Some of the ideas discussed in this blog are published in my new book called "The Stonehenge Bluestones" -- available by post and through good bookshops everywhere. Bad bookshops might not have it....
To order, click

Tuesday 30 September 2014

Flat Holm Erratic Hunt (5): the Pleistocene context

Oblique aerial photo of Flat Holm. Lighthouse Point, at the southern tip of the island, is in the distance.  West is to the right. East is to the left. The famous pink erratic is located at the edge of the beach at bottom right (West Beach).

It's worth reminding ourselves -- in the light of the erratic hunting going on on Flat Holm and Steep Holm -- what the Pleistocene context is, according to other authors.  I'll stay out of it for the moment.......

On this map you can pick out Lavernock Point to the south of Cardiff, and you can see the location of Flatholm and Steep Holm in the mouth of the Severn Estuary, where it broadens out into the Bristol Channel proper.  Click to enlarge.  Note that there is a deep channel (over 20m deep) between the two islands.

There is a complex sea floor here, with a great deal of sediment movement, given the vast tidal range and the fierce currents that sweep in and out twice a day. 

You can see from this Landsat image just how stained the water is within the estuary -- sediment is carried well out into the Bristol Channel on every falling tide.

How many times might Flat Holm and Steep Holm have been glaciated?   The current consensus (not that we should pay too much attention to it) is that this part of the Severn Estuary has been glaciated just once -- during the Anglian Glaciation (coinciding with Marine Isotope stage 12) -- about 450,000 years ago.  This is a reconstruction of the Anglian ice limit by Pawley et al -- and in view of the fact that glacial deposits are now known from well within Somerset this may be taken as a very conservative line drawn on a map:

The authors have effectively taken the coastline of Cornwall, Devon and Somerset as coinciding with the ice edge.  Note that on this reconstruction, both Flat Holm and Steep Holm will have been affected by glacial ice at this time.

A more extreme reconstruction is that of Kellaway, supported in part by Olwen Williams-Thorpe and her colleagues in 1991.  They considered that the ice reached Stonehenge, as shown on this reconstruction:

Note that these authors have very bravely added erratic trails or erratic transport routes on this map.  I have highlighted them with colour.  As far as Flat Holm is concerned, the researchers have shown an ice stream crossing the outer end of the St David's Peninsula before pushing up the Bristol Channel and reaching the Penarth - Flat Holm - Weston-super-Mare area.

It is widely assumed that the Wolstonian glaciation affected Wales; but there is not enough evidence to enable any sort of reconstruction to be done.

When we come to the Devensian glacial episode, the limit most frequently cited is that shown by Pawley et al as a stippled area on the map above.  That limit as published by the Geological Survey was inaccurate, and we now know that Devensian ice affected both Lundy Island and Caldey Island.  But as shown on the following map, it is unlikely that any Irish Sea ice or Welsh ice affected Flat Holm around 20,000 years ago.

That having been said, the ice edge may not have been very far away, ie just inland of Cardiff -- so there would have been a strong possibility of meltwater streams from the ice edge carrying down onto the floor of the estuary large quantities of debris from the eastern part of the South Wales Coalfield.  According to some reconstructions, the ice front extended beyond the present coastline between Cardiff and Newport.  Remember that the sea was not present at the time -- it was still more than 100m lower than its present level.  So a mechanism existed for the introduction of large quantities of fluvio-glacial and glacial material from the north into the vicinity of the hill mass which we now refer to as Flat Holm.


A note about the sediments on the bed of the estuary:

Today, the estuary is known as a ‘drowned valley’. The bedrock floor of the Severn Estuary upstream of a line from Lavernock Point to Brean Down is covered with areas of gravel, sand and mud. These also cover bedrock in the Bridgewater Bay area, but downstream of the Lavernock-Brean Down line, bedrock is exposed on the floor of the Estuary9. Movement of sediment within the Severn Estuary is complex due to the high tidal range and the strong tidal currents that prevail here. Not only are silt and mud carried seaward into the Estuary by the rivers, they are held in the Estuary, deposited and moved again over considerable periods of time. In addition, sand is carried up the Estuary and landward9. This sand is not derived from erosion of the shoreline rocks, but from the deposits of glacial rivers during the last ice age. The surface layers of the sandbanks are largely unconsolidated and may be moved and redeposited by the strong tides. Some of the near-shore sandbanks provide an element of protection against coastal erosion by reducing wave energy. The dredging of sand for the aggregates industry is therefore closely monitored.

Ref 9:  Kellaway, G.A. & Welch, F.B.A., 1993. Geology of the Bristol district. Memoir for 1:63 360 geological special sheet (England and Wales). London: HMSO, xii+199pp.

Flat Holm Erratic Hunt (4): previous records

Erratic pebbles have been known on Flat Holm for almost 200 years, and the earliest record seems to be that of Buckland and Conybeare in 1824.  They recorded pebbles of "chalk-flint, red sandstone, quartz, flinty slate and porphyry....." on the limestone bedrock and on the beaches which effectively double the size of the island when the tide is low.

The island is in the middle of the Bristol Channel, 4.7 km from Lavernock Point near Penarth on the Welsh side and almost 10 km from Weston-super-Mare on the English side

In 1971 Geoffrey Kellaway collected 100 pebbles from East Beach.  This will not have been a totally random collection -- presumably he just picked up the things he thought were interesting.  So any statistics and percentages cited are meaningless.  Nonetheless, the list of pebbles from this collection, as identified by RK Harrison, is interesting.

Laminated cherty dolomitic siltstone (E40285)
hornfelsed siltstone (E40286)
spicular chert (E40288)
dolomite pseudo-microbreccia (E402910
graphic microgranite (E40289, E40293, and E40294)
feldspar porphyry (E40287)
fluxioned feldspar-rich lava (E40290)
silicified welded crystal lithic tuff (E40292)

Presumably thin sections were made from all these pebbles.  The remaining -- maybe less interesting -- rocks were:

jaspers and other fine-grained quartzose rocks, dark grey to purple (24%)
purple-brown quartzitic sandstone (Old Red Sandstone?)  15%
Cretaceous flints (13%)
fine-grained and indurated quartzites (? Precamrian)  11%
Carboniferous limestone and dolomite 10%
coarse and medium grits 10%
purple and green andesite (?) 5%
 grey quartz-feldspar-porphyry - less than 5%
?obsidian - trace
vein quartz and other igneous rocks - trace

Harrison considered that there were two groups of pebbles -- relatively local and far-travelled.    The former were readily distinguished -- eg limestones, flints, dolomites, purple-brown quartzitic sandstones.  The latter were "highly variable in lithology, and provenances could not be given with any certainty.  Some -- eg vein quartz and Precambrian (?) quartzites -- might have been secondarily derived from ORS.  The andesites and welded tuffs might have come from North Wales or the Lake District.  The graphic microgranite might have come from Northern England -- for example, Carrock Fell.

The non-local material, thought Harrison, must have come from a glacial deposit, probably on the bed of the estuary.  This is reasonable, given that remnants of glacial deposits exist onshore in the area to the east of Flat Holm  (Gilbertson and Hawkins, 1977).  Also, the gravel spit that forms a spectacular low-tide feature at the eastern end of Steep Holm, although consisting mostly of local rocks, includes flints and greensand cherts. 


Source for this information:  Ch 6 in:  Whittaker A and Green GW 1983., Geology of the country around Weston Super Mare. Mem. Geol. Surv GB Sheet 279 with parts of 263 and 295.  Thanks to Chris Lee for drawing this to our attention.

Flat Holm Erratic Hunt (3): the pink granite erratic

More on the famous pink erratic boulder on the limestone rocks of Flat Holm's West Beach.  In 1936 a sample was collected from this boulder -- E17444 -- and a thin section was made.  It was examined by Dr AJ McGregor and identified as a pink leucocratic soda-rich granite; but he was unable to offer any possible provenance.

Another analysis (of the same thin section?) by Dr Stuart Baskerville of the University of Glamorgan was communicated to Chris Lee in 2011.  The petrology was described as "igneous quartz, with some slightly sutured boundaries; microperthite (quite fresh); minor hornblende; possible biotite."

Dr Baskerville said the sample was possibly from an alkali granite, granidiorite or adamellite, but he was concerned about a lack of significant amounts of mica.  All of the feldspar was thought to be microperthite -- and this might indicate equal amounts of plagioclase and potash feldspar, pointing towards adamellite (?).  The texture was described as holocrystalline, phanerocrystalline.

As far as a possible origin was concerned, Dr Baskerville suggested that the specific granite which best fitted the bill was the (Late Pre-Cambrian?) Coedana Granite of Anglesey.  But he emphasised that this identification was highly speculative.

All other suggestions gratefully received........

Flat Holm Erratic Hunt (2): the big erratics

These are the big erratics found on the island.  The pink granite erratic is of course well known, but the strange conglomerate found by Chris on the coast between Castle Rock and North-west Point is a very unusual one which had us all flummoxed.  More info will follow when we get it.  In the meantime, are there any opinions?

The famous pink granite erratic resting on a broken limestone surface at the north end of West Beach --directly to the west of the farmhouse.

Sid and Chris examining the mysterious conglomerate boulder.

The conglomerate found on the foreshore in the NW part of the island.  It does not look like the Cambrian Basal conglomerate or the Ridgeway Conglomerate of Devonian age -- one guess is that it might come from the Millstone Grit series.

And a false alarm!  I was quite excited when I saw this boulder on West Beach -- almost black in colour, very hard and heavy, and with a dense rectangular fracture pattern....  I thought it was basalt, but then realised that there was an outcrop not far away of incredibly hard (silicified) chert within the limestone series.  Great care is needed in erratic hunting......

Flat Holm erratic hunt -- a grand day out (1)

 Danger -- geologist at work.  Sid hunting for clues on Coal Beach.

Yesterday, I had a very jolly day on the island of Flat Holm, in the company of geologists Sid Howells and Chris Lee, hunting for erratics.  Grateful thanks to Linda Burnell and her colleagues from the Flat Holm Society for facilitating the visit and looking after us so well.    And thanks too to Sid and Chris, who were able to identify many things with which I was unfamiliar.  We were blessed with a calm sea for both crossings from Cardiff Bay in a RIB powered by almost 500 HP of motor power -- and the crossings took about 10 mins each way.  The return journey was in the dark -- a slightly spooky experience when travelling at 30 knots or so.  Very exciting......

On the island, we were  blessed by bright calm conditions and by a lack of sea gulls.  There is a very dense gull colony here, but luckily the breeding season is long gone, and the birds had departed.  In the early summer protective clothing is required. 

Lots of interesting things were discovered on the island.  Sid and Chris will be getting second and third opinions on many tentative sample identifications, but already we have some preliminary conclusions about where the main erratic sources may be.  More of that anon......

Sunday 28 September 2014

Rhosyfelin footpaths

Following my recent chat with Dave W on this blog, I thought I'd better check out the footpath situation -- and I have discovered that there are TWO footpaths affected by this dig site, not one.

This is the map that pops up on "Where's the Path" when one looks at this area:

What we see are two paths / public rights of way, leading from the cottage near the ford towards the S and SW.  The path is shown from the edge of the garden.  One path runs very close to the edge of the rocky spur, just where the 2014 digging has been going on, and the other runs up the middle of the little channel between the spur and the road.  The path is not shown leaving the road on the hairpin bend.  That path, if it is indeed a designated right of way, appears to have been completely obliterated by the 2011-2014 digs.

So if the path as shown is correct there has either been a deliberate blocking of the path or an attempt to prevent public access, which would either be a criminal offence, or else, with the assistance of the Highways Authority (presumably the National Park in this case) a temporary diversion of some sort.  That presumably would involve some notification to the public -- and I'm not aware that there has been any such notification.

Then things get even more messy.  If you look at the 1:25,000 map for the area it looks as if this second path leaves the road right on the hairpin bend, more or less as we see on this old OS 6" map:

That must be an old trackway, running beneath the word "felin" -- and you can also pick it out on the satellite image.  So the path running past the tip of the spur is shown by the dashed lines, but the other path isn't.

All very confusing.  But is is important, because I and lots of other people will want to know whether, if we go and wander around the dig site and take photos, we are actually on a designated public footpath.  Alternatively, are we technically trespassing?

I will try to find out from the National Park what the truth of the matter might be.........

Friday 26 September 2014

Weathered and abraded rock surfaces at Rhosyfelin

These are some of the rock surfaces currently exposed at Rhosyfelin within the area of the 2011-2014 dig.  They are all quite heavily weathered, and some are so deeply weathered that you can scrape the surface away with a sharp tool and get a soft grey powder from a deep gouge in the surface.  That speaks to me of long-continued exposure to cosmic bombardment, sunshine, wind and rain over many thousands of years.

More to the point, these surfaces (some on exposed bedrock outcrops, and some on detached slabs) are heavily abraded.  By which processes of erosion?  Well, the only ones that make sense here are glacial action and abrasion by torrents of meltwater during the deglaciation phase. This of course is supported by the abundance of large rounded boulders and stones exposed in the lowest part of the dig, under the present-day grassy valley floor.  There is no way that you could find these abraded surfaces in these positions if this was a quarry.  If the rock face was a man-made feature, you would expect all of this material to be jagged, angular and sharp-edged, like the material you find upslope on some of the more recent rock-falls:

......... or maybe the quarry owners employed apprentice quarrymen to go around polishing rock surfaces and rounding off sharp edges in case anybody might get hurt by them?

Rhyolite inclusion in a quartz block

Something for Rob and Richard.  I hadn't noticed this before, but that big detached block of quartz resting on the bedrock floor just upslope of the "proto-othostat" is very interesting indeed.  It's very brittle and sharp-edged, so it hasn't been subjected to much erosion -- and that suggests strongly to me that it is more or less in situ.  But if you click to enlarge these pics you'll see that there is a large bluish rhyolite inclusion within it.  There are traces of fine fractures within the rhyolite, but it doesn't seem to be strongly foliated.  I don't suppose there is any great significance in this, but it's a nice geological curiosity. 

Precisely what might have been the mechanism of formation?  I thought quartz was often created as a "secondary" mineral in pre-existing volcanic or sedimentary deposits?  But here it looks more as if the rhyolite is the thing that is secondary.....

Pembrokeshire's latest tourist attraction......

Just been for a little expedition to Rhosyfelin.  It looks as if the diggers, having dug, left in a bit of a hurry this year.  It's rather a pity that they left the whole place in such a mess -- with a flimsy plastic sheet held down by a few pebbles.  It will be ripped to shreds by the wind and no doubt by all those locals and tourists who will not be able to resist coming to have a look at the "quarry" over the coming months.

Part of me is appalled, and part of me delighted -- since it means that geomorphologists now have a full year in which to take a serious look at this place, and maybe to do some digging of their own......

How to manufacture a Neolithic quarry

Somebody said to me the other day, with respect to Craig Rhosyfelin: "Well, there's that lovely smooth rock face.  It looks like a quarry, doesn't it?  So probably the archaeologists are quite correct......"

That got me thinking about the "archaeological artifice."  We have encountered it already, with respect to many of the small details in the archaeological dig.  If you dig under the big "proto-orthostat" and take away all the fine sediments but leave the bigger stones behind, you can then say to yourself (and to the world): "Hey!  Look! The quarrymen here have put props and pillars under this stone in order to hold it up or to help move it......."  Totally illogical, of course, but when you have a ruling hypothesis to guide you, every little scrap of evidence comes in handy in helping you to convince yourself that you are in possession of the truth.

The same thing applies to the "quarry" rock face at Rhosyfelin.  When the excavations started in 2011 there was no rock face here.  There was an overgrown jumble of fallen rocks with a few jagged outcrops projecting through the trees, gorse and bracken.  Gradually, over four years of digging, all of that material has been taken away.  The trees and shrubs have gone, and much of the shattered rock debris has been taken away with the aid of a digger and by many pairs of willing hands.  In a few places the debris has been left behind:

But where the clearing work has gone on to an advanced stage, we can pick out the old ground level and the rock exposures that were covered in vegetation -- sometimes halfway up the face.  Consciously or unconsciously, the face has been "tidied up" in order to confirm the idea that there was a genuine worked quarry face here.  This is a 2012 photo, and below it is one from 2014:

The technique, in the creation of this artifice,  seems to have been to take away the finer sediments and the rock rubble and to leave many of the larger stones behind, with their positions mapped meticulously.  The archaeologists will of course say, if asked, that they are simply revealing a pre-existing rock face and bringing it back to its pristine condition -- as it was in the Neolithic.  That's what they do, and that's what they undoubtedly believe. 

But from where I stand, they are revealing a rather fine fracture plane in a complex rhyolite crag which has been subjected at various times to glacial, fluvio-glacial and periglacial processes.  The rock face that we see today is no different in essence from thousands of other joint- or fracture-controlled rock faces all over the British Isles which are partly covered in rockfall debris.

So the rock face at Rhosyfelin is an entirely natural feature, which has been subjected to careful cosmetic surgery because the archaeologists didn't like the look of it as it was when they found it.

Thursday 25 September 2014

The Gwaun-Jordanston Meltwater Channel System

The Quaternary Research Association is seeking nominations for the 50 most important Quaternary sites in Britain.  This is my first nomination.  Citation as follows:

Gwaun-Jordanston Meltwater Channel System

Site focal point:  Cwm Gwaun, Pembrokeshire (National Grid Reference SM992353)

Photos:  landscape image -- photographer unknown.   Computer-generated image:  courtesy Jonathan Lee.  Map:  Brian John


The meltwater channel complex inland of Fishguard is one of the most spectacular in the British Isles.  It is generally referred to as the "Gwaun-Jordanston subglacial meltwater channel system" because it extends over a distance of 14 km E-W and 8 km N-S.  The main channel is over 12 km long, and cuts a curving swathe through the northern foothills of Preseli, in effect isolating the upland of Carningli and Dinas Mountain from the rest of Preseli.  The channel is steep-sided and flat-floored.  For more than half its length it has no feeder or distributary channels, but at Llanychaer the system becomes highly complex, with an anastomosing pattern evident.  Meltwater has spilled westwards through the Criney, Esgyrn, Llanwern and Nant-y-Bugail channels, with further bifurcations near Scleddau. There is one big right-hand bank feeder channel at Cwm-onen, which has carried meltwater south-westwards from the north-west face of Mynydd Dinas.  In addition to these major features there are abundant smaller "hanging" channels, rock knobs and steep chutes, some of which still carry small streams falling through cataracts.  Smaller shallow "overspill"channels also occur on many interfluves and spurs.  All of the major channels have humped long profiles.  There are abundant glacial, fluvio-glacial and periglacial deposits within the channel walls, in places masking channel floors to a depth of tens of metres.  Holocene peats and other organic-rich sediments are also widespread, particularly in the Esgyrn and Criney Channels. 

It was proposed by Charlesworth in 1929 and 1963 that the channels were all formed by meltwater spilling westwards from a series of pro-glacial lakes impounded against the face of Mynydd Presely by Devensian ice which left behind abundant traces in the "South Wales End Moraine."   Glacial Lake Teifi was proposed as the largest of these lakes, but Charlesworth also referred to Lake Nevern, Lake Gwaun and Lake Manorowen.  The channels were therefore designated as "overflow channels".  However, as pointed out by Bowen and Gregory (1965), John (1970) and many other researchers subsequently, the geomorphology and sedimentological features of these channels point clearly to formation beneath wasting ice by meltwater flowing under hydrostatic pressure.

It is assumed by some researchers (eg. Campbell and Bowen, 1989) that the channel system is of Devensian age.  On the other hand John (1970) argued that the channel system is so complex, with so many "hanging" channels and other disconnected features, that it must be of composite origin and age.  His current favoured hypothesis is that the channel system dates back at least as far as the Anglian glaciation, and that it has been re-used by subglacial and surface meltwaters several times since then.  Since Glacial Lake Teifi is now thought to have been formed during a phase of ice advance across the north Pembrokeshire coast (rather than during a phase of ice wastage, as assumed by Charlesworth) it is difficult to imagine that the lower part of the Gwaun Channel can have been used by escaping meltwater, let alone being created at this time, in view of the fact that such a scenario would have involved meltwater flowing directly against the prevailing regional ice surface gradient.

It is now proposed that the channel complex was initially cut by subglacial meltwater flowing under the control of an ice surface gradient sloping from NE towards SW.  That means that ice from the Welsh Ice Cap must have been dominant at the time, and that the Irish Sea Glacier was either absent from Cardigan Bay or incapable of resisting the ice flowing from the Welsh uplands. Virtually nothing is known of such a glacial episode.   Most of the other evidence from West Wales indicates that the dominant ice flow across this area was from NW towards SE, during both the Anglian and Devensian glacial episodes.  Much work still needs to be done before the history of Pleistocene landscape change in this region is fully understood.  But the channel complex is both spectacular and important for the elucidation of Pleistocene history in Wales.

Interesting encounter at Rhosyfelin

Update from Rhosyfelin:  The area in front of the tent has been partly filled, with some soil and many boulders thrown in on top of a protective sheet.  The steep exposures to the left have been degraded and converted into more gentle slopes -- no doubt to prevent accidents.  But some parts of the exposures are still visible.   A lot of grass seed has also been scattered. (The photo shows the dig as it looked while work was still going on.)

I called in at Rhosyfelin today, to check it out.

The good news is that the pit has NOT been filled in -- the lower part of the dig (where the work was going on this year) has been covered with a sheet, and lots of big boulders have been thrown in on top of it to keep it in place.  But quite a few of the new exposures are still visible, and can easily be freshened up if any geologists or geomorphologists want to look at them.

The bad news is that I met a fellow there with two or three others who were looking around.  His wife told me that he was Head of the Geology Department at a well-known university which will remain nameless;  I think I was supposed to be impressed.  He was holding forth about the characteristics of the "quarry" -- so I asked, ever so politely, as one does, what it was that made him think it was a quarry.  "Of course this is a quarry!" he said with disdain.  "This is very quarryable material, so it's a quarry!"  I tried, ever so gently to talk a little about glaciers and meltwater, but he was entirely disinterested, and seemed not to know what I was talking about.  And so we went our separate ways......... 

Tuesday 23 September 2014

Fossil ice wedges

Ice wedges are formed in continuous or discontinuous (or sporadic) permafrost areas when the frozen ground contracta and when a wedge-shaped mass of ice then develops to fill the void.  Sometimes the wedge grows laterally and widens, pusing the flanking sediments aside and disrupting the pre-existing stratigraphy.  Later on, when a thaw sets in, the wedge of ice melts out, and sediments slip into the void and eventually fill it up.  These are called fossil ice wedges or "ice wedge casts" -- and they are a very good indicator -- in fluvio-glacial gravels in particular -- of very cold or tundra conditions which set in after the gravels are laid down. I have seen quite a few of these features in Pembrokeshire.

The top illustration is from the Mullock Bridge gravel pit near Dale, where the wedge formed in a Devensian kame terrace.  The lower example (from the old Mathry Road gravel pit near Letterston, Pembs) is not so well formed, and is best referred to as a "fossil frost fissure" -- it probably did not survive for more than a few years or decades.

The big question about these fossil permafrost features is this:  what age are they?   The gravels were probably laid dowm during ice wastage shortly after 20,000 years ago.  So the permafrost features might have formed shortly after that.........  On the other hand, there was a distinct cold snap referred to as the Younger Dryas, about 12,000 years ago.  That did not last for much more than a thousand years -- but it was certainly cold enough for permafrost to have formed in Pembrokeshire -- and indeed there is other evidence to support that contention.

Monday 22 September 2014

Rhosyfelin and Aber-mawr

These pictures explain why I am distinctly underwhelmed by Rhosyfelin -- although I agree it is a very beautiful and photogenic spot for a picnic.........

The top pic is from Rhosyfelin and the lower one from Aber-mawr, on the north Pembrokeshire coast.  At both localities there are low cliffs made of heavily fractured bedrock, and rockfalls and the incorporation of slabs and broken debris into adjacent sediments have been going on intermittently for at least 100,000 years.  Both sites have also been affected by the incursion of Irish Sea Ice during the Devensian glaciation -- and at both sites many of the slabs are incorporated into till and fluvio-glacial deposits.

Who needs quarrymen anyway?  Life is much simpler without them......

Sunday 21 September 2014

Drover's Roads and summer farms

There are some amazing new images on North Pembrokeshire on the web -- from satellite imagery obtained earlier this year -- presumably in the early spring.  These have nothing whatsoever to do with Stonehenge or the Ice Age, but I thought I'd share them anyway!  With the sun very low, and with long shadows, lots of landscape details stand out clearly.

The upper image shows the old Drover's Road (used in particular in the late 1700's and early 1800's) running between Carn Goedog and Carn Breseb, en route towards Carn Alw and Eglwyswrw.  Many thousands of animals must have been driven this way, heading from the rich farmlands of Pembrokeshire towards the growing industrial towns of the Midlands.  The trackways are quite spectacular.

The lower image is a classic one, showing Hafod Tydfil, on the northern flank of Preseli.  Any farm in Wales which has "Hafod" in the name must have started as a summer settlement, used for grazing in the summer months when the land close to the home farm (or "hendre") was needed for the growing of crops. The Norwegian saeter was very similar, as were the transhumance cottages in the Alps.   Hafod Tydfil must have started as a very small hovel, but eventually a farmhouse was built here, and it was farmed as a viable unit for many years.  Ronald Lockley lived and farmed here for a while after WW2.  Now the farmhouse is ruinous, and the fields are just used as permanent grazing -- but the place still stands out as an "island" of cultivation in the midst of the wilderness.  I was caught in a thunder storm here once, and was stupid enough to take shelter near that big clump of trees.......

Saturday 20 September 2014

Stonehenge -- did the rhyolite fragments come from bluestone monoliths?

(Thanks to Rob for drawing our attention to this article from 2013.  It reports that Ixer and Bevins have now been involved in a "lumping" exercise by combining the three rhyolite groups A-C into one single category which has quite a lot of variation within it.  Makes a change from the "splitting" which geologists usually do.  Doing a lumping exercise has the merit of simplicity, since we are all confused enough as it is, but there is also a danger of creating complacency and causing non-geologists to assume uniformity where there is none.  I am even more confused because the authors now seem to assume that all three types of foliated rhyolite from Rhosyfelin were to be found on a single monolith at Stonehenge which has now been destroyed.........  we will no doubt have that explained in due course.  See also this post:
But what is especially noteworthy in this article is the contribution from Mike Pitts, who makes exactly the same point as I am making below.  A link to the paper is added at the end of the original post.)

The rhyolite outcrops at Rhosyfelin. A good place for making Mesolithic microliths and Neolithic axes?  That's a question worth asking.......

 Weights and specimen numbers from the Stonehenge Layer.  These graphs have no statistical significance, but they are interesting.  These are just some of the total number of fragments examined from all contexts

Let's agree with Rob Ixer and Richard Bevins that some of the rhyolite debitage in the Stonehenge landscape has come from the Pont Saeson area near Brynberian --  and more particularly from the Rhosyfelin locality.  Of the 4000 or so "bluestone" fragments examined, over 1200 (approaching 30% by number) of those have been made of foliated rhyolite similar to those identified around Rhosyfelin.  That's  about 19% of the total, by weight -- and surprisingly, that figure is greater than the 16% for spotted dolerite. Bear in mind that there are far more spotted dolerite orthostats than there are rhyolite ones -- so as Rob and Richard have pointed out, that means that any rhyolites in the Stonehenge area have been bashed about in preference to bashing about the spotted dolerites.  Maybe that's simply because it was easier to do, since flaky and fractured rhyolites are easily broken up, and dolerites are far harder.......   Interestingly, there is no segment on the diagram above for unspotted dolerite.

Given that there is huge sampling bias here (archaeologists have only examined a few parts of the Stonehenge debitage / soil layer and have automatically homed in on "interesting" samples while rejecting "boring" ones) we can say certain things about the material that HAS been collected and examined.

1.  We are talking about very small weights here.  What is the total weight of all the rhyolite fragments examined?  Perhaps Rob will tell us, but it may be no more than a few kilos.  The weight of a few small stones or axes?  I take the point that Rhosyfelin was not an established source of axes, and that there is no known "axe group" identified by the Implement Petrology Group with the right petrological characteristics, but that does not mean that Rhosyfelin was not used by some individuals as a source of material for the making of axes.  If that happened, those axes (which may not have been very good ones) could also have been traded, brought to Stonehenge, and then destroyed because they were unfit for purpose.  (Forgive me.  I feel a story coming on.......)

2.  I'm not sure why it should be assumed that because there are "Rhosyfelin rhyolites" in the debitage at Stonehenge, they must have come from one or more destroyed orthostats.  That's a reasonable hypothesis, given those enigmatic stumps that look as if they are made of the right material (32d and 32e, if we have the numbering right) -- but it's also a perfectly reasonable hypothesis to suggest that the fragments might have come from "inconvenient" pebbles, chunks or boulders that happened to be lying around in the Stonehenge landscape and which were simply smashed up because they were in the way or were of no use.

3. There are three different rhyolite fabrics in the frame here.  These are, as defined by Ixer and Bevins in 2011:

A. Dark/black, sharp, flinty rhyolite ± joint planes. Rare, pale-coloured, flinty rhyolite is probably weathered dark rhyolite.
B. Rhyolite with a planar fabric. Rare, extreme examples of this group initially were classed as ‘slate/phyllite’.
C. Rhyolite with a pronounced planar and lensoidal fabric ± joint planes.

Forgetting about Groups D and E for the moment, they say "the other three groups are significant in percentage terms both by weight and number."

I accept that there is considerable variation in the characteristics of the rhyolites across the rock face exposed at Rhosyfelin.  But do all three types occur on this single rock face?  From my careful reading of the paper, I am not sure that this has been established.  So even if some very accurate provenancing can be demonstrated (ie to within a few metres) the possibility remains that substantial amounts of the Stonehenge rhyolitic debitage have come from other sampled or unsampled locations across the Pont Saeson area either from currently exposed rock outcrops or from others that are currently invisible.  (Even if you look at the published thin sections showing the "Jovian fabric" about which all the fuss has been made, the samples from Stonehenge and Locality 8 are similar but NOT identical, as the authors recognize.)

Top image -- Jovian fabric from Locality 8 at Rhosyfelin.  Bottom image -- Jovian fabric from one of the Stonehenge samples. The fabrics are similar but not identical, as Ixer and Bevins point out.

My point is that no matter how good the geological provenancing of these rhyolite fragments may be, the orthostat obsession has led to the quarry obsession, and that in turn has led to all sorts of skulduggery in the interpretation of perfectly natural features observed during the 2011-2014 digs.........  Ah, geology as the root of all evil!

Quotes from the Ixer / Bevins papers called "Chips off the Old Block" and "CRAIG RHOS-Y-FELIN, PONT SAESON IS THE DOMINANT SOURCE OF THE STONEHENGE RHYOLITIC ‘DEBITAGE":

** Since 2008 there has been a systematic lithological investigation of debitage found within the Stonehenge Landscape; most of these lithic fragments were re-examined in April 2013. Just over 4000 bluestone debitage fragments weighing between 0.1 and 8500g have been lithologically classified macroscopically, the majority being from the April 2008 Darvill and Wainwright excavation within the Stonehenge stone circle (3657 fragments) but also including lithics from the Heelstone Ditch excavations (171) (Pitts, 1982; Ixer and Bevins, 2013), the Stonehenge Avenue including Trenches 44 (20) and 45 (71), Aubrey Hole 7 (54), and from surface finds and test pits in the area close to the western end of the Stonehenge Greater Cursus (31) (Ixer and Bevins, 2010). ............  In addition small numbers of bluestone fragments have been identified from locations in the vicinity of the Stonehenge Landscape although no bluestones have been identified from West Amesbury Henge. In addition over 100 thin sections, polished blocks and polished thin sections have been described for the major non-dolerite classes of debitage and these have been compared with polished thin sections from all the sampled bluestone orthostats.

**  Although there are subtle but distinct differences between different rhyolitic outcrops at Pont Saeson, including those on Craig Rhos-y-felin, they share a distinctive petrography that is unrecognised from elsewhere in the Fishguard Volcanic Group. Texturally all are foliated usually with an associated lensoidal fabric where deformed lithic clasts occur and carry a similar dominant mineralogy, although, in addition, some outcrops have rare, unusual minerals.

**  Almost all (>99.9%) of the Stonehenge rhyolitic ‘debitage’ can be petrographically matched to rhyolitic rocks found within a few hundred square metres at Pont Saeson and especially to those found at Craig Rhos-y-felin. However, it is possible in a few cases, where the petrography of these Welsh in situ rocks is so distinctive, to suggest an even finer provenance to within square metres, namely to individual outcrops.

**  Well over 1200 rhyolitic lithics belonging to Groups A – C have been identified and their distribution is widespread throughout the Stonehenge Landscape including close to the Stonehenge Greater Cursus area but no parent orthostat has been securely identified although hitherto unsampled buried stones SH32d or SH32e are strong candidates. The majority of the rhyolite lithics are small struck flakes but at least one large rough out weighing 190g (April 2008 excavation context 12/10; Roman) has been recognised.  It is interesting that it does not belong to any of the established axe groups established by the Implement Petrology Group (Clough and Cummins, 1979, 1988). 



A re-examination of rhyolitic bluestone ‘debitage’ from the Heelstone and other areas within the Stonehenge Landscape
by Rob A. Ixer and Richard E. Bevins, with a contribution from Mike Pitts
Wilts Arch & Nat Hist Mag 106 (2013), pp 1-15.


Recently it has been proposed that the Stonehenge rhyolitic debitage can be distributed into five petrographical groups (A-E) (and that at least three of them (A-C) are from rocks cropping out at Craig Rhos-y-felin). This supersedes an earlier classification scheme of this important category of Stonehenge material. The earlier 1980s scheme, based on lithics found close to the Heelstone, divided the rhyolites into two groups (A and B) and sub-divided the larger into two further
sub-groups (Bi and Bii). Re-examination of this earlier material together with other Stonehenge rhyolites has allowed the two schemes to be compared and integrated.
The original 1980s Group A lithics are identical to the present Group B, (both are small groups). This group is described in detail so completing the petrographical descriptions of the Stonehenge rhyolitic debitage. Despite bearing feldspar megacrysts this group shares sufficient petrographical characteristics with rocks from Craig Rhos-y-felin to support the view that that location is the geographical origin of the group.
Lithics belonging to the 1980’s groups Bi and Bii, however, are randomly distributed amongst the present A and C groups and there are no strict correspondences. The designation Bi and Bii should therefore be abandoned.
Using the new scheme it should now be possible to map more precisely the distribution of the rhyolitic debitage in the Stonehenge landscape to inform such questions as to the number of rhyolite orthostats originally present and their fate.

Extract from the contribution by Mike Pitts

Such issues of context are extremely important if
we are to understand the significance of the different
types of stone used at Stonehenge. Context has been
conspicuously absent from most debate, which has
focussed on whether the smaller megaliths were
brought to Stonehenge from Wales, or found lying
on Salisbury Plain. Debris at Stonehenge or nearby
is typically treated as if it were a proxy for megaliths.

Distinctions between different types of bluestone are
assumed to matter only to modern scientists, not
to the people who built or used Stonehenge. None
of this is helpful (archaeologists and geologists are
equally at fault).

It is notable that all the samples matched in this
study to Craig Rhos-y-felin come from debitage and
not from megaliths (although Ixer and Bevins (2011a
and b) have suggested that buried megalith SH32e
may also come from Craig Rhos-y-felin). One of the
distinctive features of the rhyolitic rocks is that they
are flinty – they have a good conchoidal fracture.
That makes them relatively easy to break up, if they
are standing as monoliths at Stonehenge. But it also
makes them suitable for making portable artefacts.
There are flaked bluestone ‘tools’ from Stonehenge
(including some from the stone floor). Which of
these are made from debris created when stones
were dressed on site? Which are made from broken
up megaliths? And which were made in Wales and
brought to Stonehenge by people visiting, perhaps
on a pilgrimage of some kind? Clearly the distinction
has important implications for how we understand

These are questions that future research can
approach through excavation in Wales and at
Stonehenge and study of the debris. Very limited
excavation at Stonehenge would allow modern
identification of the stumps and other pieces of
megalith at the site, and perhaps the matching of
some to Ixer and Bevins’ rhyolitic groups. The
dressed dolerite fragment and apparent tools from
the 1979–80 excavations suggest that re-excavation of
Hawley’s stone dumps would be productive, despite
the lack of context. The search for sarsen quarries,
too, is critical.

A good start, as Ixer and Bevins suggest, would
be to re-examine all previously excavated rhyolitic
debitage. As they say this needs to reach beyond
Stonehenge. Over 30 years ago I compiled a list
of Stonehenge-type rocks found away from the
monument, noting that there had been ‘surprisingly
little consideration of the stones imported to the site
for use as megaliths’ (Pitts 1982, 124–126). Perhaps
now this might change.

The point about Rhosyfelin rhyolites being used for tool making is a good one, and needs to be answered.  Mike also wonders why we should always assume that the debitage at Stonehenge has come from monoliths rather than from other stone debris or from broken up tools which were imported onto the site by traders or visitors.  In my view this is much more likely than the "smashed monolith" hypothesis.  After all, why would the Stonehenge builders go to the trouble of carrying a rhyolite monolith (or several of them) all the way to Stonehenge in the knowledge that it was no use at all as a standing stone, and then to smash it up on site?  It would have been far simpler, if they were more interested in rhyolite tools from Rhosyfelin, to do the work at Rhosyfelin itself, leaving all the debris behind.  Back to my point -- was it the tools that were smashed up, rather than the monoliths?

Friday 19 September 2014

The buried stones at Durrington Walls

The map above is from Tim's blog -- the purple line shows where the stones are located. 

Here is an extract from his site, for those who are interested in the throwaway line in the recent BBC Stonehenge programme:

This relates to part of the southern bank at Durrington Walls.  It appears that there have been no excavations so far in order to check out whether these really are recumbent buried stones, or something else entirely........  no doubt Tim will report on any developments.

Using powerful ground-penetrating radar, which can ‘x-ray’ archaeological sites to a depth of up to four metres, investigators from Birmingham and Bradford universities and from the Ludwig Boltzmann Institute in Vienna have discovered a 330 metre long line of more than 50 massive stones, buried under part of the southern bank of Durrington Walls.
“Up till now, we had absolutely no idea that the stones were there,” said the co-director of the investigation Professor Vince Gaffney of Birmingham University.
The geophysical evidence suggests that each buried stone is roughly three metres long and 1.5 metres wide and is positioned horizontally, not vertically, in its earthen matrix.
However, it’s conceivable that they originally stood vertically in the ground like other standing stones in Britain. It is thought that they were probably brought to the site shortly before 2500BC.
They seem to have formed the southern arm of a c-shaped ritual ‘enclosure’, the rest of which was made up of an artificially scarped natural elevation in the ground.
The c-shaped enclosure – more than 330 metres wide and over 400 metres long – faced directly towards the River Avon. The monument was later converted from a c-shaped to a roughly circular enclosure. (source)

Parchmarks at Stonehenge in 2013

There's a mention of the parchmarks paper on Tim's web site, here:

Antiquity Issue 341 - September 2014

Simon Banton, Mark Bowden, Tim Daw, Damian Grady and Sharon Soutar

Parchmarks at Stonehenge, July 2013.... Volume: 88 Number: 341 Page: 733–739

Despite being one of the most intensively explored prehistoric monuments in western Europe, Stonehenge continues to hold surprises. The principal elements of the complex are well known: the outer bank and ditch, the sarsen circle capped by lintels, the smaller bluestone settings and the massive central trilithons. They represent the final phase of Stonehenge, the end product of a complicated sequence that is steadily being refined (most recently in Darvill et al. ‘Stonehenge remodelled’, Antiquity 86 (2012): 1021–40). Yet Stonehenge in its present form is incomplete—some of the expected stones are missing—and it has sometimes been suggested that it was never complete; that the sarsen circle, for example, was only ever finished on the north-eastern side, facing the main approach along the Avenue. A chance appearance of parchmarks, however, provides more evidence.

The Altar Stone-- recumbent or erect?

I just noticed in this image of a "Pristine Stonehenge" from the recent BBC programme that the Altar Stone is shown here (in the shadow) as recumbent -- ready to be used for ritual slaughters and other nasty going-on..  I know Aubrey Burl always argued that it was meant to be recumbent and had never been set upright in a socket -- but others (like Tony Johnson) have insisted that it was simply the biggest of the bluestones, knocked over during a nasty sarsen accident.

I haven't kept up on this issue.  What is the latest thinking?

Operation Stonehenge - Part 2

"Operation Stonehenge -- what lies beneath?"  Part 2. BBC2 18th September.

Once again, a strange mixture of fantasy and fact, with assorted gruesome reconstructions of the brutal lives led by our highly sophisticated ancestors...... and the shaman looked like the sort of fellow one wouldn't have liked to get on the wrong side of.....

But some interesting information.  A few thoughts:

1.  It was simply assumed that the sarsens all came from the Marlborough Downs -- no attempt to support that hypothesis with actual evidence.  I was not convinced.

2.  It was suggested by Katy Whitaker that all the sarsens were beaten with hammer stones over every square inch of their surfaces, to give a pristine white surface which must have made the whole monument -- when complete -- a truly spectacular appearance.  Nice pictures, but I was not convinced.

3.  Wolfgang Neubauer suggested that the sarsens were dragged from the Marlborough Downs by the most direct route possible, which means along the route of the Avenue -- although it was not there at the time.  There was mention of "glacial striations"  -- by which I suppose they meant the wonderful periglacial stripes.  But it was unclear from the commentary whether they thought that the dragging of the sarsens caused the grooves and scratches shown in the surveys, or whether they simply used these existing grooves to ease the transportation by sledges etc.  All speculation, and of course meaningless if the sarsens were really collected up in the Stonehenge area.

4.  Tony Johnson demonstrated how Stonehenge was laid out with the aid or ropes and pegs.  All fair enough, but that does not demonstrate that Stonehenge ever was finished or accurately built......

5.  The crop marks -- interesting evidence which MIGHT show that there were pits at some of the locations where stones are assumed to have been positioned -- but I have problems with this (and so does Kostas) since a place where there is a hidden pit would have deeper soil and one would assume -- better moisture retention in case of drought.  On the images below the lower one is computer enhanced, and is not to be trusted.  Another thing is that even if there were sockets in the places where the crop marks are prominent, that does not tell us that there ever were any stones in them, and this is absolutely NOT "compelling evidence that Stonehenge was completed."  Even if there were big sarsens in these supposed sockets, they could have been moved about and relocated.  If they played about with the bluestone settings over many centuries, is it not possible that they did exactly  the same with the sarsens?

6.  On the bluestones, the commentary said "geological analysis proves they were quarried in Wales" -- which is of course nonsense.   There was remarkably little about the bluestones in the programme, which went on to suggest, with great conviction, that the family of six whose teeth and bones have been analysed actually came from West Wales, providing a link with the bluestones.  That is over-egging the pudding to a considerable extent -- the published evidence does very little to support that hypothesis.

7.  I felt really sorry for the poor fellow who seems to have been crushed beneath a falling monolith, but was greatly relieved to hear that he survived to tell the tale.......  and as for that nasty shaman, the less said about him, the better......

All in all, another curate's egg of a programme, which told us relatively little we didn't know already.