The latest on Stonehenge Sarsen Sources - July 2025

In 2020, a landmark geochemical study by Nash et al. proposed that most Stonehenge’s sarsen stones, came from West Woods, about 25 km to the north in the Marlborough Downs. Using portable X-ray fluorescence (pXRF) and Bayesian PCA of trace elements such as zirconium and niobium, they tied 50 of the 52 stones to a geochemical signature matching this source. Among the most carefully matched was Stone #58, part of a central trilithon, sampled in 1958 via the now-famous “Phillips Core.”

The latest of a series of papers in the journal Archaeometry, debating the interpretation of geochemical data and its implications for the monument has just been published. This exchange highlights the nuances of scientific provenancing: data can be reanalysed differently, leading to conflicting conclusions. To clarify the sequence of claims and counterclaims chronologically, based on the provided documents and the broader context.

The Original Claim: Nash et al. (2020)

• In a landmark study, David Nash and colleagues used portable X-ray fluorescence (pXRF) to analyse 52 sarsen stones at Stonehenge, taking multiple readings per stone. For Stone #58 they also analysed the sample of it, the pXRF matched the stone to 49 other stones so they could use the sample as being representative of them.
• They also sampled sarsen outcrops from 20 potential source areas across southern England.
• Using Bayesian principal component analysis (BPCA) on immobile trace elements (e.g., normalised to zirconium to account for silica dilution during rock formation), they concluded that 50 of the 52 stones share a common chemistry, likely originating from a single source: West Woods.  A variant of PCA that incorporates prior information, BPCA is beneficial in geological provenancing because it handles missing data and reduces dimensionality while preserving meaningful geochemical trends.
• Two outlier stones had different signatures, suggesting multiple sources overall.
• This implied organised transport from a concentrated area rather than scattered local boulders.

The Challenge: Hancock et al. (2024)

• R.G.V. Hancock and team reinterpreted the same dataset from Nash et al. (2020), focusing narrowly on Stone #58.
• They used absolute elemental concentrations (not normalised) and selected different discriminating elements (e.g., Ga, Fe₂O₃, Hf).
• Their analysis suggested the data do not definitively match Stone #58 to West Woods. Instead, they argued: “as a minimum, the Wiltshire sites of Clatford Bottom and Piggledene with West Woods a distant third, appear as potential sources of stone #58. None of these locations, however, has elemental concentrations from one up to three samples that would geochemically merge with those of the Phillips core samples.”
• They criticised Nash's choice of elements (e.g., Hf's correlation with Zr making it less useful) and highlighted issues like detection limits and data quality variability in the original dataset.
• Key point: They aimed to "replicate" Nash's findings but ended up questioning the certainty, emphasising that sarsen geochemistry can vary even within a single outcrop due to formation processes.

The Comment: Nash and Ciborowski (2025)

• Nash and T.J.R. Ciborowski respond directly to Hancock et al. (2024). They identify three main "problems" with Hancock's approach:
1. Use of absolute concentrations: Sarsens form through silica cementation, which dilutes trace elements. Nash argues for normalising data (e.g., to Zr, an immobile element) to correct for this; Hancock's raw data ignores dilution effects, potentially leading to misleading comparisons.
2. Selection of discriminating elements: Hancock used elements like Fe and Ga, which can be mobile during weathering or silicification, making them unreliable for sourcing. Nash prefers immobile ones (e.g., Zr, Nb, Th) that better reflect the parent sediment's signature.
3. Single-sample comparisons: Hancock compared single samples from outcrops to the Phillips' Core, ignoring intra-site variability. Nash used multi-sample ranges and statistical envelopes to account for heterogeneity in silcretes.
• They defend West Woods as the probable source, noting that Hancock's reanalysis overlooks petrological (rock structure) evidence and wider silcrete literature. They include figures showing how dilution affects concentrations and petrographic images of zircon grains.

The Response: Hancock et al. (2025)

• In the “Archaeometry - 2025 - Hancock - On Sourcing Stonehenge Sarsen Stone #58...", Hancock and colleagues counter Nash's criticisms.
• On absolute concentrations: They argue silica dilution has minimal practical impact in this dataset, as most samples are 97–100% SiO₂. They point to examples where lower-SiO₂ samples (silicified sandstones) show extreme variations unrelated to dilution.
• On element selection: Minor differences exist (e.g., Nash used Rb, Th, U; Hancock used Ga, Fe₂O₃), but Hancock claims these don't explain the divergent conclusions. They question Nash's use of Hf (due to its correlation with Zr) and Fe₂O₃ (not significantly different across high-SiO₂ samples).
• On single samples: Hancock clarifies they didn't rely on single samples for their own analysis but used Nash's provided data. They note anomalies in some Nash samples (e.g., from Monkton Down and Totterdown Wood) that don't fit neatly.
• They reiterate that the source of Stone #58 remains uncertain and call for re-examination of Nash's claims. They also touch on potential glacial transport of sarsens, adding another layer to the debate.

Further Support from Recent Studies: Harding et al. (2025)

A recent paper by Harding et al. (published online January 2025 in Proceedings of the Prehistoric Society) applies the same pXRF geochemical techniques to two outlier sarsens in the Stonehenge landscape: the Cuckoo Stone and Tor Stone, located on opposite banks of the River Avon.

Key findings:

• Both stones exhibit geochemical signatures statistically indistinguishable from the majority of Stonehenge sarsens, pointing to West Woods as their likely origin.
• They were probably transported and erected in the early Late Neolithic (early 3rd millennium BCE), contemporary with Stonehenge Phase 1 and about 400-500 years before the main sarsen structures.
• Visibility analysis shows the stones were intervisible and positioned to form a "planned portal" across the River Avon, integrating into the broader Neolithic landscape.

This study, co-authored by Nash and Ciborowski, reinforces the West Woods sourcing hypothesis and extends the timeline of sarsen movement into the Stonehenge area. It also counters ideas of local sarsen abundance by arguing geological conditions on Salisbury Plain were unsuitable for forming large boulders like those at West Woods.

• Unresolved Issues: Methodological debates persist, but glacial claims are a red herring. Future work might use advanced zircon dating or quarry surveys to confirm West Woods.
• Implications for Stonehenge: Human transport from West Woods implies Neolithic ingenuity. Dismissing glaciation strengthens this narrative, avoiding unproven shortcuts.

 

What to Make of It All?

This back-and-forth is classic academic discourse: science thrives on scrutiny, and geochemical sourcing is inherently tricky for materials like sarsen, which form in variable groundwater environments. Here's a hopefully balanced take:

• Strengths of Nash's Position: Their original study is comprehensive, using non-destructive methods on the actual monument and statistical rigour to handle variability. Normalising for dilution is standard in geochemistry, and West Woods fits archaeologically (it's a dense sarsen field, suitable for quarrying multiple stones).
• Strengths of Hancock's Position: Reanalysing existing data is valid and cost-effective. They highlight real issues like data quality (e.g., detection limits) and question over-reliance on certain elements. If dilution effects are overstated, absolute concentrations might suffice for this specific dataset.
• Points of Agreement: Both sides acknowledge sarsen heterogeneity and the limitations of single-sample data. They agree the dataset is valuable but interpret its nuances differently.
• Unresolved Issues: The debate centres on methodology rather than new evidence. Stone #58 might be from West Woods, but Hancock's work introduces doubt. Were all sarsens from one source? Future work could involve more sampling, advanced techniques (e.g., laser ablation ICP-MS for precise zircon analysis), or integrating geophysical surveys of potential quarries.
• Implications for Stonehenge: If West Woods is confirmed, it suggests sophisticated Neolithic logistics over 25 km. If not, it opens doors to multi-source models, perhaps involving local boulders. This doesn't upend Stonehenge's story but refines it, science is iterative.

The geochemical back-and-forth remains a healthy academic exercise, with Nash's normalised, multi-sample approach appearing more robust for silcretes, while Hancock highlights valid data limitations. West Woods still leads as the probable source for most sarsens, supported by archaeological fit (dense boulder fields for quarrying).

However, Hancock's invocation of glacial transport based on John (2009), positing that ice sheets might have naturally deposited sarsens near Stonehenge, reducing the need for human effort, warrants particular scepticism. This idea, echoed in fringe theories for Stonehenge's bluestones (smaller Welsh stones), lacks empirical support and contradicts established geology. During the Last Glacial Maximum (LGM, ~27–19 ka ago), the British-Irish Ice Sheet (BIIS) covered Scotland, Ireland, most of Wales, and northern England, but stopped well north of Wiltshire's Marlborough Downs and Salisbury Plain. Southern England was periglacial, cold tundra with permafrost, but ice-free, as glaciers flowed from highlands without extending south. No glacial erratics (transported boulders) of any kind, let alone sarsens, have been found on Salisbury Plain.

Recent studies (2024–2025) further debunk glacial theories for Stonehenge stones. A July 2025 analysis of the "Newall boulder" (a bluestone fragment) found "no evidence to support an interpretation that it is a glacial erratic. Nash's dismissal, that glacial ideas have "no basis in fact”, is thus justified; invoking them risks pseudoscience, as sarsens form from local Palaeogene sands, not distant bedrock.

• Unresolved Issues: Methodological debates persist, but glacial claims are a red herring. Future work might use advanced zircon dating or quarry surveys to confirm West Woods.
• Implications for Stonehenge: Human transport from West Woods implies Neolithic ingenuity. Dismissing glaciation strengthens this narrative, avoiding unproven shortcuts.

In summary, the debate refines provenancing without resolution yet, but glacial transport is soundly refuted.

References: 

Ciborowski, T. J. R., D. J. Nash, T. Darvill, et al. 2024. “Local and Exotic Sources of Sarsen Debitage at Stonehenge Revealed by Geochemical Provenancing.” Journal of Archaeological Science: Reports 53: 104406. https://doi.org/10.1016/j.jasrep.2024.104406.

Hancock, R. G. V., M. P. Gorton, W. C. Mahaney, S. Aufreiter, and K. Michelaki. 2024. “Stonehenge Revisited: A Geochemical Approach to Interpreting the Geographical Source of Sarcen Stone #58.” Archaeometry 67, no. 1: 1–19. https://doi.org/10.1111/arcm.12999.

Hancock, R. G. V., M. P. Gorton, W. C. Mahaney, S. Aufreiter, and K. Michelaki. 2025. “ On Sourcing Stonehenge Sarsen Stone #58: A Response to Nash and Ciborowski's Comments.” Archaeometry 1–3. https://doi.org/10.1111/arcm.70028.

Harding, P. et al. (2024) ‘Earliest Movement of Sarsen Into the Stonehenge Landscape: New Insights from Geochemical and Visibility Analysis of the Cuckoo Stone and Tor Stone’, Proceedings of the Prehistoric Society, 90, pp. 229–251. doi:10.1017/ppr.2024.13.

John, B. S., & Jackson, L. Jr. (2009). Stonehenge's Mysterious Stones. Earth, 54(1), 36–42 https://www.researchgate.net/publication/270162075_Stonehenge's_mysterious_stones

Nash, D. J., and T. J. R. Ciborowski. 2025. “Comment on: Stonehenge Revisited: A Geochemical Approach to Interpreting the Geographical Source of Sarsen Stone #58.” Archaeometry 1–14. https://onlinelibrary.wiley.com/doi/10.1111/arcm.13105.

Nash, D. J., T. J. R. Ciborowski, J. S. Ullyott, et al. 2020. “Origins of the Sarsen Megaliths at Stonehenge.” Science Advances 6, no. 31: 1–8. https://doi.org/10.1126/sciadv.abc0133.