022. The Army Basing Programme, Stonehenge and the Emergence of the Sacred Landscape. Internet Archaeology 56. https://intarch.ac.uk/journal/issue56/2/full-text.html
• Wessex Archaeology grey literature reports (2014–2020) on the Bulford Army Basing Programme excavations (synthesised in the above sources).
• News reports published 18 June 2026 (e.g. The Guardian, BBC) on the Bulford solstice post pits and associated finds.

Prototype Stonehenge Structure at Bulford

Reconstruction of summer solstice celebrations as they might have appeared at Bulford 5000 years ago

 © Wessex Archaeology

Factual Archaeological Report – June 2026

Compiled from Wessex Archaeology announcements and contemporary reporting.

Executive Summary

A Neolithic wooden structure consisting of two large post pits aligned with the solstices was excavated at Bulford, Wiltshire, approximately 5 km (3 miles) from Stonehenge. Radiocarbon dated to around 3000 BC, it is contemporary with the earliest phase of activity at Stonehenge (earthwork construction) but predates the erection of the large sarsen stones by roughly 500 years. The structure was far simpler than later Stonehenge phases and likely served as a basic solar marker or observation point for seasonal/ritual gatherings.

Discovery and Excavation Context

The site was investigated by Wessex Archaeology, led by archaeologist Phil Harding, as part of pre-development archaeological work ahead of new Ministry of Defence housing on Salisbury Plain (Army Basing Programme). Initial evaluation trenching took place around a decade earlier; detailed post-excavation analysis, including alignment studies, has been completed recently.

The find lies outside the Stonehenge World Heritage Site but within the broader Neolithic ritual landscape of Salisbury Plain. The same project area also yielded Late Neolithic pits with special deposits, segmented ring ditches (possible hengiform origins later modified into barrows), an Anglo-Saxon cemetery, and modern military features. The solstice-aligned posts form the focus of recent public announcements.

Description of the Structure

The monument comprised:

• Two large post pits, each approximately 0.5–1 m wide and ~1 m deep, positioned about 120 m apart.
• The pits originally held wooden posts estimated at 2–4 m in height (the timber has long since decayed).
• Alignment confirmed by archaeoastronomer Dr Fabio Silva to match summer solstice sunrise and winter solstice sunset as they appeared around 2950 BC.
• A smaller pit aligned with the two main posts contained a rare disc-shaped (discoidal) flint knife, possibly placed deliberately and symbolically linked to the sun.

The structure was a minimal “gunsight” style marker rather than a complex monument. It would have allowed precise observation of the solstices, consistent with early farming communities’ need to track seasonal cycles.

Artefacts and Ritual Evidence

Dozens of smaller pits surrounded the main posts and contained deliberately deposited items (not domestic refuse). Key finds include:

• Grooved Ware pottery in the Woodlands style (distinct from Durrington Walls style found nearby).
• Finely worked flint tools, notably a rare discoidal flint knife with high craftsmanship.
• Animal bones (cattle, pig, red and roe deer, aurochs), antler digging tools.
• Chalk objects: a small spherical “ball” and a large concave “bowl”, both showing carving marks.
• Evidence of careful, sequential refilling of pits suggests ritual or symbolic deposition rather than waste disposal.

Dating and Chronological Context

Radiocarbon dating of organic material (bone, antler) from associated pits places the main activity at approximately 3000–2900 BC. This aligns with Stonehenge Phase 1 (construction of the earthwork enclosure and ditch) but is roughly 500 years earlier than the placement of the large sarsen stones.

Feature / Phase	Approximate Date	Relation to Stonehenge
Bulford post pits & associated activity	~3000–2900 BC	Contemporary with Stonehenge Phase 1 (earthworks)
Stonehenge sarsen stones & trilithons	~2500 BC	Approximately 500 years later
Local ring ditches (Bulford)	Late Neolithic origins; Early Bronze Age modification	Part of the wider evolving ritual landscape

Interpretation

Phil Harding described the discovery as one of the greatest finds of his career, noting that even two post pits reveal much about the thinking, behaviour, and cosmology of people 5,000 years ago. Matt Leivers (Wessex Archaeology) highlighted the long-term continuity of marking significant celestial events in the landscape, describing it as evidence of “the religion of the stone age made manifest in the ground.”

The structure indicates early and accurate interest in solar alignments, almost certainly linked to seasonal concerns of early farming communities. Winter solstice marking may have been particularly important as a ritual response to the “dying” of the light and the hoped-for return of spring. It is plausible (though not proven) that some of the people using the Bulford site were involved in the earliest construction phases at Stonehenge itself.

The find does not overturn existing models of Stonehenge’s development but adds important detail to the pre-stone phases and demonstrates that solstice awareness in the region began earlier and in simpler forms than the iconic monument we see today.

Broader Implications

This discovery reinforces the picture of a rich, evolving Neolithic ritual landscape across Salisbury Plain, with multiple foci of activity. It underscores the value of systematic developer-funded archaeology in revealing sites that would otherwise remain unknown. The preservation of the nearby ring ditches as a Scheduled Ancient Monument is welcome.

Further detailed publication of the full post-excavation analysis by Wessex Archaeology is expected and will provide additional stratigraphic, artefactual, and environmental data.

Further Reading & References

Primary news reports (June 2026)

• The Guardian – “Solstice-aligned 5,000-year-old monument ‘once in a lifetime find’, say archaeologists” (detailed quotes from Phil Harding and Matt Leivers).
• BBC News – “Simpler, older version of Stonehenge found three miles from famous site” (excellent explanation of the alignment work by Dr Fabio Silva).
• Associated Press – Straightforward factual report on the discovery.

Official project sources

• Wessex Archaeology – Discovery led by Phil Harding reveals 5,000-year-old ‘prototype’ for Stonehenge solar alignment.
• Wessex Archaeology – Bulford project page (background on the wider excavations, Neolithic pits, ring ditches, and earlier phases of the same Army Basing Programme work).
• Wessex Archaeology – Army Basing Programme overview (context for the Larkhill and Bulford discoveries).

Additional context

• Phil Harding has worked extensively on the Stonehenge landscape and previous Bulford excavations (2015–2017 phases). Full post-excavation reports from the current work are expected in Wessex Archaeology’s grey literature series and will eventually be available via the Archaeological Data Service (ADS).
• For broader Neolithic landscape studies around Stonehenge and the River Avon, see earlier publications arising from the same Army Basing Programme (e.g. work at Larkhill).

Woodhenge: slightly solstitial?

 

How much sun is actually built into Woodhenge? The short answer is: a real orientation, an unreliable line, and a very short working life. The longer answer is more interesting, and rather less than the signage promises.

The claim, and its pedigree

The standard story is easy to tell. Woodhenge is a Late Neolithic timber monument two miles north-east of Stonehenge: six concentric oval rings of posts (Maud Cunnington's rings A–F), later wrapped in a bank-and-ditch henge with a single causewayed entrance. The long axis of those ovals runs broadly north-east to south-west — the same axis as Stonehenge — and is said to point at the midsummer sunrise in one direction and the midwinter sunset in the other. Hence the tidy textbook line, repeated on the English Heritage signboards, that the monument was “built to align with the summer solstice sunrise.”

The pedigree is respectable. Cunnington herself drew the comparison with Stonehenge in her 1929 excavation report, and noticed that a clear sightline ran between the posts in the solstitial direction. Alexander Thom surveyed the site in the late 1950s and reached broadly the same conclusion. And the most authoritative modern treatment — Clive Ruggles and Amanda Chadburn's Stonehenge: Sighting the Sun (Historic England / Liverpool University Press, 2024) — accepts a deliberate solstitial orientation of the timber rings as part of a genuine mid-third-millennium practice across the Stonehenge landscape.

So this is not a fringe claim to be knocked down. It is a mainstream one, endorsed by the people who have spent their careers being sceptical about exactly this sort of thing. Which makes it worth asking, precisely, what survives once the scepticism is applied evenly — including to Woodhenge.

What “aligned” is allowed to mean

Ruggles and Chadburn draw a distinction that does most of the useful work here. They separate three grades of solar orientation: monuments that are broadly solstitial (precision of the order of 5°, e.g. Maes Howe); monuments that pinpoint the solstice in space (around 0.5°, as the Stonehenge axis does); and monuments that could in principle pinpoint the solstice in time (of the order of 0.01° — the kind of precision Thom once imagined at Kintraw, and which has not survived scrutiny).

The crucial point is that even Stonehenge, the precise end of the credible range, fixes the solstice in space but not in time. For several days either side of the actual solstice the sun rises and sets in sensibly the same place on the horizon, so an alignment of this precision tells you where on the skyline to look, but cannot tell you which day is the solstice. Stonehenge could flag a window of days for ceremony; it could not function as a calendrical instrument. That is the ceiling. Any claim that Woodhenge did better than Stonehenge would need extraordinary evidence, and there is none.

The line that will not stay still

Here is the first genuinely sceptical difficulty, and it is one Cunnington flagged herself: an oval has a long axis, but Woodhenge's posts are irregularly spaced, and the exact azimuth of the axis is correspondingly hard to fix. This is not a quibble. The whole claim rests on a single line drawn through a slightly lopsided arrangement of postholes, and where you put that line depends on choices.

It depends, first, on which rings you trust. Cunnington noticed that rings A and B appeared to share an entrance orientation closer to that of the henge than to the inner rings. Chadburn (2010) and Chadburn and Ruggles (2017) accordingly suggested that perhaps only part of the monument — plausibly rings C, D, E and F — followed the “astronomical” axis at all. The monument may not be a single aligned object so much as an aligned core inside a less-aligned shell.

It depends, second, on whose survey you use. Cunnington and Thom arrived at slightly different azimuths (Ruggles 2006 discusses the discrepancy), and Thom's figure has the additional weakness that he was surveying the modern concrete marker posts, set in the 1920s to indicate the holes, rather than the excavated holes themselves. A survey of markers is a survey of an interpretation. None of this is fatal, but it means the “alignment” comes with error bars wide enough to matter, and that the often-quoted single azimuth is more confident than the underlying geometry warrants.

The entrance, and the awkward dates

If Woodhenge were straightforwardly a solar monument, you would expect its most prominent architectural feature — the henge entrance — to honour the same axis. It does not. The bank-and-ditch entrance sits on a noticeably different alignment from the timber rings, and that mismatch is the loose thread that the recent dating programme has pulled.

Funded by Historic England for Sighting the Sun, and reported in detail by Chadburn and Marshall (Historic England Research Report 94/2024), radiocarbon dating and Bayesian modelling of Cunnington's curated charcoal and of antler picks from the 1970 ditch section produced two distinct events. The timber rings were raised in roughly 2635–2575 cal BC (95% probability) — close in time to the sarsen settings at Stonehenge. The enclosing ditch and bank came significantly later, most probably in the range 2465–2345 cal BC: on the order of two centuries afterwards. (This supersedes the older single-phase reading of about 2300 BC derived from the 1970 Evans and Wainwright trench.)

That gap reframes the entrance problem rather neatly. The henge builders, two centuries on, did not align their entrance on the sun because by then there was probably nothing solar left to align with: the timber rings, the actual carriers of the orientation, would have rotted or been removed. The solstitial sightline, in other words, appears to have been a feature of the early timber monument and to have lapsed long before the earthwork that now defines the site to visitors.

A short working life

This is the conclusion Ruggles and Chadburn themselves emphasise, and it deserves more prominence than it usually gets. The solstitial alignments at both Woodhenge and the Durrington Walls Southern Circle appear to have been short-lived: at Woodhenge the timbers decayed; at Durrington the circle was buried inside a vast henge. Within perhaps a century of construction, the sightlines at both had ceased to be usable for actual observation (Ruggles & Chadburn 2024, 109–111; Chadburn & Marshall, forthcoming).

A monument that can be observed along for a century or so is a real thing, but it is not the timeless solar observatory of popular imagination. It also sits awkwardly with the more romantic processional narratives — crowds walking between aligned monuments at the turning points of the year — if the alignments at the different sites were not all operational at the same time. That is precisely the open question the authors flag rather than answer.

The landscape did some of the aligning

The deepest sceptical point is one the leading researchers raise against themselves. How much of the “alignment” is astronomy, and how much is geography that happened to run the right way?

It has been argued for Stonehenge that the site was chosen partly because natural periglacial striations in the chalk already ran in an approximately solstitial direction (Parker Pearson 2012). The Durrington Southern Circle and the Lark Hill posthole alignment both face down dry valleys that lead off in broadly solstitial directions. If a builder erects a monument facing down a valley whose natural axis is already near the solstice, the resulting “alignment” is real but partly inherited from the ground, not computed from the sky. The intention may have been to monumentalise a place already felt to be charged, with the solar coincidence read as confirmation rather than designed from scratch. Woodhenge, on its low rise above the Avon, is at least a candidate for the same reading. Distinguishing deliberate astronomical design from appropriated topography is genuinely hard, and the honest position is that we have not done it.

It is worth remembering, too, that broadly sun-facing orientation is common in the period without implying precision: the long barrows of a millennium earlier tend to face the sunrise/sun-climbing sectors of the horizon, but under the influence of several factors at once, and well-known solstitial monuments such as Newgrange look like one-offs within multifactorial orientation patterns rather than members of a precise tradition. A north-easterly facing in this landscape is not, by itself, strong evidence of solar intent.

Why the modest version is the credible one

There is a useful discipline that cultural astronomers have enforced since the early 1980s, summarised by the statistician Peter Freeman as “observe everything” and “report all you observe.” The failure mode it guards against is selecting the one feature that fits your hypothesis and quietly ignoring the rest. Ruggles and Chadburn deploy exactly this principle to dismantle a series of recent overreaches: Darvill's reading of a 365¼-day calendar in the numerology of the stones; the 2km “mega-circle” of pits around Durrington Walls, much of which turns out to be natural sinkholes and Bronze Age features; and the two Cursus pits said to mark solstice sunrise and sunset from the Heel Stone, which are merely two among several.

The reason the Woodhenge orientation survives the same treatment is that it is a modest claim. It does not require precision Stonehenge itself lacks, it does not require a calendar, and it does not require selecting one feature from many: the long axis of the rings is the obvious feature, and it does point, broadly, the right way. Strip out the embellishments and what is left is defensible: a deliberate, broadly solstitial orientation of the early timber rings, of modest precision, in a landscape where that practice was briefly current.

What the signage overstates

Set against the literature, the public framing — “built to align with the summer solstice sunrise” — is not wrong so much as over-confident in three specific ways. It implies a single, well-determined azimuth, where the axis is in fact contested and hard to fix. It implies the whole monument is the aligned object, where the dating suggests the alignment belonged to the early timber rings and not to the henge that visitors actually see. And it implies a standing, enduring relationship with the sun, where the working life of the sightline was probably brief. The midwinter-sunset half of the claim is weaker still at the site, since the south-western horizon rises and blocks the setting sun, which is part of why the north-eastern, sunrise direction is the one usually emphasised.

A reasonably sceptical verdict

The defensible position, on the best current evidence, is narrow but real:

• The early timber rings were probably deliberately oriented on the solstitial axis, in step with a genuine practice seen at Stonehenge, Durrington and Lark Hill around 2600–2500 BC.
• The orientation is broadly to moderately precise at best; the exact azimuth is not securely known and the two historic surveys disagree.
• The alignment was short-lived and had probably lapsed by the time the henge earthwork — whose entrance ignores the solar axis — was built two centuries later.
• How much of the orientation reflects astronomy and how much an already-solstitial patch of landscape remains genuinely unresolved.

What does not survive is the observatory: there is no evidence Woodhenge measured the solstice, predicted it, or kept a calendar, and good reason to think it could not have. The fair summary is that Woodhenge was a monument oriented on the solstice, for a while, not a monument that observed it. That is a smaller and more interesting claim than the boards suggest — and, conveniently, the one the strongest research actually supports.

---

Sources

Cunnington, M. E. 1929. Woodhenge: A Description of the Site as Revealed by Excavations. Devizes.
Ruggles, C. L. N. 2006. “Interpreting Solstitial Alignments in Late Neolithic Wessex.” Archaeoastronomy: The Journal of Astronomy in Culture 20: 1–27.
Chadburn, A. and C. L. N. Ruggles. 2017. “Stonehenge World Heritage Property, United Kingdom.” In Heritage Sites of Astronomy and Archaeoastronomy in the Context of the UNESCO World Heritage Convention: Thematic Study no. 2. Paris: ICOMOS, 41–62.
Ruggles, C. L. N. and A. Chadburn. 2024. Stonehenge: Sighting the Sun. Liverpool University Press / Historic England.
Ruggles, C. and A. Chadburn. 2024. “Missing data.” Cosmovisiones / Cosmovisões 5 (1): 99–109. DOI: 10.24215/26840162e007 (open access).
Chadburn, A. and P. Marshall. n.d. Woodhenge, Durrington, Wiltshire: Radiocarbon Dating and Chronological Modelling. Historic England Research Report Series 94/2024.
Ruggles, C., A. Chadburn, M. Leivers and A. Smith. 2021. “A Possible New Sightline in the Stonehenge Landscape.” Journal of Skyscape Archaeology 7 (1): 144–156.
Parker Pearson, M. 2012. Stonehenge: Exploring the Greatest Stone Age Mystery. London: Simon & Schuster.
Magli, G. and J. Belmonte. 2023. “Archaeoastronomy and the alleged ‘Stonehenge calendar’.” Antiquity 97 (393): 745–751.

Neolithic Trackway Mats: A Practical Addition to the Megalith Transport Debate

 

The following is a rewrite of an AI-generated proposal for a novel theory of megalith transport presented here in a more discursive form, with the original references retained. Whether the idea is genuinely novel I cannot say with confidence; apologies in advance if it has been proposed before and I've missed it.

---

The standard question in megalith transport research is how you move a 40-tonne block of stone across a significant distance. The better question — the more tractable one — is where exactly the problem is hardest. Not, in general, on firm dry ground, but at the transitions: river crossings, valley floors, bogs, marshes, and the glacial till that underlies much of lowland Britain. A stone on a sledge on compacted chalk moves fairly efficiently with enough people. The same stone on waterlogged peat doesn't move at all, or sinks.

That bottleneck is the starting point for an idea worth taking seriously: the use of temporary ground mats — woven or bundled from branches, reeds, bracken, hides, or timber — laid ahead of the stone to provide a firm running surface and then leapfrogged forward as progress demanded.

The principle is immediately recognisable to anyone who has watched a large crane or a heavy construction vehicle manoeuvred across soft ground on a modern site. Portable timber or composite mats are standard equipment precisely because wheeled and tracked vehicles, like sledges, will sink into soft substrates under point-loading. The mat distributes that load across a larger footprint, keeps the surface consistent, and can be picked up and moved on. The engineering logic is not complicated. It did not require the twentieth century to discover; it required only the observation that something wide and firm under a heavy thing stops it sinking. Neolithic communities operating on soft terrain, transporting substantial weights, would have arrived at that observation through necessity.

The relevant Neolithic evidence for matting capability is indirect but real. The Sweet Track in the Somerset Levels — dated to around 3800 BC — demonstrates sophisticated knowledge of working with soft, waterlogged ground using locally available timber. What it also demonstrates is that the Neolithic repertoire included permanent trackway construction. The mat theory proposes something simpler and more portable: a temporary surface, reused, not buried, and therefore archaeologically invisible.

That last point is both the theory's plausibility argument and its weakness. The absence of evidence for trackway mats is entirely explicable — organic material used, reused, and never deliberately deposited would leave nothing. But that same explanatory convenience makes the idea difficult to test. One could argue that almost any transport aid made entirely of organic material is equally possible and equally unverifiable, which is a reason for epistemic caution rather than dismissal.

Where the proposal gains traction is in its compatibility with existing models. The experimental work reported by Richards and Whitby (1997) — moving a 40-tonne block on an oak sledge on greased timber rails, requiring around 130 people on a 1:20 slope — assumes stable, reasonably firm ground for the rail system itself. Rails placed on soft ground shift and sink under load. Mats placed beneath and around rails would stabilise them, maintain level running, and prevent the whole system losing its geometry partway through a soft patch. Parry (2000), reviewing the mechanics of sleds on lubricated timber rails, notes the sensitivity of the system to ground conditions; mats address exactly that sensitivity. The systems are complementary rather than competing.

The reassessment of roller evidence is relevant context here. Harris (2018a) reviews the history of the roller hypothesis — four centuries old, still dominant in public understanding — and finds it weakly supported by the experimental and ethnographic literature. The images from Nias and Sumba, long cited as illustrations of logs being used as rollers, are now read by some researchers as showing timber slipways instead (Harris 2018b). If the field is already moving towards static slipway systems, the integration of ground matting with those systems becomes more rather than less plausible.

The lipid residue evidence from Durrington Walls (Shillito 2019) is relevant in a different way: the fatty residues interpreted as pig fat in Grooved Ware vessels may represent the grease applied to sledge runners and rails. The same organisational capacity that could render down and deploy animal fat as a lubricant could equally construct and manage a system of portable mats. These are complementary elements of what was evidently a carefully planned and resourced transport operation.

What would strengthen or weaken this proposal? Experimental archaeology testing mat-assisted transport on reconstructed soft-ground conditions would be the obvious next step — how many people, how much material, what realistic gain on a waterlogged valley floor? Harris's UCL doctoral thesis (2020) provides a quantitative framework for labour estimates in prehistoric monument construction; integrating mat construction and deployment into that kind of model would give the idea more analytical grip.

The proposal should also be seen against the broader backdrop of the Le Plasker evidence from Brittany (Blanchard et al. 2025) and the Bougon experiments, which collectively show that Neolithic communities across Atlantic Europe were operating with greater organisational sophistication than the "drag it with ropes" model implies. Long-distance deliberate sourcing, communal labour mobilisation, and phased logistical planning are not in question. The means of managing difficult terrain is the gap. Mats are a low-technology, high-practicality answer to that gap — precisely the kind of answer that archaeological absence can neither confirm nor rule out, but that engineering logic makes hard to dismiss.

---

Key References

• Richards, J. and Whitby, M. (1997) 'The engineering of Stonehenge', in B. Cunliffe and C. Renfrew (eds), Science and Stonehenge, Proceedings of the British Academy 92. Oxford, 231–56. Link

• Parry, R. H. G. (2000) 'Megalith mechanics', Proceedings of the Institution of Civil Engineers — Civil Engineering 138(4): 183–192. https://doi.org/10.1680/cien.2000.138.4.183

• Harris, B. (2018a) 'Roll Me a Great Stone: A Brief Historiography of Megalithic Construction and the Genesis of the Roller Hypothesis', Oxford Journal of Archaeology 37: 267–281. https://doi.org/10.1111/ojoa.12142

• Harris, B. (2018b) 'Moving megaliths: time to park the rollers', British Archaeology, December 2018. Link

• Harris, B. (2020) Landscapes of labour: a quantitative study of earth-moving and stone-shifting in prehistoric northern Wessex. Doctoral thesis, UCL. Download (286MB)

• Shillito, L.-M. (2019) 'Building Stonehenge? An alternative interpretation of lipid residues in Neolithic Grooved Ware from Durrington Walls', Antiquity 93(370): 1052–1060. https://doi.org/10.15184/aqy.2019.62

• Blanchard, A. et al. (2025) 'Le Plasker in Plouharnel (fifth millennium cal BC): a newly discovered section of the megalithic complex of Carnac', Antiquity 99(406): 915–934. https://doi.org/10.15184/aqy.2025.10123

Paul Whitewick on the Altar Stone and Glacial Transport

How Did They Build Stonehenge? Moving 40-Tonne Stones: Stonehenge Access All Areas, Ep 4

Bluestone Transport, were oxen used?

Julian Richards discusses whether Oxen were used to move the stones, for the full story look out for his video on https://www.youtube.com/@julianrichards1483

I completely agree that “pure people power” (plus rollers, levers and a lot of coordinated effort) was almost certainly the main story for the big sarsens.

That said, the recent isotopic work on the young female cow tooth from Stonehenge’s ditch has added a really interesting extra strand for the earlier bluestone phase. The strontium signatures point to her having maybe come from the Preseli area in Wales — exactly the same source region as the bluestones — and she was there around the right time. The lead spike in her enamel is open to more than one interpretation; it could be calving stress, but it’s also the sort of signature that can appear with other physiological demands.

On sarsen.org I’ve floated a very tentative, speculative thought that perhaps a few cattle played a supporting role in the broader logistics of getting the (much smaller and lighter) bluestones to the site — not teams of oxen yoked up dragging multi-tonne stones for hundreds of miles, but maybe helping with shorter hauls, local manoeuvring on site, or simply travelling with the human group. The wider European evidence for some early cattle traction in megalithic contexts makes it worth at least considering, even if a major new British osteological study suggests heavy, routine traction wasn’t common in the Neolithic here.

None of this contradicts the central importance of human labour — it just adds a possible extra tool in the Neolithic toolkit for certain tasks or phases. One tooth obviously doesn’t rewrite the whole story, and I’m very happy to be wrong!

I’ve written more about the tooth and the thinking here: https://www.sarsen.org/2026/02/a-speculative-hypothesis-neolithic.html and the follow-up here: https://www.sarsen.org/2026/03/the-ox-that-moved-mountain-or-didnt.html