A monument born from climate change

Geology and archaeology have recently come together at Stonehenge in a spectacular fashion. It now appears that the 5-m-long Altar Stone at the centre of this Neolithic stone circle on the Salisbury Plain of Wiltshire, southern England, was sourced from some 700 km away, from Devonian rocks of northeast Scotland (Clarke et al., 2024). In sharp contrast, the silicified sandstone sarsens used to build most of Stonehenge (Fig. 1) came from only 30 km away to the north, from West Woods, Marlborough in Wiltshire (Nash et al., 2020, 2021; Worsley, 2021).  

What is the geological history of the Stonehenge sarsens? When and how were these silica-cemented sandstones formed? Here I explore the links between geology, climate history and archaeology by proposing that the West Woods sarsens plausibly correlate with silcretes found across southern England and northern France. I argue that they formed during a warm episode in Earth’s history some 55.8 million years ago (Ma) known as the Paleocene-Eocene Thermal Maximum (PETM) that is, in turn, linked to the history of the early Iceland mantle plume.  

Figure 1 | Plans of Stonehenge showing (A) the area of the monument enclosed by earthworks and (B) detail of the stone circle. Image credit: © Nash et al. (2021) PLoS ONE 16(8): e0254760; https://doi.org/10.1371/journal.pone.0254760. Published under a CC BY 4.0 licence (https://creativecommons.org/licenses/by/4.0).

Beyond West Woods 

The sarsens at West Woods, north of Stonehenge, appear to be resistant remnants of a former Paleogene cover that was laid unconformably atop Cretaceous chalk over a wider area than is now preserved (Worsley, 2019). Estimates indicate that several hundred metres of chalk were removed by pre-Thanetian erosion in the region, implying that the unconformity is a major regional feature (Gale & Lovell, 2018). The scattered sarsen remnants lie near the present-day western erosional edge of the Paleogene of the London Basin (Fig. 2).

Figure 2 | Outcrops of Paleogene strata in the London Basin. The scattered sarsen remnants at West Woods, Marlborough lie near the present-day western erosional edge of the Paleogene of the London Basin, and correlate along strike with the Collier’s End puddingstone (approximate locations shown as blue stars). The Bagshot Formation, a series of upper Eocene sands and clays deposited in shallow water, some fresh, some marine; the Thames Group, Eocene clays, some silty, sandy or gravelly, deposited in a marine shoreface to outer marine shelf; the Thanet Formation, late Paleocene fine-grained sand, silt and clays deposited on an inner to outer marine shelf; the Lambeth Group, Late Paleocene to Early Eocene silty or sandy clays and occasional sandstone and conglomerates deposited in fluvial, estuarine, lagoonal environment. Also illustrated is the Neogene Crag Group, Pliocene to Pleistocene marine and estuarine sands, gravels, silts and clays deposited in a shallow tidal marine embayment on the western margins of the North Sea Basin. (1996TextViewer BGS©UKRI Creative Commons Attribution-ShareAlike 3.0 Unported Licence, CC BY-SA; https://webapps.bgs.ac.uk/Memoirs/docs/B01360.html)

Along strike to the north-east, on the northern boundary of the London Basin, lie the Hertfordshire Puddingstones, and notably a conglomeratic outlier of Hertfordshire Puddingstone preserved in situ and exposed at Collier’s End. Stratigraphic analyses (Lovell & Tubb, 2006) and mapping (Lovell et al., 2023) indicate that the Collier’s End puddingstone lies at or close to the boundary of the Paleocene and Eocene (Fig. 3).

Based on the strength of the along-strike correlation, the West Woods sarsens and Collier’s End puddingstone are plausibly silcretes of common origin, formed at a similar time during the PETM. 

the PETM lasted for over 100,000 years, ample time for silcrete formation

Overlying the Collier’s End puddingstone is the early Eocene London Clay Formation, which was regionally extensive. Some may question whether there was enough time for silcretes to form – that is, be deposited and cemented – during the PETM, before the sea level transgression that laid down the London Clay. However, the PETM lasted for over 100,000 years, ample time for silcrete formation given the enhanced solubility of silica at high temperatures. Additionally, early Eocene deposits overlying the Collier’s End puddingstone (the Blackheath Beds, Fig. 3) contain reworked puddingstone and sarsen, presumably derived from silcretes formed during the PETM and then exposed to erosion in the early Eocene prior to deposition of the London Clay. 

Figure 3 | Depositional sequences of the early Paleogene succession of southeast England. The Collier’s End Pebble Bed (CEPB) contains concretions of Hertfordshire Puddingstone and is interpreted as late Paleocene in age, formed at or close to the Paleocene-Eocene Thermal Maximum (PETM). Stratigraphy is based on the interpretation by Knox (1996). (Figure reproduced from Lovell (2016) Proceedings of the Geologists’ Association172, 301-310; http://dx.doi.org/10.1016/j.pgeola.2015.08.004).

The Silcrete Story

Silicified sandstone blocks are found across much of southern England and northern France. These silcretes contain varying amounts of sandstone and conglomerate. The sandy end-member facies is known as sarsen, and the conglomerate-rich end-member facies is referred to as puddingstone (named after its resemblance to plum pudding).

The sarsens and puddingstones were deposited in the Anglo-Paris Basin, which encompasses the London and Hampshire basins, as well as the Paris Basin. The Anglo-Paris Basin formed around 100-66 million years ago during the Late Cretaceous, when sea levels were high and much of this region was submerged beneath a shallow sea creating extensive chalk deposits, however the basin continued to evolve during the Paleogene. 

© Science Photo Library

The PETM landscape 

What can we learn from the composition of the Stonehenge sarsens? Geochemical data reported by Nash and colleagues (2021) show that these sarsens are chemically pure, composed of over 99% silica, with minor quantities of rounded flint pebbles. The purity of the unusually quartz-rich sands suggests they came from an intensely weathered source area of Mesozoic rocks, while the flint was likely derived from the upper Cretaceous chalk. I propose that both the sands and pebbles were plausibly swept south-eastward by PETM floods toward the shore of the Paleogene North Sea, but into what landscape were these sediments carried?  

The Collier’s End puddingstones provide rare insight into the PETM landscape

The Collier’s End puddingstones provide rare insight into the PETM landscape (Lovell et al., 2023). These pebble beds cut into mottled and red-stained mudstones (in the lowermost part of the Reading Formation, part of the Lambeth Group; Fig. 3) that are interpreted as mudflats (Fig. 4). I propose that the Collier’s End puddingstone and West Woods sarsens both formed in a hot Paleogene landscape, in a flat and muddy coastal or near-coastal area trending roughly northeast-southwest across what is now southern England. They were deposited by floods that carried in many flint pebbles in the east, as well as quartz-rich sand, then silicified under the anomalous heat of the PETM, which enhanced the solubility of silica. If the West Woods sarsens and Collier’s End puddingstones are indeed silcretes of common origin, we should expect to find pebbly sarsens and sandy puddingstones – an expectation confirmed by observations of sand-rich puddingstones at Collier’s End (Fig. 5).

Figure 4 | The Collier’s End Pebble Bed. The CEPB lies on an erosional surface of red-stained mottled mudstones (in the lowermost part of the Reading Formation) at Plashes Farm, Hertfordshire. Scale 50 cm. (Image credit: Lovell et al. 2023. Proceedings of the Geologists’ Association 134, (5-6), 503-516; https://doi.org/10.1016/j.pgeola.2023.05.002).

Regional extent 

How extensive was this postulated episode of silcrete formation at the time of the PETM? Silcretes are found in southern England beyond Wiltshire and Hertfordshire, but the formation dates are not well constrained. For example, Isaac (1983) suggests that residual silcrete deposits in east Devon (to the west of the London Basin) date from the early Paleogene, while Bristow (1993) proposes that sarsens from south Somerset and surrounding the Hampshire Basin are also of Paleogene age. Puddingstones at Portesham, Dorset (Lovell, 2016) appear to be part of the regional development of Paleogene silcrete described by Isaac and Bristow. 

Evidence of similar geological processes has also been found in northern France where silcretes in the Paris Basin are stratigraphically constrained to the Paleocene-Eocene boundary (Quesnel et al., 2014) supporting the theory of a regional silcrete event driven by the extreme heat of the PETM.  However, there is also a proven later episode of silcrete formation in France. Green (2016) assigns an early Eocene age to puddingstones used to build Roman querns – a circular tool used to grind grain – found in northern France. Rather to his surprise, Green reports that the querns contain “…natural casts of fossil brackish and marine molluscs…” that are identified as lower Ypresian and interpreted as equivalent to the Abbey Wood facies of the Blackheath Beds, Harwich Formation, in southern England (Fig. 3).

It is possible then that an early phase of regional silicification occurred across the Anglo-Paris Basin linked to the PETM, followed by a second phase of apparently more local silicification perhaps linked to an early Eocene hyperthermal event circa 54 Ma. There may have been other phases of Cenozoic silcrete formation. A Neogene or Quaternary age of formation is proposed for scattered silcretes lying south of the London Basin (Ullyott et al., 2004). That much later date does not invalidate the Paleogene linkage for Stonehenge proposed here.  

Figure 5 | A sand-rich puddingstone found in-place at Collier’s End. The fine quartz sand was originally in place on the surface of a concretion of Hertfordshire puddingstone, the matrix of which is silicified fine quartz sand. The concretion was recovered from a temporary expose of eh Collier’s End Pebble Bd in a recent road cut. Coin is 30mm diameter. (Image credit: © Bryan Lovell)

Pulsating plume 

What is the justification for the proposed correlation of various distinctive Paleogene events across the Anglo-Paris Basin? Over the last thirty years, an integrating hypothesis concerning the Paleogene of northwest Europe has gained strength: thermal pulses in the Iceland mantle plume spread out in the low-velocity zone of the upper mantle causing widescale, episodic uplift and subsidence (so called dynamic topography) and exert a first-order control on the development of depositional sequences in the Paleogene of the North Sea (White & Lovell, 1997).  

In this framework, regional uplift of the Cretaceous sea floor in the early Paleogene was caused by a major hot pulse in the Iceland plume that created both transitory uplift and long-term uplift caused by magmatic underplating of the crust (Brodie & White, 1994; White & Lovell, 1997). This early Paleogene episode of uplift, which exposed the chalk sea floor to erosion, was followed by additional pulses of uplift that controlled regional sea level, coastlines and depositional sequences in the Paleogene of the North Sea and offshore Scotland, as well as the Anglo-Paris Basin (White & Lovell, 1997; Knox, 1996; Gale & Lovell, 2020).  

Message from the past 

An important message from the PETM may be read in a new geological time-trail in the Evolution Garden of London’s Natural History Museum, which opened in the summer of 2024. Near the younger end of the trail are specimens of puddingstone from Collier’s End, Hertfordshire, sarsen from West Woods, Wiltshire, and columnar basalt from Iceland.  These three different types of rock are linked by the Iceland mantle plume. A hot pulse in the plume at c.56 Ma appears to have triggered the PETM. That hot pulse in the plume led to major and rapid discharge of carbon to the atmosphere. The resulting very warm climate led to formation of the main body of silcretes now found in the Anglo-Paris Basin.

The Evolution Garden at the Natural History Museum, London, UK (© Marissa Lo).

Yo-yo tectonics associated with the pulsating plume led to episodic supply of sand from Paleogene Scotland to the North Sea Basin and to episodic erosion of near-coastal areas, as evidenced by a series of now-buried Paleogene landscapes along the continental margin of northwest Europe (Conway-Jones & White, 2022). Additionally, significant volumes of greenhouse gases released by volcanism associated with pulsed plume activity likely triggered hyperthermal anomalies and episodic changes in global climate, including the PETM (Conway-Jones & White, 2022).

As an aside, the plume hypothesis for uplift of southern England is much preferred to the previous widespread notion that uplift was the result of crustal shortening associated with Alpine or Pyrenean tectonics. According to Gale and Lovell (2018), that far-field crustal shortening would have generated a maximum of 25 m of regional uplift of the chalk. That might bring the floor of the chalk sea above sea level, but “…would provide only a fraction of the c.125 m uplift above sea level required to produce the denudation of c.500 m thickness of the chalk surface that took place across the Chilterns–East Anglia dome before transgression of the Paleogene sea.”

Lessons from ancient climate change 

Silcretes have proved useful. Sarsens from West Woods were moved a short distance to build Stonehenge (Nash et al., 2020, 2021; Worsley, 2021), while puddingstones in Hertfordshire were quarried by the Romans for use as querns to grind corn (Green, 2016; Jones & Green, 2023; Lovell & Tubb, 2006). But these silcretes offer more than just historical significance – they hold clues to an ancient episode of abrupt climate change.

silcretes offer more than just historical significance – they hold clues to an ancient episode of abrupt climate change

During the PETM, carbon was released at an alarming rate due to volcanic activity, leading to a sharp rise in global temperatures. Earth eventually recovered and the silcretes of Stonehenge and Hertfordshire serve as a reminder of the resilience of our planet. However, the rate of carbon release during the PETM pales in comparison to today’s human-driven emissions that owe nothing to volcanoes. A stark lesson from the silcretes of the PETM is that our planet will adapt successfully to a sudden change in climate. The survival of human societies, however, is far less certain.

Author

Dr Bryan Lovell
Emeritus Senior Researcher in Earth Sciences, University of Cambridge, UK, and President of the Geological Society of London 2010–2012.

Editor’s note:

Very sadly, Bryan Lovell passed away before signing off the final version of this article. The editorial team extends our sincere thanks to Professor Andrew Gale at the University of Portsmouth, UK who kindly proofread the edited article to ensure we had retained accuracy with our edits. If any inaccuracies remain, these are entirely the fault of the editorial team and we offer our apologies. Please contact us at geoscientist@geolsoc.org.uk to flag any concerns. An obituary celebrating Bryan’s wonderful professional life is available here.

Further reading

• Bristow, C. M. (1993) Silcrete duricrusts west of the Bovey Basin. Proceedings of the Ussher Society, 8, 177-180. 

• Brodie, J. & White, N. (1994) Sedimentary basin inversion by igneous underplating, North-west European continental shelf. Geology, 22, 147-150; https://doi.org/10.1130/0091-7613(1994)022<0147:SBICBI>2.3.CO;2  

• Clarke, A. J. I. et al. (2024) A Scottish provenance for the Altar Stone of Stonehenge. Nature, 632, 570-575; https://doi.org/10.1038/s41586-024-07652-1 

• Conway‐Jones, B. W. & White, N. J. (2022) Paleogene buried landscapes and climatic aberrations triggered by mantle plume activity. Earth and Planetary Science Letters, 593, 117644; https://doi.org/10.1016/j.epsl.2022.117644 

• Gale, A. S. & Lovell, B. (2018) The Cretaceous-Paleogene unconformity in England: Uplift and erosion related to the Iceland mantle plume. Proceedings of the Geologists’ Association, 129, 421-43; https://doi.org/10.1016/j.pgeola.2017.04.002  

• Gale, A. S. & Lovell, B. (2020) Control of the Paleogene sedimentary record of the Anglo-Paris Basin by both the Iceland mantle plume and the Massif Central hotspot. Proceedings of the Geologists’ Association, 131, 652-666; https://doi.org/10.1016/j.pgeola.2020.07.001 

• Green, C. (2016) The exploitation of silcretes (sarsen and puddingstone) in England and Normandy since Stonehenge. Proceedings of the Geologists’ Association, 127, 349-358; https://doi.org/10.1016/j.pgeola.2015.12.002 

• Green, C. et al. (2016) Discovery of a second Roman quarry in Hertfordshire for manufacture of querns from Paleogene Hertfordshire Puddingstone siliceous concretions. Proceedings of the Geologists’ Association, 127, 359-362; https://doi.org/10.1016/j.pgeola.2015.12.003 

• Isaac, K. P. (1983) Tertiary lateritic weathering in Devon, England, and the Palaeogene continental environment of South West England. Proceedings of the Geologists’ Association, 94, 105-114; https://doi.org/10.1016/S0016-7878(83)80002-9 

• Jones, M. & Green, C. (2023) A Romano-British puddingstone quarry in the Hertfordshire Chilterns. Proceedings of the Geologists’ Association, 134, 86-106; https://doi.org/10.1016/j.pgeola.2022.11.001 

• Knox, R. W. O’B. (1996) Tectonic controls on sequence development in the Palaeocene and earliest Eocene of southeast England: Implications for North Sea stratigraphy. In Hesselbo, S. P., & Parkinson, D. N. (Eds.), Sequence Stratigraphy in British Geology. Geological Society of London Special Publication 103, 209-230; https://doi.org/10.1144/GSL.SP.1996.103.01.12 

• Lavoisier, A. (1789) Observations générales sur les couches horizontales, qui ont été déposé par la mer, et sur les conséquences qu’on peut tirer de leurs dispositions, relativement à l’ancienneté du globe terrestre. Mémoire de l’Académie Royale des Sciences, 351-371. 

• Lovell, J. P. B. (1977) The British Isles through Geological Time: A Northward Drift. George Allen & Unwin, London. 

• Lovell, B. (2016) Paleogene silcretes and the 55 Ma global warming event: Is there proof in the Hertfordshire Puddingstone? Proceedings of the Geologists’ Association, 127, 301-310; https://doi.org/10.1016/j.pgeola.2015.08.004 

• Lovell, B. & Tubb, J. (2006) Ancient quarrying of rare in situ Palaeogene Hertfordshire Puddingstone. Mercian Geologist, 16, 185-189. 

• Lovell, B. et al. (2023) New exposure of the Cretaceous–Paleogene unconformity and Paleocene-Eocene pebble bed in the Paleogene outlier at Collier’s End, Hertfordshire, UK. Proceedings of the Geologists’ Association, 134, 503-516; https://doi.org/10.1016/j.pgeola.2023.05.002  

• Nash, D. J. et al. (2020) Origins of the sarsen megaliths at Stonehenge. Science Advances, 6, 1-8; https://doi.org/10.1126/sciadv.abc0133  

• Nash, D. J. et al. (2021) Petrological and geochemical characterisation of the sarsen stones at Stonehenge. PLOS ONE; https://doi.org/10.1371/journal.pone.0254760 

• Quesnel, F. et al. (2014) Reconstructing the P/E continental paleosurface in and around the Paris and adjacent basins: New insights for paleogeographic, geodynamic and Paris Basin (abstract). Geologists’ Association – Geological Society of London – Society of Antiquaries conference on Puddingstones and related silcretes in the Anglo-Paris Basin, May 2014. 

• Sumbler, M. G. (1996) British Regional Geology: London and the Thames Valley. London, HMSO for the British Geological Survey. 

• Ullyott, J. S. et al. (2004) Distribution, petrology and mode of development of silcretes (sarsens and puddingstones) on the eastern South Downs, UK. Earth Surface Processes and Landforms, 29, 1509–1539; https://doi.org/10.1002/esp.1136 

• White, N. & Lovell, B. (1997) Measuring the pulse of a plume with the sedimentary record. Nature, 387, 888-891; https://doi.org/10.1038/43151  

• Worsley, P. (2019) Geology of the Clatford Bottom catchment and its sarsen stones on the Marlborough Downs. Mercian Geologist, 19, 242-252. 

• Worsley, P. (2021) The Sarsens of the West Woods, Marlborough Downs and Stonehenge. Mercian Geologist, 20, 120-129. 

Share this article