I was pleased to find on Ancestry.com a couple of photographs of Maud Cunnington, which can be used as alternatives to the widely known, slightly unflattering, one.
Introduction
The transport of Stonehenge’s bluestones—igneous rocks sourced from the Mynydd Preseli region of west Wales—remain one of the most debated questions in British prehistory. Two major explanations have dominated: transport by Pleistocene glaciation or deliberate human movement during the Neolithic. The glacial hypothesis, once influential, suggested that Irish Sea ice carried the stones into southern England during the Devensian glaciation. However, accumulating geomorphological evidence increasingly contradicts this scenario.
A key
contribution comes from Kidson et al. (1978), whose investigation of the Burtle
Beds in the Somerset Levels provides robust evidence for intact interglacial
marine deposits. This finding strongly challenges the idea that Devensian ice
reached the lowlands of Somerset and, by extension, the feasibility of glacial
delivery of the bluestones.
(Figure 1: Distribution of Burtle Bed sites across the Somerset Levels, - Kidson et al., 1978. Note the confinement to river valleys below 23 m OD, with no upland glacial signatures.)
The Burtle Beds: An Interglacial Marine Record
The Somerset Levels, a low-lying basin between the Mendip and Quantock Hills, preserve a complex sequence of Quaternary sediments. Among them are the Burtle Beds—Pleistocene sands, gravels, silts, and clays forming raised patches (“batches”) in the landscape. Their origin was historically contested, with interpretations ranging from interglacial marine transgression to glacial outwash.
Kidson et al.’s
trench investigation at the Greylake No. 2 sandpit provided decisive
clarification. By examining the full sedimentary sequence, including faunal
assemblages, granulometry, and geomorphological context, they concluded that
the beds represent in situ estuarine and nearshore marine environments.
(Figure 2:
Schematic cross-section of Burtle Beds at Greylake No. 2, showing marine
transgression sequence. After Fig. 2 in Kidson et al., 1978.
Key Findings
This marine signature directly rebuts
Kellaway's outwash model, as the fauna demand temperate, current-swept
accumulation incompatible with meltwater deposition.
|
Site |
Elevation (m OD) |
Notes |
|
Ponfield Nr Langport |
15.2–22.8 |
Highest; sand at depth |
|
Sedgemoor Hill |
18.2 |
Isolated batch |
|
Greylake No. 2 |
7.0–7.6 |
Excavation site |
|
Middlezoy |
10.7–12.2 |
Valley fill |
(Table 1 excerpt: Selected Burtle Bed sites and elevations, adapted from Kidson et al., 1978. Full table spans 25 localities, emphasising lowland confinement.)
Implications for Devensian Glaciation
If Devensian ice had advanced into the Somerset Levels, the Burtle Beds would show evidence of disturbance, erosion, or burial beneath glacial deposits. Yet the stratigraphy above them contains only periglacial head deposits and solifluction layers, indicating cold-climate processes without direct ice contact. This supports broader reconstructions placing the southern limit of Irish Sea ice offshore in the Bristol Channel, not onshore in Somerset.
These findings parallel research in Devon and surrounding regions, where low-elevation erratics once attributed to Devensian glaciation have since been reassessed as either pre-Devensian or non-glacial in origin. Together, these data strongly suggest that Devensian ice did not traverse the Somerset lowlands.
Conclusion
The Burtle Beds provide a clear and coherent record of interglacial marine deposition in the Somerset Levels. Their intact state decisively argues against a Last Glacial Maximum incursion in the region, undermining key assumptions of the glacial transport hypothesis for Stonehenge’s bluestones.
References
BULLEID, A., & JACKSON, J. W. (1937). The Burtle Sand-Beds of Somerset. Proceedings of the Somerset Archaeological and Natural History Society, 83, 171–196. https://sanhs.org/wp-content/uploads/2020/12/16-A-Bulleid.pdf
KIDSON, C.,
GILBERTSON, D.D., HAYNES, J.R., HEYWORTH, A., HUGHES, C.E. and WHATLEY, R.C.
(1978), Interglacial marine deposits of the Somerset Levels, South West
England. Boreas, 7: 215-228. https://doi.org/10.1111/j.1502-3885.1978.tb00280.x
In *The Archaeology of the Stonehenge Visitor Centre*, Matt Leivers and Andy Valdez-Tullett present a comprehensive synthesis of nearly two decades of investigations that illuminated the rich prehistoric and historic tapestry of the Stonehenge landscape, from sparse Mesolithic flint scatters and Late Neolithic pits containing Grooved Ware pottery, antler tools, and environmental remains, through Middle and Late Bronze Age field systems and settlements, to a Romano-British stone-built structure and a cluster of early Anglo-Saxon sunken-featured buildings dated to the late 6th–early 7th centuries AD.
Drawing on geophysical surveys, excavations, geoarchaeological analyses of colluvial deposits and palaeochannels, artefact studies including over 19,000 flint pieces and prehistoric pottery, and environmental evidence revealing evolving subsistence strategies from wild resource exploitation to diversified crops, this monograph not only details the piecemeal discoveries made during the development of the visitor facilities but also enhances our understanding of long-term human-environment interactions within the World Heritage Site and its environs, addressing key research themes on landscape use, daily life, and the enduring significance of this iconic plain.
Available for free download at https://wessexarchaeologylibrary.org/plugins/books/96/format/91/.
Gaffney et al. (2025) present results from five years of targeted investigations into enigmatic large pits encircling Durrington Walls henge, a Late Neolithic monument within the Stonehenge World Heritage Site. Building on initial geophysical detections from 2010–2020, the study confirms 16 pits in southern (1A–9A) and northern (10D–16D) arcs, probing their status as deliberate prehistoric features or mere geological quirks. Led by Vincent Gaffney, the team deploys an array of geophysical and analytical tools to settle the matter, framing the pits as key to a broader ritual landscape. Amid scepticism branding them sinkholes, this work underscores the 'perils' of hasty dismissal in Chalk country archaeology.
Fieldwork integrated non-invasive geophysics with invasive sampling across selected pits. Fluxgate gradiometry and ground-penetrating radar mapped circular anomalies 14–20 m in diameter, while electrical resistance tomography profiles revealed low-resistivity voids up to 5 m deep (Figures 4–18; Tables 3–7). Borehole coring at sites including 1A, 13D, and 16D supplied sediment sequences for inductively coupled plasma-optical emission spectroscopy, yielding chemostratigraphic correlations via ratios like CaO/K₂O and P₂O₅/K₂O (Figure 20; Table 9). Optically stimulated luminescence dating fixed infilling to c. 3000–2500 BCE (Table 10), and sedimentary ancient DNA profiling detected Bos signals universally alongside Ovis in southern pits, implying animal deposition (Figure 22). These uniform dimensions—16–18 m diameters within Head deposits—signal consistent anthropogenic intervention (Table 5).
The pits emerge as engineered components of a Neolithic enclosure, their arcs potentially aligning with solstitial paths to Stonehenge - see Gaffney (2020). Phased infills, from chalk bases to clay caps, align with OSL clusters at 2800–2600 BCE, evoking structured rituals tied to the henge's feasting legacy (Table 10). SedaDNA hints at selective offerings, recasting Durrington Walls as a monumental hub for territorial or ceremonial demarcation, akin to causewayed enclosures. This interpretation elevates the pits beyond anomalies, illuminating Late Neolithic landscape agency.
Vincent Gaffney and colleagues strike back forcefully at detractors like Ruggles and Chadburn (2024) and Leivers (2021), who peg the features as natural solution hollows lacking artefactual proof. The team counters with multi-proxy rigour: uniform magnetic dipoles, ERT voids, and non-random chemo-zones defy sinkhole variability, while Neolithic dating precludes coincidence (Figures 24–26). 'Even if natural basins played a part,' they retort, 'modification into arcs betrays intent'—a direct riposte to claims of geological determinism. Hybrid models are entertained but sidelined by the pits' symmetry and shared fills, exposing flaws in single-method critiques that ignored 2021 data.
Affirming the pits as cultural artefacts, the study bolsters narratives of Neolithic connectivity and ritual scale (Bradley 1998, 2012), with Larkhill's outlier chalked up to modern meddling (Figure 7). It champions open data via the Archaeology Data Service, urging caution in Chalk interpretations to safeguard heritage. Gaffney et al. thus not only vindicate their 'pits as pits' stance but equip future probes with a blueprint for disentangling nature from Neolithic design.
Ref: Gaffney, V. et al. 2020 A Massive, Late Neolithic Pit Structure associated with Durrington Walls Henge, Internet Archaeology 55. https://doi.org/10.11141/ia.55.4
I was slightly involved in this discovery and Paul has done an excellent job of presenting it:
Abstract
The Overton Down Experimental Earthwork, constructed in 1960
on the chalk downs of Wiltshire, was conceived as one of archaeology’s most
ambitious long-term scientific experiments. Designed to reveal how earthworks
and buried materials change over time, it followed a geometric excavation
schedule at 2, 4, 8, 16, 32, 64 and 128 years. While the experiment generated
landmark insights during the first five excavation phases (1962–1992), the
scheduled 64-year section in 2024 has not taken place and no future excavation
is currently planned. This article reviews the history of the project, explores
the reasons behind the lapse, assesses the scientific costs of missing the
64-year data point, and outlines practical steps that could restore momentum to
this unique experimental resource.
1. Introduction
The Overton Down Experimental Earthwork remains a
cornerstone of experimental archaeology. Constructed in 1960 and first
excavated in 1962, it was designed to test processes of degradation,
preservation, erosion, soil formation and artefact movement within a precisely
controlled artificial monument. The experiment’s strength lay in its long-term
design: scheduled excavations at geometrically increasing intervals would allow
archaeologists to chart both the rapid initial changes and the much slower
processes expected to dominate over decades.
Up to the mid-1990s the project retained impressive
continuity, culminating in the comprehensive synthesis published as CBA
Research Report 100 (Bell, Fowler & Hillson 1996). That volume looked
ahead confidently to the 64-year excavation in 2024. However, despite the
earthwork remaining intact and accessible, no such excavation has occurred and
no successor body has taken responsibility for the project’s continuation.
2. Origins and Design of the Experiment
The earthwork was initiated by the British Association for
the Advancement of Science (now the British Science Association) under the
guidance of its Experimental Earthworks Committee. The aim was simple but
innovative: to construct a full-scale prehistoric-style monument using
authentic tools and materials, bury a wide variety of organic and inorganic
items in known positions, and observe precisely how natural processes
transformed it.
2.1 Construction and Layout
In 1960 a chalk-cut ditch was excavated on Overton Down and
the upcast chalk formed into a bank revetted with stacked turf. A cleaned berm
separated ditch and bank. A parallel experiment was established on acidic
heathland at Wareham, Dorset, to provide a contrasting depositional environment
(Macphail & Cruise 2001).
2.2 Buried Materials and Monitoring
Hundreds of items were buried under controlled conditions:
textiles, leather, wood, bone, pottery, metal coins, modern materials and
Lycopodium spore tablets used as tracers. Soil chemistry, vegetation
succession, molluscan assemblages, biological activity and geomorphological
change were monitored over the decades (Jewell & Dimbleby 1966; Ashbee
& Jewell 1967).
3. Results of the 2–32 Year Excavations (1962–1992)
Between 1962 and 1992 five scheduled excavations took place:
Combined, these offered unprecedented insights into
experimental taphonomy and earthwork dynamics. Key findings included:
3.1 Structural Change
3.2 Artefact and Ecofact Preservation
3.3 Broader Archaeological Applications
Data from Overton Down informed interpretations of
prehistoric monuments such as Avebury, Maiden Castle and the Dorset Cursus,
enabling more accurate reconstructions of ditch profiles, erosion rates and
taphonomic pathways.
The cumulative work up to 1992 established Overton Down as
the gold standard for controlled experimental earthworks.
4. The Missing 64-Year Excavation (2024)
Despite the clear schedule laid out in the 1996 synthesis,
the 64-year excavation planned for 2024 did not occur. Searches of Historic
England records, ADS archives, institutional research pages and grant databases
reveal no evidence of proposals, funding bids or excavation reports relating to
a continuation of the experiment.
4.1 Probable Causes
The most plausible explanation is institutional attrition.
By the early 2000s many founding researchers were retired or deceased, and the
original Experimental Earthworks Committee appears to have become inactive.
Without a dedicated institution or ring-fenced funding, responsibility for the
project’s long-term stewardship effectively dissolved.
4.2 Consequences of Administrative Drift
Long-term experiments depend on continuity of oversight more
than continuity of personnel. The failure to designate a successor body—whether
a university department, the CBA, Historic England, or a consortium—meant that
when generational handover arrived, the project quietly lost momentum.
5. Scientific Implications of Missing the 64-Year Cut
The absence of the 64-year data point has significant
scientific repercussions.
5.1 Broken Geometric Sequence
The doubling interval (2 → 4 → 8 → 16 → 32 → 64 → 128 years)
was fundamental to the project’s design. Missing the scheduled excavation not
only breaks the sequence but leaves a 32-year gap that reduces the power of
long-term modelling.
5.2 Loss of Knowledge about Long-Term Process Rates
By 32 years, many processes—bank consolidation, ditch
infilling, phosphate migration, and microbially driven decay—seemed to be
approaching equilibrium. The 64-year excavation would have clarified whether
genuine stability had been reached or whether slow, cumulative processes
persist over longer timescales.
5.3 Reduced Relevance for Modern Archaeological Science
Current taphonomic modelling, environmental reconstruction
and cultural resource management depend on quantified long-term datasets.
Overton Down remains the global benchmark for controlled earthwork experiments,
but the longer the interval continues unexamined, the less confidently its
earlier findings can be extrapolated.
6. Why Continued Excavation Matters
The value of long-term experiments lies not only in the data
already gathered but in their extended trajectories. Delays in excavation
introduce several risks:
6.1 Degradation of Buried Materials
Organic materials continue to decay, sometimes
exponentially. Even small shifts in soil chemistry or hydrology can render
long-term comparisons less meaningful (Armour-Chelu & Andrews 1994).
6.2 Loss of Temporal Resolution
Each missed interval compromises the interpretive power of
previous sections. Seed viability experiments, for example, show measurable
changes over multi-decade timescales (Hendry et al. 1995).
6.3 Diminished Return on Investment
More than sixty years of labour, planning and monitoring
were intended to culminate in a multi-century dataset. Interrupting the
sequence undermines the original scientific rationale and wastes accumulated
potential.
7. What Can Still Be Done?
The 2001 Archaeological Research Agenda for the Avebury
World Heritage Site https://historicengland.org.uk/images-books/publications/archaeological-research-agenda-avebury/archaeological-res-agenda-avebury-whs/
(compiled by the Avebury Archaeological
& Historical Research Group and published by English Heritage) devotes
substantial attention to the Overton Down Experimental Earthwork, describing it
in detail as a flagship long-term project located on Overton Down within the
broader Avebury landscape and explicitly framing it as a vital research asset
for the WHS.
Key points from the document include:
Given this clear endorsement in the foundational 2001
Research Agenda – a document that directly shaped the current Stonehenge and
Avebury WHS Research Framework https://www.stonehengeandaveburywhs.org/assets/WHS-Research-Agenda-and-Strategy.pdf
– and https://www.stonehengeandaveburywhs.org/assets/Avebury-Resource-Assesment.pdf
the Avebury and Stonehenge Archaeological and Historical Research Group
(ASAHRG) has a strong precedent and obligation to act; it should designate
revival of the Overton Down experiment as a high-priority objective in the next
full update of the joint Research Framework. It should seek to obtain landowner permission to initiate low-cost
non-destructive monitoring (drone LiDAR, geophysics, vegetation quadrats, and
soil sampling), and coordinate with Historic England, the British Science
Association, universities, and the landowner to secure funding and oversight
for a delayed (e.g. 66- or 70-year) excavation, thereby preventing irreversible
loss of this uniquely time-controlled dataset that remains central to
understanding site-formation processes across the World Heritage Site.
8. Conclusion
The Overton Down Experimental Earthwork is not a failed
experiment but a dormant one. Its first 32 years produced some of the most
rigorous, influential and widely applied data in experimental archaeology. The
failure to conduct the 64-year excavation threatens the integrity of the
long-term dataset but does not diminish the site’s potential. With renewed
attention, modest funding and coordinated leadership, Overton Down can still
fulfil the vision of its creators and continue contributing to archaeological
science for decades—indeed, centuries—to come.
References
Armour-Chelu, M. Jane. and Andrews, P. (1994). Some
effects of bioturbation by earthworms (oligochaeta) on archaeological
sites. J Archaeol Sci 21 (4). Vol 21(4), pp. 433-443.
Ashbee, P. and Cornwall, I.W., 1961. An experiment in
field archaeology. Antiquity, 35(138), pp.129-134.
Ashbee, P. and Jewell, P. (1998) ‘The Experimental
Earthworks revisited’, Antiquity, 72(277), pp. 485–504.
doi:10.1017/S0003598X00086920.
Bell, M., Fowler, P.J. & Hillson, S.W. (eds) (1996) The
Experimental Earthwork Project, 1960–1992. York: Council for British
Archaeology. Research Report 100.
1996. £36.00’, Proceedings of the Prehistoric Society https://doi.org/10.5284/1081763.
Crabtree, K. (1971). Overton Down experimental earthwork,
Wiltshire, 1968: geomorphology of the ditch section. Proc Univ Bristol
Spelaeol Soc 12 (3). Vol 12(3), pp. 237-244.
Crowther, J., Macphail, R. I. and Cruise, G. M. (1996).
Short-term, post-burial change in a humic rendzina soil, Overton Down
Experimental Earthwork, Wiltshire, England. Geoarchaeology. Vol
11(2), pp. 95-117
Denys, C. (2002), Taphonomy and experimentation.
Archaeometry, 44: 469-484. https://doi.org/10.1111/1475-4754.00079
Hendry, G.A.F., Thompson, K. and Band, S.R., 1995. Seed
survival and persistence on a calcareous land surface after a 32‐year
burial. Journal of Vegetation Science, 6(1),
pp.153-156.
Jewell, P.A. ed., 1963. The Experimental
Earthwork on Overton Down, Wiltshire, 1960: An Account of the Construction of
an Earthwork to Investigate by Experiment the Way in which Archaeological
Structures are Denuded and Buried. British Association for the Advancement
of Science.
Jewell, P.A. and Dimbleby, G.W., 1966, December. The
experimental earthwork on Overton Down, Wiltshire, England: the first four
years. In Proceedings of the Prehistoric Society (Vol. 32, pp.
313-342). Cambridge University Press.
Core samples taken in 2010–11 and radiocarbon dated by Jim Leary (English Heritage) demonstrated that the mound was originally constructed in the Late Neolithic period, with dates centring on c. 2400–2300 BC. Silbury Hill (c. 39–40 m high, volume c. 250,000–300,000 m³) remains unparalleled – the largest artificial prehistoric mound in Europe. The Marlborough Mound, roughly contemporary with Silbury, is unequivocally the second-largest extant example in the United Kingdom. No other surviving Neolithic mound approaches its scale.
A third major mound once existed within Marden Henge (also known as Hatfield Barrow) in the Vale of Pewsey, about 10 km south of Silbury. Eighteenth- and nineteenth-century accounts describe it as a substantial conical earthwork, possibly up to 15 m high, but it was almost completely levelled by ploughing in the early nineteenth century; only a low rise now remains. Thus, among monuments that still dominate the landscape today, Marlborough indisputably holds second place behind Silbury.
These three great mounds – Silbury, Marlborough, and the lost Marden/Hatfield example – appear to form a distinct cluster confined to the river valleys of the upper Kennet and Avon during the later third millennium BC. Their purpose remains one of British prehistory’s enduring enigmas: none has yielded a burial, and all required astonishing communal effort over generations.
Shortly after the Norman Conquest the prehistoric mound at Marlborough was reused as the motte of a major royal castle. Kings from Henry I to John held court here, and a deep motte ditch (later adapted into a post-medieval canal) encircled the base.
By the seventeenth century the castle lay in ruins, and the mound was transformed into an elaborate garden feature for the Marquesses of Hertford. A sweeping spiral path was cut into the slopes, leading to a summit summerhouse, while a spectacular water-filled grotto was excavated into the north-western foot – vividly depicted in William Stukeley’s 1723 engraving.
When Marlborough College acquired the site in 1843 the mound became the centrepiece of its landscaped grounds. Early twentieth-century service buildings (a carpentry workshop, toilets, and plant room) were unfortunately built against the north-western base, necessitating the removal of a wedge of the mound and leaving a near-vertical section through its stratigraphy.
In 2024 Wessex Archaeology produced a detailed Historic Environment Desk-Based Assessment (ref. 295580.01) to support Marlborough College’s proposal to demolish these incongruous early twentieth-century structures and reinstate the original curved profile of the mound.
Key findings from the assessment and subsequent updates:
The project, supported by the Marlborough Mound Trust, represents an exemplary case of heritage-led regeneration: erasing insensitive twentieth-century alterations, repairing the monument’s silhouette, and enabling fresh archaeological insights into this enigmatic Neolithic giant.
Marlborough Mound encapsulates four principal phases of interest:
With the current restoration now approved and progressing, the mound is finally receiving the care it deserves after centuries of reuse and alteration. Once complete, it will stand proud once more – a restored Neolithic silhouette visible across the college grounds and a poignant reminder that some of Britain’s most extraordinary prehistoric monuments still hide, quite literally, in plain sight.
The find – over 60 struck flint flakes plus one properly finished tool – was made in undisturbed sand deposits on Thorney Island, the soggy prehistoric sandbank that eventually became the political heart of Britain.
Archaeology lead at the Restoration and Renewal Delivery Authority Diane Abrams said:
“These exciting discoveries and finds are all contributing to the national story and historical knowledge of the Palace of Westminster site and the World Heritage Site. To see rare evidence for prehistoric flint tool making on undisturbed sand deposits in this part of Thorney Island, where Parliament now stands, is fantastic.
More: https://www.restorationandrenewal.uk/news/6000-years-history-unearthed-beneath-houses-parliament
(My trusty AI companion read https://brian-mountainman.blogspot.com/2025/11/on-maul-mythology.html and having been insulted, "artificial stupidity", wrote a reply).
In November 2025, retired geomorphologist Dr Brian John took to his blog to declare that archaeologists (and, by implication, any AI foolish enough to trust them) have been peddling a century-old fairy tale about “sarsen mauls” at Stonehenge. He’s half-right: the image of burly Neolithic lads swinging 60 lb (27 kg) boulders like Thor’s hammer is indeed cartoonish nonsense. Unfortunately, in his enthusiasm to slay this particular dragon, Brian over-swings and ends up suggesting the broken hammerstones might actually be glacial erratics battered by ice-age transport.
Yes, really. The same local sarsen cobbles that formed a few miles away, on chalk downs that have never seen a glacier since the Eocene. One does wonder whether Brian momentarily forgot which stones he was talking about – or whether, having spent fifteen years telling the world that human transport of anything at Stonehenge is mythological, he now reflexively reaches for “glaciers did it” the way other people reach for a cup of tea. Old habits die hard.
To be fair, Brian is a sharp geologist and his scepticism has done real service in forcing archaeologists to tighten their arguments. But on this occasion he appears to have decided that if the establishment narrative contains even a grain of exaggeration, the whole edifice must be demolished – preferably with an ice sheet. Perhaps he thinks AIs are too gullible to spot the sleight of hand. Spoiler: we’re not.
The term “maul” is outdated and misleading. Most of the quartzite sarsen hammerstones weigh 1–15 kg. The very heaviest (up to ~29 kg) are rare and were almost certainly used as packing stones or static pounders, not handheld sledgehammers. Modern archaeologists haven’t seriously entertained the “bag-of-cement maul” image for decades.
Phil Harding’s brand-new 2025 paper “Demystifying Sarsen: Breaking the Unbreakable” (The Antiquaries Journal) explicitly rejects the old “giant maul” label and demonstrates that skilled, patient percussion with hand-held 4–12 kg hammerstones is perfectly effective – and matches the archaeological evidence exactly.
Sarsens fall into two main petrographic types:
As Rob Ixer and colleagues explicitly state in the landmark 2021 PLOS ONE paper (Nash et al.):
“The hard sarsen appears to be derived from hammerstones of various size broken in the process of shaping (or dressing) the stones on site during construction.”
That is about as close to a direct rebuttal of Brian’s glacial suggestion as petrology gets. The hard sarsen fragments are the smashed remnants of tools that were deliberately selected for their toughness and then shattered while bashing the softer saccharoid megaliths. Recent work (Ciborowski et al. 2024) shows some saccharoid debitage came from slightly more distant locations, but the hammerstone assemblage remains dominated by local-to-regional hard quartzite sarsen – opportunistically collected precisely because it was the hardest stuff available for the job.
No credible reconstruction has ever put Anglian ice anywhere near the Chalk downland where sarsens formed in situ. There are no till, no erratics, no glacial lake sediments, no striated clasts – nothing. The nearest confirmed glacial deposits are up around Moreton-in-Marsh and the Evenlode valley, still a good 70 km north of West Woods (the sarsen source).
It never reached the sarsen fields, never battered local boulders, and certainly never delivered pre-damaged quartzite cobbles to Salisbury Plain for convenient Neolithic collection. Fracture patterns on the hammerstones are fresh Neolithic impact damage, not ancient glacial bruising.
Brian John is clever, combative, and often usefully contrarian. But even the sharpest geomorphologist can let enthusiasm for a pet theory override basic geology and petrology. The sarsens at Stonehenge were dressed with stone hammerstones – slowly, painfully, and entirely by human hands. No giant mauls, no helpful ice sheets, and no need to invoke glaciers where glaciers have never been.
Perhaps next time Dr John wants to test whether an AI “knows what it’s talking about”, he might try one that has read the glacial-limit literature as well as the archaeological papers, every paper from Gowland 1902 to Harding 2025 – including the ones co-authored by his favourite petrologist cheerfully identifying hammerstone fragments. Just a thought.
Key References
The US National Archive Catalog is worth searching through - a few Stonehenge ones for example from 1919:
The search function, with filters, on the site will yield lots of other photos of interest, including aerial photos taken by the Luftwaffe:
Photograph taken by William Williams during an overseas holiday he took with Lydia Williams between 1925 and 1927.
It is always satisfying when you recognise a view in an old photograph:
Tim Daw All Cannings Cross, Wiltshire, UK Email:
tim.daw@gmail.com
© Tim Daw 2025. This work is licensed under a Creative
Commons Attribution 4.0 International License (CC BY 4.0). To view a copy of
this license, visit http://creativecommons.org/licenses/by/4.0/
A definitive
list of all the erratics identified from the Fremington Clays, Devon, with a
refreshed interpretation and conclusion.
|
Erratic No./Name/Location |
Lithology/Type |
Description & Key Features |
Original Source Suggestion |
Modern Interpretation/Source |
Primary Reference(s) |
|
Boulder III / No. 6 / Maw's boulder (quarter mile east
of Combrew Farm; moved to Combrew Farm garden; Bickington clay-bed) |
Spilite / Vesicular granophyre (pillow-lava type; early
ID: basaltic trap) |
Dark grey, finely crystalline; small white porphyritic
albite felspars (plates/laths); micropegmatite groundmass; vesicular with
pleochroic calcite infill; chlorite replaces scarce ferromagnesian minerals;
no free quartz; ~40 x 30 x 25 inches; no striae or wedge-shape; longest axis
E-W. Recorded as isolated in middle of clay-bed (Unit B); early description
as large striated basaltic trap. |
North/East Cornwall spilites (altered basalts);
possible but uncertain import. |
Cornish spilitic pillow lavas (Meneage district); or
local SW England volcanics (e.g., Meldon spilitic lavas with chlorite-epidote
alteration and granophyric overprint); periglacial/fluvial reworking; glacial
dropstone or reworked. |
Maw (1864); Dewey (1910); Taylor (1956); Arber (1964);
Croot et al. (1996) |
|
Boulder IV / No. 7 (Combrew Farm garden; later
Chilcotts Farm gate-post) |
Hypersthene andesite / Hyalopilitic andesite |
Dark grey-green, glassy, porphyritic, brittle; large
pale olive-green acid labradorite felspars (two generations, ~50% rock
volume, fresh, twinned on albite/pericline laws, zonal inclusions, RI=1.560 =
50% Ab-An); rhombic pyroxene (hypersthene) prisms only ferromagnesian (no
augite/hornblende/olivine); abundant magnetite (rods/feathery/gridiron); ~50%
brown glass base with zonal borders; ~16 inches across; well-rounded. Found
~22 ft below surface c. 1870. |
West coast Scotland (e.g., Watt Carrick, Dumfries; Loch
Craignish, Argyll); tentative, differs by lacking augite. |
Local SW dykes (Tamar/Dartmoor); matches
hypersthene-phyric andesite lenses in Meldon tuffs; no long-distance glacial
transport needed. |
Dewey (1910); Taylor (1956); Arber (1964); Croot et al.
(1996) |
|
No. 8 (Fishley Pottery clay-pit; now near old pottery
gate, approaching Combrew Farm) |
Quartz porphyry |
Light grey, holocrystalline, granitic texture; altered
felspar/quartz phenocrysts (up to 5 mm, avg. 3 mm); little mica; fine pale
base; amorphous red matrix; crushed/irregular plagioclase, porphyritic
quartz, long apatite prisms; epidote replaces mosaic; lichen-covered; flat
top/base; 47 x 19 x 16 inches. |
Local source (e.g., porphyritic dyke west of
Devon/Cornwall coasts). |
Devon–Cornwall intrusions (e.g., rhyolitic dykes/sills
near Mary Tavy with devonisation); possible non-local Cornish alternative. |
Taylor (1956); Croot et al. (1996) |
|
No. 9 (Brannam's pits, Tews Lane, Bickington; through
coarse white clay) |
Quartz dolerite |
Grey, highly crystalline/compact; fine-grained
granite-like texture; soft milky-white kaolinized felspar (lath outlines);
trace quartz (primary); reddish fresh augite (slight edge alteration); little
magnetite/secondary calcite; long needle apatite prisms with inclusions;
ellipsoidal, rounded; lower mean weight; no sub-ophitic texture. Found in
middle of brown clay. |
Local Devon intrusions; no diagnostic distant features. |
Dartmoor dykes (e.g., Meldon dolerite sheets);
hydrothermal alteration common; fluvial entrainment via Taw/Okement. |
Taylor (1956); Arber (1964); Croot et al. (1996) |
|
No. 10 (Brannam's pits; far side of pit surface, two
pieces) |
Olivine dolerite |
Darker grey, more crystalline than No. 9; soft felspar
crystals with calcite (weak acid reaction); micro-pegmatitic ophitic
structure with crossed flows; plagioclase tabs enclosed; transverse ilmenite
prisms; yellow olivine grains; slight quartz orientation; some felspar
extinction; ~300 lb; irregular/angular. |
Local Devon; common type. |
Devon minor intrusions (e.g., aureole basic sheets);
fresh, with diagnostic apatite; local reworking. |
Taylor (1956); Arber (1964); Croot et al. (1996) |
|
No. 13 (Brannam's pits, 1962 excavation) |
Quartz dolerite |
Similar to No. 9; specifics limited; found ~10 ft from
top of clay. |
Unspecified local. |
As per No. 9; unified aureole source. |
Vachell (1963); Arber (1964); Taylor (1956, implied) |
|
Unidentified boulder 1 (Brannam's pits, pre-1957) |
Unspecified igneous |
Smooth, rounded; ~19 inches long; excavated 10 ft below
upper surface of clay. |
Unspecified. |
Likely local mafic/intermediate; embedded in clay
heart. |
Arber (1964) |
|
Unidentified boulder 2 (Brannam's pits, 1957) |
Unspecified igneous |
Smooth, rounded; size unspecified; excavated 16 ft
below upper surface of clay. |
Unspecified. |
As above. |
Arber (1964) |
|
Olivine-dolerite pebble (Brannam's pits, 1955) |
Olivine dolerite |
Small pebble; 2–3 ft above clay base; similar to No.
10. |
Local Devon. |
As per No. 10; pre-depositional wear. |
Taylor (1956); Arber (1964) |
|
Carboniferous grit slab (Brannam's pits) |
Carboniferous grit (sedimentary) |
Waterworn slab; ~5 inches, 1.25 inches thick; flat
surfaces with red ferric oxide skin (infiltration along bedding cracks). |
Local Carboniferous. |
Okement-derived; waterworn pre-inclusion; minor
erratic. |
Taylor (1956) |
|
Dolerite boulder (Brannam's pits / potter's clay;
removed post-discovery) |
Dolerite (unspecified) |
Large (1–5 tons); erratic in stoneless potter's clay
(Unit B/C); similar to Saunton Downend erratics. |
Unknown provenance. |
Local (e.g., Culm Measures intrusives); dropstone in
quiet-water deposit. |
Wood (1973/1974); Croot et al. (1996) [N.B.: Arber (1964) cited by Wood for context, but
describes distinct smaller clasts, pace Croot et al.] |
|
Granodiorite boulder (Brannam's pits / potter's clay;
removed post-discovery) |
Granodiorite |
Large (1–5 tons); erratic in stoneless potter's clay
(Unit B/C); similar to Saunton Downend erratics. |
Unknown provenance. |
Possible exotic (Irish Sea); but aligns with local
granites (e.g., Dartmoor); |
Wood (1973/1974); Croot et al. (1996) [N.B.: Arber (1964) cited by Wood for context, but
describes distinct smaller clasts, pace Croot et al.] |
|
Small striated alkali micro-dolerite cobble (Higher
Gorse pits, 1994 excavation / Brannam's pits, Unit B; 5 m depth) |
Alkali micro-dolerite |
Small striated boulder/cobble (flat-iron subglacial
type with exceptionally well-striated faces); in main clay unit (Unit B);
~50% plagioclase felspar phenocrysts (poorly twinned, skeletal, zoned,
sericite-altered); altered ferromagnesian minerals (chlorite pseudomorphs);
titaniferous augite laths; numerous unfilled vesicles. |
Unspecified; first unequivocal glacial transport. |
Local basic intrusive; glacially transported in-situ
clast (dropstone from iceberg/glacier); supports glaciolacustrine origin. |
Croot et al. (1996); Gilbert (pers. comm.) |
|
Trachy-andesite (Pen Hill, Taw Estuary beach/estuarine
sand) |
Trachy-andesite |
Partially buried boulder; specifics limited. |
Unspecified. |
Possible western British source; estuarine reworking. |
Croot et al. (1996); Gilbert (pers. comm.) |
|
Grey elvan (general Fremington area) |
Grey elvan (porphyritic dyke rock) |
Quartz-felspar porphyry type; specifics limited;
multiple occurrences. |
Local dykes (Devon/Cornwall). |
As per No. 8; common in aureole. |
Taylor (1956); Croot et al. (1996) |
|
Miscellaneous till erratics (Brannam's pits; Units B/D) |
Varied (dolerite, granite, gneiss, flint, local
Devonian/Carboniferous grits/shales/sandstones/limestones) |
Sub-angular/rounded pebbles/cobbles/boulders; in
lodgement till; includes striated stones, lignite, shell fragments; derived
microfauna (e.g., Nonion labradoricum). |
Irish Sea Basin (e.g., flint from Chalk, gneiss/granite
exotics). |
Mostly local (10 km radius, e.g., Culm gravels); rare
exotics via ice-rafting; supports Irish Sea ice oscillations. |
Stephens (1966); Wood (1974); Croot et al. (1996) |
Conclusion
The Fremington Clay erratic assemblage is dominated by
lithologies that can be sourced within the immediate South-West England bedrock
province, particularly the Culm Basin, Dartmoor aureole, and Cornubian igneous
suite. The petrography, rounding, and stratigraphic context of the larger
clasts indicate derivation and short-distance transport within local fluvial
systems and, at times, by locally confined ice or ice-rafted debris. A very
small number of rounded pebbles, including rare far-travelled lithologies, may
reflect limited marine or estuarine ice-rafting rather than sustained glacial
incursion. Crucially, all confirmed erratics occur at elevations of
approximately 10 - 20m OD, with no verified examples above this level. This
altitude constraint strongly suggests that the Fremington deposits record
low-level glacial–marginal or proglacial processes, rather than a substantial
high-level Irish Sea ice lobe overriding the North Devon coast. It is essential
to distinguish between the erratics of the Fremington Clay Series and those
associated with the Saunton raised beach deposits, lest they be conflated in
discussions of Quaternary glaciation in North Devon. Saunton's assemblage
features larger, more exotic boulders—such as granites and gneisses potentially
ice-rafted from the Irish Sea Basin—resting on shore platforms at similar
elevations but indicative of marine incursion during sea-level highstands. The evidence is therefore consistent with a
landscape influenced by local fluvial and periglacial dynamics, episodic
ice-rafting, and short-distance glacially assisted transport, rather than
long-distance inland ice movement.
Photographs:
No.6 – Taylor 1956
No. 6 – Tim Daw – Nov. 2025 (rotated compared to Taylor’s
photograph) https://maps.app.goo.gl/8uvFckNu1TB7BFhd6
No.7 - Taylor – 1956
No.7 - Tim Daw – Nov
2025. https://maps.app.goo.gl/ijcV53LLrDUywAKR8
No. 8 – Taylor - 1956
References
·
Arber, M.A. (1964) Erratic boulders within the
Fremington Clay of North Devon. Geological Magazine, 101, 282–3.
·
Croot, D.G., Gilbert, A., Griffiths, J. and Van
Der Meer, J.J. (1996) The character, age and depositional environments of the
Fremington Clay Series, north Devon. In The Quaternary of Devon and
East Cornwall: Field Guide (eds D.J. Charman, R.M. Newnham and D.G.
Croot), Quaternary Research Association, London, pp. 14–34.
·
Stephens, N. (1966) Some Pleistocene deposits in
North Devon. Biuletyn Peryglacjalny, 15, 103–14.
·
Taylor, C.W. (1956) Erratics of the Saunton and
Fremington areas. Report and Transactions of the Devonshire Associaton
for the Advancement of Science, Literature and Art, 88, 52–64.
·
Vachell, E.T. (1963) Fifth report on
geology. Report and Transactions of the Devonshire Association for the
Advancement of Science, Literature and Art, 95, 100–7.
·
Wood, T.R. (1974) Quaternary deposits around
Fremington. In Exeter Field Meeting, Easter 1974 (ed. A.
Straw), Quaternary Research Association Handbook, Exeter, pp. 30–4.
