More on the West Kennet granidiorite mystery

Some of the coherent cobbles found at West Kennet.  They are interpreted as "corestones" or solid rock remnants which have survived while the outer parts of a boulder (or several) are weathered away.

Cunyan Crags -- a possible source area

Here is a new article on the mystery of the strange rotten granidiorite bits and pieces found at West Kennet. The article is not out yet in printed form -- it is dated as 2025.  It's a detailed paper, and although the authors (as ever) tend towards an explanation of the boulder as "yet another possible example of Late Neolithic long distance prehistoric transport" they admit to being puzzled.  Are they correct in indicating that the 77 bits have possibly all come from one boulder derived from the Cunyan Crags area of Cheviot, in Northumberland?  Time will tell.........

Of course, boulders such as this one could have been picked up by overriding ice and carried southwards -- rather like the Whin Sill erratics described by Olwen Williams-Thorpe and others many years ago. But why on earth would anybody want to collect the boulder (if it was just one) from a till exposure or from a glaciated ground surface and cart it off to West Kennet?

https://brian-mountainman.blogspot.com/2021/11/breaking-news-another-igneous-erratic.html

https://brian-mountainman.blogspot.com/2012/01/whin-sill-erratic-cluster-in-stevenage.html

https://brian-mountainman.blogspot.com/2012/04/chunk-of-whin-sill-in-london.html

"Geochemical provenancing of igneous glacial erratics from Southern Britain, and implications for prehistoric stone implement distributions" by Olwen Williams-Thorpe, Don Aldiss, Ian J. Rigby, Richard S. Thorpe, 22 FEB 1999, Geoarchaeology, Volume 14, Issue 3, pages 209–246, March 1999

http://onlinelibrary.wiley.com/doi/10.1002/%28SICI%291520-6548%28199903%2914:3%3C209::AID-GEA1%3E3.0.CO;2-7/abstract

Put this down as one of the mysterious boulders of Southern England and the Channel coasts..........

==================================

Rob Ixer, Richard Bevins, Nick Pearce, Duncan Pirrie, Josh Pollard, Alex Finlay, Matthew Power and Ian Patience. 2025 "Exotic granodiorite lithics from Structure 5 at West Kennet, Avebury World Heritage Site, Wiltshire, UK." Wiltshire Archaeological & Natural History Magazine, vol. 118 (2025), pp. 1–18




ABSTRACT


Seventy-seven pieces of very weathered pyroxene-bearing granodiorite corestone excavated from trenches 2, 3 and 9 within Structure 5 of West Kennet in 2019 and 2021 and varying from small pebbles to >500grms cobbles, have a total weight of 22kg. Detailed petrographical and geochemical analyses of typical samples show them to share an unusual (for Britain) and distinctive mineralogy and petrography and also suggest they are all from a single outcrop/subcrop. The essentially unaltered pyroxene-bearing granodiorite carries ‘large’ skeletal zircon crystals, which are a determinative characteristic. Petrological comparisons with similar British granodiorites show that its origin is to be found within the large, 60km and lithologically highly diverse Cheviot Igneous Complex of Northumberland, more than 450km from West Kennet. Three Cheviot samples were selected for comparative analysis, one chosen for its petrographic similarity to the corestones, as suggested by previous workers, a second, close to the first and also to significant Neolithic activity at Threestoneburn Stone Circle, and finally a third based on petrography and notable topography, namely Cunyan Crags.  Only the last sample shares a sufficient number of similarities that there warrants further investigation in that area. The corestones are highly exotic with regard to their find spot and although it is difficult to conceive of any practical use for them, West Kennet provides yet another possible example of Late Neolithic long distance prehistoric transport, a distance of between 450km if taken from outcrop and 150km if collected from secondary glacial drift sources, although North Sea coastal glacial tills as a source for the stones appears unlikely and from East Anglia very unlikely. The original Cheviot Hill location remains unidentified but is being actively sought.

The Courtmacsherry "glacial" raised beach

  I have been looking again at the evidence for the Courtmacsherry raised beach, and am still very puzzled.  The dating of it is hugely dependent upon the reliability of OSL dating, and I'm a bit worried about some of the other grounds cited for it being formed at a time of high RSL in MIS 3 or 4.  I have nagging doubts, and remain to be convinced that it is not Ipswichian in age, like most of the other raised beaches in the SW of the British Isles.

I have been reading the excellent article by Colm O'Cofaigh et al (2012), and have concerns about the following:

Extracts from p 175

Undoubtedly there is extensive evidence for the beach along the
south coast of Ireland forming in a cold-climate environment (see
above), and indeed some of the evidence which underpins this
interpretation was first noted by Wright and Muff (1904). This
includes the presence of erratics within the raised beach gravels
and interbedding of the beach gravels with periglacial slope
breccias. Previous investigations of the sand facies that overlie the
raised beach at sites such as Howe's Strand and Broadstrand have
interpreted it as ‘blown sand’ (e.g. Synge, 1978). This is inconsistent
with the sedimentology of these sands which exhibit well defined
hummocky and swaley cross-stratification consistent with
a shallow marine rather than aeolian setting. Thus the facies
sequence of raised beach/HCS sands/SCS sands indicates
submergence following beach formation. Periglacial slope deposits
and isolated angular clasts of local bedrock within these sands
indicate the maintenance of a ‘cold’ depositional environment
during this submergence.

We suggest that high
relative sea level and the submergence indicated by the facies
sequence of raised beach deposits overlain by shallow marine sands
are related to glacioisostatic depression during ice sheet build-up on
land combined with still high eustatic sea level prior to the LGM. The
interpretation of high RSL during ice sheet build-up is consistent with
observations of erratics in the beach gravels (proposed byWright and
Muff to record deposition from floating ice) and the presence of
periglacial slope breccias interbedded with the beach and overlying
shallow marine sands.

Colm Ó Cofaigh, Matt W. Telfer, Richard M. Bailey, David J.A. Evans, 2012
Late Pleistocene chronostratigraphy and ice sheet limits, southern Ireland
Quaternary Science Reviews 44 (2012) 160-179.

First, the evidence of erratics within the raised beach gravels is NOT a ground for claiming that the beach is a cold-climate feature. Second, how do the authors know that the slope breccia is a genuine cold-climate deposit rather than one associated with rockfalls such as one sees all the time on a temperate coastline subject to wave attack?  Third, the transition from shoreline deposits to shallow-water marine deposits suggests a transgression or rising sea level, but that does not of itself prove glacio-isostasy or the advance of a glacier from the north or from anywhere else.  Raised beaches may be formed at a time of falling sea-level, or rising sea level, or indeed oscillations over a few metres over thousands of years.

The raised beach deposits are apparently devoid of organic materials -- which is a pity!  A shell fauna such as that seen in the Gower raised beaches would be rather valuable in sorting this matter out........

I'm struck by the similarity between the southern Irish sites and West Angle, at the mouth of Milford Haven.  Here there are interglacial sands, silts and clays which have been overridden and partly destroyed by advancing ice which left behind a characteristic Irish Sea till.  I have dealt with West Angle in many posts on this blog (use the search box), and I do not see any features there that might support the idea of an Early or Mid Devensian glaciation or a high RSL on the Pembrokeshire coast coinciding with a global sea-level at about - 80m. This of course implies isostatic depression of the coastline by around 80m, which in turn implies an ice load imposed by a glacier c 300 - 400m thick.  That's a big ask for the Early or Mid Devensian........

The surveyed exposure at West Angle, showing a sequence of supposed interglacial marine and estuarine deposits overlain by till.

Quaternary events on the North Devon coast

Prof Nick Stephens's map of ket Quaternary features in the South West.  Many of the features on this map are now known to be incorrect, but note Lake Maw -- centred on the Somerset Levels, withy an overflow to the south

Been looking again at this major contribution:

Bennett, J. A., Cullingford, R. A., Gibbard, P. L., Hughes, P. D., & Murton, J. B. (2024). The Quaternary Geology of Devon. Proceedings of the Ussher Society, 15, 84-130. 

https://ussher.org.uk/wp- content/uploads/benettetal1584130v2.pdf

I think we are getting pretty close to a consensus here:

Quote:

Glaciation appears to have modified the landscape south of the Bristol Channel relatively little, although this is not surprising given that the ice from both the Irish Sea and South Wales did not penetrate far beyond the present north Devon and Somerset coast. Presumably the relatively easily deformed marine deposits flooring the Bristol Channel allowed the ice to flow along or across the valley until reaching the resistant rocks on today’s northern coast of Devon. Here, river valleys parallel to the coast and also truncated may in fact be ice- marginal landforms (Rolfe, 2015; Gibbard, et al. 2017). The BRITICE project (Clark et al., 2020) takes a broad view of the glacial limits in Britain but suggests that the pace of ice retreat from the Celtic Sea may have varied according to bed topography. In addition, Scourse et al. (2021) propose short- lived events of ice advance down the Celtic Sea that may have affected the north Devon coast and the further possibility that the ice margin had advanced as far as the shelf break at 25.5 ka.

Lake Maw is still a puzzle!  In my correspondence with Philip H and Philip G we have pondered on Lake Maw.  Use the search box to look at earlier posts.  If ice really did cross the Bristol Channel and hit up against the clifflines of N Devon and N Cornwall, in the Late Wolstonian or Late Devensian (and maybe at other times too!) then an ice barrier must have blocked the Bristol Channel and impounded a considerable meltwater lake in the Severn Estuary  -- referred to by the older workers (including Frank Mitchell and Nick Stephens) as Lake Maw.  But in that case, where are the lacustrine clays that might be expected in this lake?  

A similar lake -- Glacial Lake Teifi -- filled a broad depression in the Teifi Valley when Irish Sea ice moved in from Cardigan Bay from the W and NW.  No shorelines have been identified (which means that its surface level was anything but stable) but there are abundant laminated lake clays in many locations, as described by Fletcher and Siddle (1998) and Etienne et al (2006). At the time of its greatest extent it was more than 10 km wide and 40 km long.

doi:10.1016/j.quascirev.2005.03.019

The interesting thing about this lake is that it formed during the ADVANCE of the ice from the Cardigan bay, rather than during its dissolution.  It appears that the ice advanced over the lake deposits, creating complex structures and depositing Irish Sea till and glaciofluvial sediments as it did so,

The full sequence of events is not yet elucidated......

High glacial sea levels?

Newgale storm beach -- the most prominent feature of its type in Wales.  Erratics galore.....

 In several journal articles recently I have encountered variations on this statement:  "The raised beach contains erratic pebbles, demonstrating that at the time of its formation glacier ice was active in the vicinity."   How weird is that? 

Let's just look at the present-day storm beaches and quiet water beaches of Pembrokeshire:  Poppit, Ceibwr, Parrog, Newport, Abermawr, Druidston, St Brides, Manorbier, Dale, West Angle,  and so on and so on.  Hundreds of beaches.  They ALL contain erratic cobbles.  The percentages are not very high -- maybe up to 10% in places.  But I defy anybody to name a beach in Pembrokeshire that does NOT contain far-travelled erratic materials. 

The explanation is simple.  Beaches are generally built of clasts derived from adjacent bedrock outcrops and exposures of unconsolidated sediments.  Those sediments include till, glaciofluvial sands and gravels, slope breccia and silts and clays derived from past glacial and periglacial episodes.  They are ubiquitous -- found on all of the coasts regardless or orientation of configuration.  Because they are unconsolidated, they are easily destroyed by wave action across quite a wide tidal range, and incorporated into beach materials.  Because of the Holocene sea level rise, and because the sea has been advancing at the expense of the land, there is an "accumulator effect" seen at its most spectacular in Newgale, where the storm ridge has marched eastwards for many thousands of years, incorporating glacial and glaciofluvial deposits from locations now several miles offshore.

So let's move to the raised beaches. There are plenty of them -- for example at Poppit, Ogof Golchfa, Abermawr, Broad Haven South, Ragwen Point, and Ballum's Bay.  

Cemented raised beach at Ragwen Point

I don't think I can recall a single raised beach in West Wales which is free of erratic material, and TN George, DQ Bowen and many others have made exactly the same point about the raised beaches of Gower.  There is no reason to think that the processes operating during the Last Interglacial were any different from those of the present interglacial, even though the climate may have been slightly warmer and the sea level might have been slightly higher.  There were clearly abundant glacial and water-lain deposits lying around to provide at least some of the materials required for beach building.

So why should it be assumed -- anywhere -- that because there are erratics in a raised beach there must have been an active glacier ice edge not very far away?

Coastal erratics -- St Brides Haven, Saunton and Whitesands

 

Dolerite erratics resting on a broken rock platform of old red Sandstone at St Brides Haven in Pembrokeshire. They are washed by high tides. There is no stratigraphic context, but they were most likely transported into this location by the ice of the Irish Sea Ice Stream in the late Devensian glaciation.

The Baggy Point "giant erratic" and the pink granite erratic at Saunton.  These are sealed by sandrock at the base of the local Pleistocene stratigraphic succession -- and this means they were probably in place on the rock platform before the Ipswichian interglacial.  Like the St Brides erratics, these big boulders are probably washed out from ancient glacial deposits, just like the rounded "boulder bed" boulders seen at Whitesands in Pembrokeshire.

Erratic boulders resting on the rock platform at Whitesands.  Here the raised beach is not visible, and the boulders are sealed in slope breccia assumed to date from the Early or Mid Devensian.

The MIS 4 (Early Devensian) glaciation

 

New Quay, Ceredigion.  Some rather interesting exposures....

It's interesting to see the debate about how the MIS 4 (Early Devensian) might have affected the western parts of Britain.  If you put "Early Devensian" into the search box on this blog, you will see that over the years I have revisited this issue over and again, seeking to explain details of the local Quaternary stratigraphy with reference to modelling studies and also field evidence from the south coast of Ireland, north Wales and elsewhere.

I have also highlighted this issue in some recent posts, arguing that it is a mistake to pull in evidence and suppositions about the "giant erratics" of the Bristol Channel and to pretend that they are related to MIS 4 glacial events.  I now think that these boulders are simply the last remnants of ancient glacial deposits, visible by chance on "washed" rock platforms near present sea level and also found at higher altitudes (60m, 80 m and even higher) above the modern cliffline.

But the evidence from southern Ireland (Courtmacsherry and Fethard, for example) is important and persuasive, and we have to accommodate it in whatever narrative we develop.

At the moment, in West Wales, we have not much to go on,  apart from evidence of two (at least) glaciations -- one before the Ipswichian Interglacial and the other later, in the Late Devensian.  Key evidence of a pre-Ipswichian glaciation comes from Black Mixen, near Lydstep, and Ceibwr on the north Pembrokeshire coast.  Look them up using the search box.

The only evidence of two recent (Devensian) tills separated by evidence of a melting interlude comes from New Quay -- and I am sure that one of the tills is the Irish Sea Till and the other is a till derived from the Welsh Ice Cap.  Could one of the tills date from MIS 4 and the other from MIS 2?  I'm not sure about that, since that would require a "non-glacial interval" of around 40,000 years.

Many researchers in the New Quay - Llanon area have been confused by the sequence of deposits. At New Quay I recorded a lower Irish Sea till and a local till above it -- the two separated by evidence of a warmer non-glacial interval.  Others, including Eddie Watson, recorded at Llanon a "basal" Welsh till overlain by Irish Sea till.  There is much discussion in the literature about a "Llansantffraid soil" which might be interglacial or interstadial, and abundant cryoturbation and other features indicative of intense permafrost conditions. During his lifetime I disagreed on many things with Prof Dai Bowen, but I do agree with him that there seems to have been an oscillating ice front on the Welsh coast in the New Quay - Llanon area during the Devensian........ but which till deposit (if any) dates to MIS 4?  And can it be correlated with the MIS 4 glacial deposits of Southern Ireland?

My interpretation (1965) of the stratigraphic sequence on the south 

coast of Ceredigion.

Quote from one of my earlier posts:

It's interesting that Etienne et al (2005) refer to another exposure at Llannon where a shelly Irish Sea till overlies "subglacial tills of Welsh provenance". This site is also described in the "Quaternary of Wales" volume edited by Campbell and Bowen, on p 65. They refer to a Welsh till overlain by an Irish Sea till, and argue for a complex interaction between Welsh Ice and Irish Sea ice along the coast of Cardigan Bay. They do not quite resolve the issue of the intense periglacial disturbance of the surface of the Welsh till, which seems to have taken place prior to the laying down of the Irish Sea till. Was there a prolonged periglacial episode with deep permafrost between the two glacial phases?

I am also very sceptical about the raised beaches of West Wales being interpreted as "cold climate" features just because they have erratics in them.  As far as I can see, all of the raised beaches contain erratics of various sizes, and that is not surprising since they are formed from whatever deposits happen to be available for wave attack and the removal of fines.  Raised beaches are essentially "lag deposits" reworked by the action of tides and waves.  Besides, faunal and microfossil evidence suggests that the bulk of raised beaches are genuine representatives of warm or interglacial climates.

This is all rather interesting.  I am looking at my West Wales stratigraphy to see whether there are sites that will be susceptible to "alternative" explanations as to the chronology and the severity of the glacial / periglacial climate.  One problem is that most of the sections I have examined over the years are in coastal valley locations, where a thick accumulation of pseudo-stratified slope breccia does not necessarily indicate a long period of climate dominated by permafrost and freeze-thaw processes. Rockfalls and scree development continues in such locations even when the climate is relatively mild and one has to be very careful about "climatic indicators".....

Much thinking going on........

==========

PS.  Let's not forget that there has been a considerable debate in Eastern England about a pre-LGM glaciation (assumed to be from the Early Devensian) that was more extensive than the LGM.   Our late colleague Allan Straw was instrumental in proposing this to other specialists:

http://www.emgs.org.uk/Mercian/Mercian%20-%20Individual%20papers/Mercian%20Geologist%20volume%2019%202016-2019/Mercian%202019%20v19%20p207%20MIS%204%20glaciation%20in%20eastern%20England,%20Straw.pdf

See also Prof Peter Worsley:

Worsley, P., 1991. Possible early Devensian glacial deposits in the British Isles. 47-51 in Ehlers, J., Gibbard, P.L., & Rose, J. (eds.), Glacial Deposits in Great Britain and Ireland. Balkema: Rotterdam.

Cwm Cerwyn, Preseli

 

 

Thanks to Hugh360 for this excellent image of Cwm Cerwyn, seen from the upland ridge near Cerrig Marchogion. If you ignore that traces of man-made quarrying, you can see clearly that this is an incipient cwm or cirque, never fully developed but nonetheless showing signs of active ice here.  I think the mound which is visible just left of centre is made of bedrock, but it will reward further research......

I know there is a till sheet here, but I have never managed to discover the till thickness. I am not sure whether we can refer to a real terminal moraine marking the edge of a small glacier........

The last glacier here might have been in the Younger Dryas (Zone III).   But in this location, on the margin of glaciation, I can envisage a small glacier coming and going on many occasions during the Quaternary. 

Ice-rafted erratics, Norfolk

Palaeo-rivers and site locations in the southern North Sea.  The theory is that glaciers from Scotland and Scandinavia flowed into the northern part of the basin, and that there was a floating / calving ice front from which icebergs transported debris southwards, to be dumped or dropped on tidal or freshwater mud flats........

This is an interesting paper, dealing with a number of small erratic clasts (up to 15 kg in weight) found in sediments on the Norfolk coast.  These are of course very different from the giant erratics found on the coasts of Devon and Cornwall, and the geomorphological / sedimentological setting is very different too.  But an interesting paper, reminding us that there were some very old glaciations and that erratics were carried by glacier ice -- and also in mobile icebergs in deep water -- southwards towards the Straits of Dover.  So our thoughts turn to the strange erratics found on the coasts of the English Channel -- might they have come from the North Sea at a time of substantially higher RSL?

Something to think about........

 https://nora.nerc.ac.uk/id/eprint/14695/1/Larkin_et_al.pdf

Nigel R. Larkin, Jonathan R. Lee & E. Rodger Connell. 2011. Possible ice-rafted erratics in late Early to early Middle Pleistocene shallow marine and coastal deposits in northeast Norfolk, UK. Proc. Geol. Assoc. 

Abstract

Erratic clasts with a mass of up to 15 kg are described from preglacial shallow marine deposits (Wroxham Crag Formation) in northeast Norfolk. Detailed examination of their petrology has enabled them to be provenanced to northern Britain and southern Norway. Their clustered occurrence in coastal sediments in Norfolk is believed to be the product of ice-rafting from glacier incursions into the North Sea from eastern Scotland and southern Norway, and their subsequent grounding and melting within coastal areas of what is now north Norfolk. The precise timing of these restricted glaciations is difficult to determine. However, the relationship of the erratics to the biostratigraphic record and the first major expansion of ice into the North Sea suggest these events occurred during at least one glaciation between the late Early Pleistocene and early Middle Pleistocene (c. 1.1−0.6 Ma). In contrast to the late Middle (Anglian) and Late Pleistocene (Last Glacial Maximum) glaciations, where the North Sea was largely devoid of extensive marine conditions, the presence of far-travelled ice-rafted materials implies that earlier cold stage sea-levels were considerably higher.

=========================

6. Conclusions

Concentrations of erratics within WCF (Wroxham Crag Formation) coastal deposits at Sidestrand and West Runton in northern East Anglia are considered the product of melt-out from (possibly grounded) icebergs.

The provenance of the erratics implies that these icebergs were derived from glaciers that were eroding bedrock in the Southern Uplands, Midland Valley and southern Grampian Highlands of Scotland, and Oslofjord in southern Norway.

The age of these erratic-bearing beds can be broadly constrained to a period from the late Early Pleistocene to early Middle Pleistocene interval (c. 1.1–0.6 Ma, a time period that spans the ‘Menapian’ (MIS 34)) to late ‘Cromerian Complex’ (MIS 16) stages.

These erratics demonstrate both the existence of restricted glaciations in Scotland and Norway, and their periodic expansion into the North Sea Basin prior to the maximum extent of the ice sheets during the Anglian Glaciation (MIS 12) of the Middle Pleistocene.

This research supports the work of Sejrup et al. (1987) and Ekman (1999) that argues that both the BIIS and SIS were active in the North Sea Basin on at least one occasion well before the Anglian stage of the Middle Pleistocene.

The deposition of the erratic-bearing beds during these early glaciations appears to coincide with higher glacial sea-levels than occurred during the late Middle and Late Pleistocene.

==============

https://www.northfolk.org.uk/geology/erratics.html

The formation of the Straits of Dover comes into the frame here too:

https://brian-mountainman.blogspot.com/2017/04/the-creation-of-straits-of-dover.html

Gibbard, P. Europe cut adrift. Nature 448, 259–260 (2007).
https://doi.org/10.1038/448259a

The initiation of the megaflood responsible for the cutting of the Straits of Dover.  Note that in the North Sea the water level is assumed to have been substantially higher that that in the proto-English Channel.  Note also the large icebergs running aground and dropping erratics!

The Early Devensian (MIS 3) Dilemma

 

Thanks to Prof Ian Shennen for drawing this to my attention.  It's a very interesting article from 2021, drawing together vast amounts of data to give us a picture of absolute sea level change (not RSL !!) over the last 80,000 years -- the approx span of the Devensian.

Evan J. Gowan, Xu Zhang, Sara Khosravi, Alessio Rovere, Paolo Stocchi, Anna L. C. Hughes, Richard Gyllencreutz, Jan Mangerud, John-Inge Svendsen & Gerrit Lohmann. 2021. A new global ice sheet reconstruction for the past 80 000 years. Nature Communications 12, 1199. https://doi.org/10.1038/s41467-021-21469-w

Abstract
The evolution of past global ice sheets is highly uncertain. One example is the missing ice
problem during the Last Glacial Maximum (LGM, 26 000-19 000 years before present) – an
apparent 8-28 m discrepancy between far-field sea level indicators and modelled sea level
from ice sheet reconstructions. In the absence of ice sheet reconstructions, researchers often
use marine δ18O proxy records to infer ice volume prior to the LGM. We present a global ice
sheet reconstruction for the past 80 000 years, called PaleoMIST 1.0, constructed independently
of far-field sea level and δ18O proxy records. Our reconstruction is compatible with
LGM far-field sea-level records without requiring extra ice volume, thus solving the missing
ice problem. However, for Marine Isotope Stage 3 (57 000-29 000 years before present) - a
pre-LGM period - our reconstruction does not match proxy-based sea level reconstructions,
indicating the relationship between marine δ18O and sea level may be more complex than
assumed.

The tail end of the Ipswichian interglacial is represented at the left edge of the diagram.  Then we can see an Early Devensian cooling with ice sheet development in MIS 3, a milder episode in MIS 4 with some ice wastage, and then the last great surge in ice sheet growth during the LGM of MIS 2.  Then follows the Holocene, with ice wastage and the return of vast quantities of meltwater into the oceans, with sea level returning to its interglacial level.  We can assume that these three graphs are pretty accurate, since they combine modelling and empirical data.  

The authors point out that the only serious mismatch occurs in MIS 3, where observed sea levels appear to have been lower than the models suggest they should have been.  This might be due to problems in extrapolating from oxygen 18 levels to assumptions about sea level positions -- but no doubt that problem will be ironed out in due course.

The nice thing about these curves is that they fit very nicely into the climate change scheme which I worked out in 1965 on the basis of the West Wales Quaternary sediment sequence.  I worked out that the Ipswichian high sea level was followed on the Pembrokeshire coast by a long period of lower relative sea-level during which there were many oscillations of periglacial climate with slope breccia accumulations.  Then came the onset of full glacial conditions in the LGM, with ice affecting all parts of the county.  The Irish Sea till horizon was the definitive stratigraphic marker.   I was quite sure that whatever was going on elsewhere in the Early and Mid Devensian, glacier ice did not affect Pembrokeshire.

The full glacial cycle at Aber-mawr, which I described in 1965.  The Ipswichian raised beach was not visible at Aber-mawr at that time.  The "main head" episode, with a number of facies of slope breccia, was thought to represent the Early and Middle Devensian (MIS 4 and MIS 3).

So far so good, until geomorphologists working in SE Ireland started to discover evidence of a substantial ice load just on the other side of St George's Channel, sufficient for a good deal of crustal depression, in MIS 4.  Of which more in due course.........

Raised marine terraces, Gruinard Bay, Wester Ross

 

This is a great picture of two raised marine terraces (we might call them raised beaches) in Gruinard Bay in northern Scotland.  This is on the eastern shore of The Minch,  in an area which we might expect post-Devensian isostatic recovery to be still going on.  The marine limit here is not much above 20m. There has been much greater isostatic uplift further south, between Fort William and the Firth of Clyde, where ice sheet crustal load was at its greatest.   The two terraces (above the current HWM) are formed mostly of outwash sands and gravels from the Little Gruinard River.  I am not aware of any detailed geomorphology studies in the literature.

Incidentally, it is postulated that the pink granite erratic at Saunton (on the N Devon coast)  might have come from this location  -- but having looked at the geology map I am rather sceptical about that.........

The Bristol Channel: lost in a fog of MIS-information

The Irish Sea Ice Stream "assault" on the southern coasts of the Bristol Channel, modified after Gilbertson and Hawkins.  The caption refers to the Anglian Glaciation, but it now looks as if this was actually the situation during the Late Wolstonian Glaciation.

We are all lost in the fog, looking down into the murky depths of the Bristol Channel and trying to navigate around ice floes bearing giant erratics..........

Things are very murky indeed, and I suspect that all of us who have been involved in the discussions about the Bristol Channel (James Scourse, Nick Pearce, Rob Ixer, Chris Rolfe, Richard Bevins, Phil Gibbard, Olwen Williams-Thorpe, John Hiemstra, Danny McCarroll, Ed Lockhart and myself, to mention just a few) are more than a little confused, because we don't really know what to make of the evidence.   So in the mix we find field observations, assumptions and speculations relating to ice rafting, the positions of giant erratics, isostatic and eustatic interactions, tectonic uplift, ice edge positions, glaciological and climatic factors, the reliability of OSL and other dating methods, and much else besides.  

Part of the difficulty that we all face arises from the conflation of a number of different issues.  So arguments about ice rafting and relative sea level (RSL) positions and the interpretations of sea floor sediments  have got tangled up in arguments about the likelihood of expansive glaciation in the Early and Mid Devensian (MIS 3 - MIS 4).  Yes, these things are interrelated, but the "Devensian debate" is a hugely complex one, and it is made much more complicated than it needs to be by the obesssion that some people seem to have with the ice rafting of big erratics and the necessity of working out a mechanism for big dirty icebergs to afflict the Devon and Cornwall coasts with RSL more or less where it is today.

The pink erratic at Saunton, sealed beneath sandrock and tied into the stratigraphic sequence. This boulder (reputedly from Gruinard Bay, Wester Ross, in the north western highlands of Scotland) was probably present on the rock platform prior to the Last Interglacial. Other "freestanding" erratics might have been carried into this locality during the Late Devensian glaciation.

This, if I may say so, is a sterile debate because it is completely unnecessary unless you are intent on denying that glacier ice has, on one occasion or more, come into contact with the southern shores of the Bristol Channel.  Following the publication of the latest comprehensive study by Bennett et al (2024), it is incontrovertible that the ice of the Irish Sea Ice Stream has come into contact with the cliffline and pressed inland into embayments, all the way from the Bristol Avon to Land's End.  As James Scourse has said, some of the field evidence is unequivocal and some is equivocal.  That's no big deal, and it can also be said of all of the coasts of Wales.  Large stretches of the coastline of western Pembrokeshire are apparently devoid of evidence of glaciation, and the best evidence of Quaternary sedimentation is found in embayments and coastal valleys protected from the effects of coastal erosion and cliffline retreat in a high energy storm wave environment.  This does not mean that just a few isolated parts of the coastline were overridden by glacier ice or affected by lobes of ice  projecting beyond the position of the glacier front.  Lobes only exist where there are channels and topographical constraints.

Common sense (and glaciology) dictates that if Lundy and Flatholm were overridden by streaming ice, and if the Isles of Scilly were similarly affected, the west-facing and north-facing clifflines of Cornwall, Devon and Somerset must have presented a massive barrier which prevented the ice from moving far inland.  Of course, at the time this coastal tract was far above RSL, and the cliffs must have been partly buried by banks of scree and landslide debris -- but they must still have been high enough to hinder ice ingress towards Exmoor and the uplands of Cornwall.

The cliff rampart on the North Devon coast, near Woody Bay

The simplest explanation of the "giant erratics" resting on the coastal rock platform and also at higher altitudes is that they are glacially transported and that they are exposed by washing processes up to, and some way above, high water spring tides (HWST).  At higher altitudes (maybe up to 100m in places) the erratics are sometimes exposed at the surface but are still, for the most part, buried in till and slope deposits.  This is exactly the situation in the Fremington - Croyde - Saunton area.

Some of the boulders are sealed beneath possible Ipswichian and probable Devensian deposits, and so they must have been transported and emplaced by Wolstonian or earlier glaciers.  Other boulders which have no stratigraphic context (like the Limeslade boulder or the Flatholm pink granite boulder) might just have been carried by Devensian ice........

If one sees some merit in Occam's Razor and accepts the above narrative, discussions about tectonics, isostatic loading effects and RSL positions become immaterial, and we can move on.

My latest reconstruction of the Late Devensian glaciation of South Wales and the Bristol Channel region.  I would not be at all surprised if some of the deposits of the Fremington - Croyde area tirn out to be of Late Devensian age.

=========

So -- what about an Early and Mid Devensian glaciation?  Ah, that is another matter entirely, which deserves another post.....

The myth of ice-rafted coastal erratics

Features seen on Höga Kusten, Sweden, around and beneath the 260m marine limit.  Kalottberg hills are "capped" with moraine and glacial and fluvioglacial deposits, providing excellent conditions for forest cover to survive.  Beneath the highest strandline surfaces are washed -- which means that most of the fines have been washed out during isostatic uplift.

Beneath the highest shoreline in the Höga Kusten district, the finer materials in the morainic cover are washed out, leaving massive accumulations of boulders and smaller stones. The term "Klapperfält" is used to describe them.. Most of the boulders are local, but scattered among them are far-travelled erratics, coloured orange in this diagram.  In areas such as this, stillstands or pauses in the rate of uplift may result in the formation of storm-beach ridges.

Erratics of many different lithologies on washed rock surfaces in the Stockholm Archipelago.

Washed surfaces, kalottberg and klapperfält features can be found anywhere within the dark coloured area on this map.

It's intriguing that ice-rafted erratic boulders are still being talked about in the context of the large erratics found on the shorelines of southern Britain.  Ice rafting seems to be the "process of preference" among those who -- for whatever reason -- do not wish to accept that full glacial processes have substantially affected the Bristol Channel coasts of Cornwall, Devon and Somerset.  

About a year ago I posted these items on the blog: 

https://brian-mountainman.blogspot.com/2023/12/coastal-geomorphology-in-devensian.html

https://brian-mountainman.blogspot.com/2023/12/the-big-erratic-boulders-of-devon-and.html

........... and I have discussed the matter in many other posts as well.  Use the search box to find them.

The latest spat is with James Scourse, who argued in a very strange paper that the big coastal erratics were emplaced by floating ice at a time of high relative sea level (RSL) and cold climate.

https://brian-mountainman.blogspot.com/2024/03/new-scourse-paper-on-giant-erratics-is.html

That of course flies in the face of our understanding of eustatic / isostatic relationships during the Devensian and earlier glacial episodes.  In the latest rather crude attack on me by Pearce, Ixer, Bevins and Scourse they argue again that the big coastal erratics (like the Limeslade erratic) on the Bristol Channel coasts are "ice rafted"  -- and that there was "asynchrony between Early and Middle Devensian regional ice sheet development and global sea level".   I don't have a clue what that means, and no evidence in support of the contention is provided.  I see no sign at all of Early and Middle Devensian glacial activity in the Quaternary stratigraphy of West Wales, and I see no evidence of substantial isostatic depression at the time, or tectonic tilting.  Bennett et al, in the new paper on the Quaternary in Devon, say with respect to the coastal erratics: "...........another view is that the coastal platform erratics were delivered by icebergs calved from glacier ice farther north, an interpretation that is consistent with the widespread distribution of erratics on shore platforms on the Bristol Channel and English Channel coasts."

Even if there was a substantial ice mass in Wales during the Early and Mid Devensian, global sea level at that time was around -70m, and I can see no mechanism for a sinking of the Devon and Cornwall coast by an equivalent amount, thereby creating conditions for ice rafted debris to be dumped on the coastal rock platform.  On the contrary, when one looks at known isostatic adjustment mechanisms, there would have been a hinge line along the Bristol Channel and a rise in the relative altitude of the South-West Peninsula, making giant erratic emplacement from floating ice even less likely.

Icebergs do not, in general, deliver erratics to distant coastal platforms. Dirty icebergs moving away from floating ice fronts almost always operate in deep water, dropping their debris load as a consequence of bottom melting -- ending up as glacio-marine sediments with added dropstones. There are vast thicknesses of these materials in the Celtic Sea and in the outer Bristol Channel. (They have received impressive attention during the BRITICE studies.)  Icebergs do get stranded in embayments in the intertidal zone, and they do break up into bergy bits and release big boulders onto mud flats.  But most of the ice with "rafting" characteristics is sea ice which develops as flat sheets and then gets broken up by lateral pressure, and broken ice floes are almost always clean.  Where there is an ice foot on a polar coastline,  it is almost always a mixture of sea ice, frozen spray and detached chunks of winter snowbanks.  It sometimes contains erratic boulders, but these are generally taken from pre-existing coastal materials including beach accumulations, till and scree.  They are seldom carried in from the sea; they are more likely to be carried OUT from the coastline to be dropped in deeper water offshore. 

As I see it, the big erratic boulders scattered along the coasts of the Bristol Channel are NOT all concentrated in the current intertidal zone.  They occur at altitudes up to at least 80m.  They have to be residuals or remnants of ancient glacial deposits scattered along the Bristol Channel coasts.  As I have said before, the "concentration" of these erratics on the intertidal rock platform is more apparent than real, since these are washed surfaces similar to those associated with "kalottberg" hills around the marine limit in Sweden.  Nobody ever suggests that the boulders found at the marine limit are unique in any way;  they almost always also occur above and below the limit.  So the position of the marine limit -- like RSL on the coast of the Bristol Channel -- is interesting but not geomorphologically significant.

https://brian-mountainman.blogspot.com/2018/07/kalottberg-features-in-northern-sweden.html

It seems to me highly likely that the erratic boulders on the Bristol Channel coasts are simply the remnants of ancient glacial deposits that have been almost entirely removed on the coast by marine, periglacial and other processes, possibly over several glacial / interglacial cycles. Further inland, away from the coast, there may well be in situ glacial deposits (with giant erratics!) still awaiting discovery.

 The boulders themselves may date from a number of different glacial phases, including the Anglian, Late Wolstonian and Late Devensian, in which Irish Sea ice from the N and NW crossed the coastline and pressed inland via convenient depressions and wide valleys.   Just a few of the boulders (such as the famous pink erratic at Saunton) can be tied into the stratigraphic sequence,  sealed by sandrock or slope breccia.  At Whitesands in North Pembrokeshire  the boulder bed (which I think is related to some of the erratics on coastal rock platforms) appears to consist of boulders emplaced prior to the last interglacial and then modified by wave action.  It is not visibly associated with the Ipswichian raised beach. 

Some of the boulders of the Whitesands boulder bed, held in cemented slope breccia at the base of the Devensian sediment sequence.  The boulders were probably emplaced during an early glaciation and rounded by wave action during the Ipswichian interglacial.  This narrative probably applies to many of the Bristol Channel "giant erratics"........

Looking again at the ice-rafting hypothesis, I cannot see any tectonic or other evidence which suggests that floating ice was involved in the emplacement of any of the big erratics on the Bristol Channel coast.  On those occasions when there were dirty icebergs in the Celtic Sea, sea level was probably at least 100m lower than today, and the coastline was at least 10 km further to the west than it is now.  

One from the archive.  A "boulder bed" (including some well rounded ones) on a washed surface -- in part of the raised beach sequence on King George Island, South Shetlands.  Photo from 1965.   The marine limit in this area is at c 275m.  There is no reason to believe that |"ice rafting" had anything to do with the presence of these boulders.  They have all come from in situ glacial deposits.

Ice cliff resting directly on the current storm beach, at the eastern end of South Beaches, Byers Peninsula. Livingston Island.  The ice is at the edge of a glacial dome called Rotch Dome -- it does flow seawards, but not very fast.  Some of the ice is dirty, so debris is dropped onto the beach in frequent cliff collapses. We walked along this beach on one occasion, to look back and see a large ice collapse which obliterated the beach across which we had just walked.......

More support for a big Wolstonian glaciation

 

I have been looking again at the big article by Bennett et al (2024) on the Quaternary in Devon.  As indicated on the table above, they believe that the one glacial episode that substantially affected the North Devon coast occurred during the Late Wolstonian (MIS 6).  They consider that the glaciation of the Bristol Channel during the Late Devensian (MIS 2) was less extensive, although they accept that the evidence for that time demonstrates a local ice cap on Dartmoor. (There must, of course, also have been a Dartmoor ice cap in the Late Wolstonian, but the evidence for it is not easy to find.........)

As for the Anglian (MIS 12), the authors are cautious.  They suggest that the evidence for a big glaciation at that time is not clear in and around Devon, and therefore they prefer to stick with a "periglacial" climate label for that period. 

Commendable work, with great relevance for other parts of SW Britain. 

Quote:

The most recent research on glaciation in the Bristol
Channel was published by Gibbard et al. (2017, 2022). This
included a synthesis of new publicly available borehole and
bathymetric data, combined with a wealth of other existing
disparate data sources. Sediment boreholes throughout
the Bristol Channel confirm the area was glaciated in the
Pleistocene. Till is present below marine deposits and, in some
areas, is visible morphologically as submerged moraines (Fig.
9). In the central and eastern Bristol Channel the submerged
valley course of the palaeo-Severn is very clear in new highresolution
bathymetric surveys. This former river course and
associated tributaries cross-cut through glacial sediments in
the Bristol Channel. At least three phases of glaciation are
recorded in the Bristol Channel, one related to the southern
limits of a Late Devensian Substage (~MIS 2) Welsh Ice Cap
which reached into Swansea Bay, an earlier Devensian (MIS
4 to 3) glaciation associated with Irish Sea ice, and another
older glaciation that is associated with ice that filled the entire
outer and central Bristol Channel (Fig. 1). The age of the older
Bristol Channel glaciation is still open, although it predates
the Devensian (Late Pleistocene) and must date to the Middle
Pleistocene, and as noted above is likely to correlate with MIS
6 (i.e., the Late Wolstonian). This has implications for the age
of the glacial deposits present on land in the Barnstaple area,
which have traditionally been correlated with the Anglian Stage
(MIS 12) (e.g., Croot et al. 1996). It is more likely, however,
that these deposits correlate with the oldest glaciation of the
Bristol Channel immediately to the north, i.e., with MIS 6 and
the Late Wolstonian Substage (potentially the Moreton Stadial
of Gibson et al., 2022).

I am unconvinced about the Early Devensian  (MIS 4-3) glaciation associated with Irish Sea ice, since the evidence is just not there -- and  I suspect this idea will be dumped before too long.......

The Quaternary Geology of Devon

Glacial deposits and erratics in the Fremington area.  Note that the till cover extends over the 60m contour.  It is therefore not surprising that "high level erratics" are found at even greater altitudes elsewhere on the Devon coasts.

Been waiting for this for some time.  A very comprehensive and interesting summary.  I'm intrigued by the  suggestion that glacier ice "approached" the North Devon coast on at least three occasions from the Bristol Channel, and probably crossed it more than once.  The authors are in no doubt about the presence of glacial deposits in the Fremington area.  I would have liked a more careful consideration of the ice rafting hypothesis in the section where they talk about erratic boulders.  They say that the rivers that cut the deep river valleys were graded to low, cold period sea levels, but if that was the case you cannot have high sea levels with ice rafting of erratics when conditions were of "full glacial" status.  You can't have your cake and eat it.

Anyway, I'll give this useful article greater consideraton on another occasion.

===========

Bennett, J.A., Cullingford, R.A., Gibbard, P.L., Hughes, P.D. and Murton, J.B. 2024. The Quaternary Geology of Devon. Proceedings of the Ussher Society, 15, 84-130.

https://ussher.org.uk/journal/catalogue/volume-15-part-3-2024

Abstract

Throughout the 2.6 My of the Quaternary, Devon has occupied a critical position with respect to the evolution of Britain in that it lies close to the North Atlantic Ocean between the southern coast of the Bristol Channel and the northern coast of the English Channel. This setting results in the area being highly sensitive to climatic and environmental change. Although the county lies beyond the general limit of the major glaciations of the last 0.5 My, it was impacted by glacial ice and its associated meltwater that approached the north coast on at least three occasions. Glaciers also left deposits on the Bristol Channel floor and potentially locally on Dartmoor and possibly Exmoor. Ice-rafting of erratic rocks also occurred. Nevertheless, the whole region has been repeatedly subjected to severe cold-climate, periglacial conditions for much of Quaternary time. Under cold periods frost- dominated climates have driven the formation of a thick carapaces of slope-derived debris (head deposits), with wind erosion shaping the tors characteristic of the high moorlands. At the same time deep river valleys have been carved by continual severe seasonal snow meltwater, the streams transporting weathered rock derived from the steep slopes. The rivers that cut these valleys were graded to low, cold period sea levels, and their valleys frequently include multiple terrace accumulations, the highest of which date from the Middle Pleistocene. Apart from the Late-glacial open vegetation and Holocene blanket peatland, evidence of earlier warm-climate conditions also occur, but are limited in comparison to the cold-climate accumulations. Past and present interglacial (temperate) deposits are also known from the county. They include not only river deposits, but also cave infillings and high sea-level, fossil beaches, often including fossil evidence. Coastal erosional landforms such as wave-cut platforms and cliff formations are commonly found, and submerged offshore examples are also known, particularly from the English Channel coast. Despite all this evidence, its fragmentary nature means that the record of events in the region is both limited and intriguing. The advent of numerical age determination and other modern analytical methods have improved knowledge of the timing of some events and provenance of materials in the region, but much more work is required to fill in the substantial gaps in current knowledge.

Pearce et al: Much ado about nothing much

The map published by Scourse et al, 2021. The labels used for the glaciers are no more "correct" than the labels that might be used by others on their own maps.  The ice edge positions shown are also matters of opinion, as are the directions of ice flow........ and I happen to disagree with several of the key features on this map.

With regard to that petty and mean-spirited ad hominem attack the other day from Pearce, Ixer, Bevins and Scourse, one of the things that most intrigued me was the obsession with silly little details.  

One comment (with the use of "sic") related to my use of the term "Irish Sea Glacier" somewhere in the text, whereas I should apparently have used the term "Irish Sea Ice Stream" as defined by Scourse and many other colleagues in 2021.  Big deal.  Well yes, that is the term used for many years now to describe the ice mass flowing in the Celtic Sea arena and into the Bristol Channel.  For many years I have used the term "ice stream" myself, in this blog and in publications.  But the term "Irish Sea Glacier" is and has been used widely across the literature for many years, and is widely understood as the ice mass that extended all the way out to the Celtic Sea shelf edge.  An ice stream is simply a big glacier, flowing fast and flanked by ice that is stagnant or flowing much more slowly.  A glacier might be described as an ice stream channelled in a trough or bounded by topographic highlands.  Pearce et al apparently want the rest of us to restrict the term "Irish Sea Glacier" to the ice that flowed into the Cheshire Basin, flanked by the uplands of North Wales and the southern Pennines.  But it was no more constrained by topography than the "Irish Sea Ice Stream"  that flowed through St Georges Channel -- and I think that the term "ice lobe" better describes the characteristics of the ice that flowed into Cheshire and the north Midlands.  

It may also be argued that the use of the term "Irish Sea Ice Stream" for the ice occupying the Celtic Sea is inadequate, since the feature (possibly during several glaciations) had many of the characteristics of a piedmont glacier, as I pointed out in one of my earliest glaciology articles, in 1968:

https://brian-mountainman.blogspot.com/2011/02/glaciological-dilemma.html

https://brian-mountainman.blogspot.com/2016/07/ice-in-celtic-sea-piedmont-glacier-or.html

https://www.researchgate.net/publication/319922666_Short_Notes_Directions_of_Ice_Movement_in_the_Southern_Irish_Sea_Basin_During_the_Last_Major_Glaciation_An_Hypothesis

Thus there is no "correct" terminology in any of this.  It is disingenuous of Pearce et al to pretend that there is, and that others are "in error" if they use terms that do not conform to somebody else's labelling system.

Reference:

https://brian-mountainman.blogspot.com/2021/05/new-dating-for-lgm-irish-sea-ice-stream.html

Maximum extent and readvance dynamics of the Irish Sea Ice Stream and Irish Sea Glacier since the Last Glacial Maximum
J. D. Scourse, R. C. Chiverrell, R. K. Smedley, D. Small, M. J. Burke, M. Saher, K. J. J. Van Landeghem, G. A. T. Duller, C. Ó Cofaigh, M. D. Bateman, S. Benetti, S. Bradley, L. Callard, D. J. A. Evans, D. Fabel, G. T. H. Jenkins, S. McCarron, A. Medialdea, S. Moreton, X. Ou, D. Praeg, D. H. Roberts, H. M. Roberts, C. D. Clark
Jnl of Quaternary Science, 7 May 2021 (special issue article)

https://onlinelibrary.wiley.com/doi/10.1002/jqs.3313?af=R

https://doi.org/10.1002/jqs.3313

This all reminds me of a spat I had with James Scourse five years ago, following the publication of my QN article on the glaciation of the Isles of Scilly.  Scourse attacked me, using very intemperate language, partly because I had the temerity to disagree with a part of his sedimentary stratigraphic labelling system.  In his eyes, I dare say, his labels were 100% reliable, and not open to negotiation by anyone.  Anyway, after some insulting bluster and questioning of my competence, he had to grudgingly agree that the Devensian ice limit as drawn by me did actually have some merit..........

https://brian-mountainman.blogspot.com/2019/03/scourse-versus-john-rather-scilly-spat.html

Then there is the matter of the Anglian Glaciation, about which I have written at length on this blog.  In their QN piece, Pearce et al have a silly dig at my "contention" that the ice advance responsible for bluestone transport was during the Anglian Glaciation, and claim that there is no geochronological or other evidence to support this.  It's true that when I wrote my article I did mention the Anglian episode (MIS 12), but in doing that I was representing an almost unanimous view across the research community that this was the most extensive pre-Devensian glaciation across much of the British Isles.  

LEE, J R, ROSE, J, HAMBLIN, R J, MOORLOCK, B S, RIDING, J B, PHILLIPS, E, BARENDREGT, R W, AND CANDY, I. 2011. The Glacial History of the British Isles during the Early and Middle Pleistocene: Implications for the long-term development of the British Ice Sheet. 59-74 in Quaternary Glaciations–Extent and Chronology, A Closer look. Developments in Quaternary Science. EHLERS, J, GIBBARD, P L, AND HUGHES, P D (editors). 15. (Amsterdam: Elsevier.)

I am quite unconcerned about whether the Anglian was or was not the largest British glaciation, and if Pearce et al had bothered to look at my published output they would have seen that I have discussed at length the possibility of a very large Wolstonian glaciation -- and that I am rather convinced by the latest suggestions of Phil Gibbard and others that it was more important than the Anglian for landscape transformation (and sediment transport) in the Celtic Sea arena. 

https://brian-mountainman.blogspot.com/2024/08/towards-late-wolstonian-mis-6-ice-limit.html  

More to follow.......

The Berry House boulder, north of Shebbear

 

From Google streetview -- this is a hefty boulder, on the roadside verge at the entrance to Berry House