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

Thursday, 30 June 2016

Severn estuary heavy minerals support Irish Sea Glacier thesis

With a certain amount of generous help (and not revealing my sources) I have now been able to see some scans of pages from a Bristol University doctorate thesis from 1974.  The results of the doctorate research were never published, because the author obtained a job immediately after its completion and never, thereafter, managed to find the time to write things up.  At the base of the page I reproduce an OCR document, with many corrections because parts of the scans are indecipherable.  Please forgive the inevitable errors, but this is the best I can do.......

The most interesting point to come out of all this is that the heavy minerals do not appear to have come from secondary or sedimentary rocks in the immediate vicinity of the Severn Estuary, so they must have come from primary sources either in North Wales or in the North of Scotland.  So they must have been carried in sediments by the Irish Sea Glacier as it flowed southwards initially and then broadly eastwards up the Bristol Channel and into the Severn estuary. This is of course confirms what many geologists and geomorphologists have been saying for many years.  Dr Griffiths suggests that the "immature heavy mineral suite" and the relative abundance of the heavy minerals in question, accords with Kellaway's assumptions about ice movement -- and this also supports his assertion that there was a distinct ice stream within the glacier which carried erratic material from Scotland, Northern Ireland and North Wales.

This brought to mind the work of CB Crampton on the heavy minerals in the glacial and other deposits of the Vale of Glamorgan, in which he concluded that there were various heavy mineral suites which bore little resemblance to those of the underlying rocks, and which must have been carried either by southward-flowing Welsh ice or by eastward-flowing Irish Sea ice..

Reports and transactions (Cardiff Naturalists' Society), 1900-1981 -
Vol. XCI 1961-63
CB Crampton: Certain aspects of soils developed on calcareous parent materials in south Wales

 In my mind this confirms that Irish Sea ice must not just have affected the coasts of Somerset, Devon and Cornwall, but that it must have reached Mendip and the Bristol area as well.  The heavy mineral assemblage studied by Elwyn Griffiths cannot in its entirety have been transported upstream from the Bristol Channel into the Severn Estuary by tidal streams, since the prevailing flow and direction of sediment transport in this sort of situation has to be downstream, towards the open sea.

Now I would like to know how this research ties in with the work on Flat Holm a couple of years ago.  How many of those pebbles collected from the beaches came from Scotland and North Wales?  And what does the evidence tell us about erratic comminution over distance and over time?

The evidence accumulates........


(Note:  thanks to Dr Rob Ixer for correcting various spelling mistakes in the following -- now updated.)

Extract:  Dr Elwyn Griffiths, Ph D Thesis, University of Bristol (1974): “Sedimentary Response To The Tidal regime Of The Upper Severn Estuary. "

Extract:  Dr Elwyn Griffiths, Ph D Thesis, University of Bristol (1974): “Sedimentary Response To The Tidal regime Of The Upper Severn Estuary. "

P 10
Study of the interactions ??  between wave and tidal energy indicates that wave activity, when present, only modifies the distribution of heavy mineral concentrations already produced by tidal currents.

This implies that the degree of selective sorting is mainly attributed to differences in current velocities. The highest parts of the sandbars, areas of deposition,  experience the lowest overall velocities, and channels and constrictions of the study area experience the highest overall velocities.

Finally, an analogy to the process of current selective sorting is seen on a smaller scale at Lydney sandbar. Allen (1968) illustrates heavy mineral concentrates along the crests of straight asymmetrical ripples from this site. Segregation must have taken place, during a decreasing current when the critical erosional velocity was surpassed for the light fraction but not for the heavy fraction. As the shear stress is lowest in the troughs of ripples (Raukivi,1963) quartz grains once deposited here will remain as long as the critical erosional velocity is not surpassed.


The provenance of the upper Severn Estuary sands is best determined by the categorisation of certain heavy mineral associations.  These associations, indicative of major source rocks, have been defined by Krumbein and Pettijohn (1938); Feo-Codecido (1956); and many others.

Using Feo-Codecido’s provenance scheme it is evident that the heavy mineral suite is immature and that the major source rocks are contact and dynamothermal metamorphics and basic igneous rocks. The low content of tourmaline, RUTILE and zircon (RTZ. index) suggest that reworked source rocks are minor. The perfect crystalline forms of tourmaline, garnet and hornblende, and the angular fragments of kyanite, staurolite, diopside and enstatite, rule out prolonged weathering and recycling from a secondary source. Also, the high proportions of hornblende, garnet, kyanite, staurolite, and andalusite (see Appendix 4) and the presence of glaucophane suggest that the dynamothermal metamorphic rocks are a primary and dominant source. The only metamorphic rocks within the catchment area of the Severn river are the Malverns but the linear outcrop does not contain the above assemblage.

Because of the lack of primary source rocks in the estuary drainage basin and the catchment areas of rivers flowing into the upper reaches of the Bristol Channel, reworked sedimentary rock sources must be considered. To produce an almost saturated immature heavy mineral suite plus metamorphic rock fragments the presumed sedimentary rock source must be undergoing relatively rapid erosion and have a large areal extent within the drainage basin. Also, the former source of these rocks must have been primarily metamorphic and igneous rocks.

The Old Red Sandstone and the Triassic rocks are the most common, outcropping along the shores of the estuary, sometimes forming the estuary bedrock, and within the catchment areas of the two major rivers, the Severn and the Wye.  Allen (1965) showed that the transport path and source rocks of the Upper Old Red Sandstone of the Welsh Borderland, based on feldspar associations, was from North Wales and the Mona Complex in Anglesey. It was noted, however, that undoubtedly contributing quartz, orthoclase, some of the opaques and iron oxides, the Old Red Sandstone of the Welsh Borderland is low in garnet and other high ranking metamorphic minerals.  Similarly, the Triassic rocks which possibly contribute fluorite, apatite, and sedimentary rock fragments again do not contain the relevant igneous and metamorphic heavy mineral suites. The Carboniferous Limestone, which in the Bristol district is rich in haematite and magnetite, and the Jurassic rocks of the Cotswolds, are considered to be unlikely sources.

When trying to determine source rocks which are not within the present sedimentary basin, and in this case with no direct connections, the ensuing discussion becomes rather tenuous. The nearest possible source rocks of any real extent is the Mona Complex of Anglesey and more  probably the metamorphic rocks of N. Scotland, which are 200 km and 500 km distant, respectively. The only possible method of transport from such an isolated region would be ice action.

There are known extensive and thick glacial deposits in the lower reaches of the Severn Estuary (Hawkins, pers. comm.). Kellaway (1970) proposed that the glacial erratics and the Bluestones of Stonehenge, which are believed to be derived from Pembrokeshire, were brought into the area by eastward movement of the Irish Ice. Large garnetiferous boulders have been found at Porthleven in Cornwall and at Portishead, near Bristol, which are thought to have been transported by the Irish Sea Ice from Scotland. (Hamilton, pers. comm.)

The glacial deposits ??  found in the Bristol region may, possibly, have been dragged by the ice from the floor of the Celtic Sea. It is possible that the sands of the study area have a similar mode of origin; but to confirm this, further exacting provenance studies are necessary.


Myron of Alexandra Palace said...

If there really is glaucophane that is of significance as the only real outcrop of glaucophane/crossite is the Marquise of Anglesey monument just away from the Menei Bridge. Scottish occurrences are either unconfirmed or very small. Ballantrae.
The Porthleven stone is well known to us all and is interpreted as a dropstone so of no relevance.
The raw data Appendix 4 need to be seen in order to judge the value of this. These are not easy minerals to recognise (glaucophane apart). RTZ is almost universal in sandstones and certainly in ORS sandstones and note garnet is common in the Cosheston ORS sst but not in other ORS units including of course our beloved Altar stone.
This work and interpretation are dated but were the data sound would have interesting re-interpretive value.
It would be instructive to compare this with the recent Thomas (not that one) work on the 'enigmatic' ORS and its clasts.
Provenancing sst by their HMS is very very very difficult. Provenancing clean sst is mainly done by magnetostratigraphy these days (no good for loose blocks of course). Using dilute HCl as a guiding tool is wonderfully retro but hopeless although did bring a huge smile bringing memories of flame tests.


BRIAN JOHN said...

Thanks Myron -- very useful points. (Careful -- your latest incarnation can very easily turned into Moron..!!)

Can't agree with everything, can we? That would never do. The Porthleven erratic. As I have explained before, that cannot possibly be a dropstone, no matter what assorted experts have said over the years. "Dropstone" is not the right word anyway -- a dropstone is dropped from melting floating ice and falls through the water column onto sea-floor sediments, some way below. For the Porthleven erratic to be a dropstone, we need sea level to be well ABOVE that of the present day. What we are talking about here is a stone carried in grounded floating ice onto the shoreline and then released in situ as the stranded ice continues to melt. But that's a detail relating to terminology......

If the climate was cold enough to allow sea ice to be drifting about in the Bristol Channel, carrying large stones down from Scotland or wherever, sea-level would have been about 100m lower than it is today. So no erratics could have been dumped at Porthleven, on the shore platform, close to present sea level. That is a fundamental mistake made by generations of geologists and geomorphologists. The only situation which would have made dumping from sea ice possible in this situation is isostatic depression of the coastline by approx the same amount as the eustatic sea level drop. The more I think about this, the more I am convinced that there was not enough isostatic depression in the SW Peninsula, even if Dartmoor and Exmoor supported their own little ice caps.

There is really no alternative. The Porthlevel erratic has to be a remnant of old glacial deposits which have been subsequently eroded away and incorporated into later sediments.

So the Giant's Quoit at Porthleven is not irrelevant at all -- is is highly relevant to our understanding of glacial events in the Bristol Channel region.

Moron of Iskandirpur said...

Well there is,
The Porthleven gneiss is not the only out of place large high grade metamorphic boulder along the coast of the SW English Channel (I think you meant English Channel for Bristol Channel in this context).
The seabed off the Lizard has high grade metamorphic rocks (I do not know if they be garnetiferous but certainly gneissic) and these could well be the origin for the Porthleven gneiss. Not as exciting as Highland Scottish rocks on a field trip just a few miles away. This has been noted a few times in the literature.

BRIAN JOHN said...

Sorry -- careless of me. Yes, you are right -- the Porthleven erratic is on the English Channel side of the peninsula, not the Bristol Channel side. This is a really interesting issue -- will do another post on it. Ian West has also written about the foreign stones found on the English Channel sea bed.

Dave weston said...

Dear Myron,
You say "Using dilute HCl as a guiding tool is wonderfully retro but hopeless although did bring a huge smile bringing memories of flame tests.

Similarly, using water to extinguish fires is equally retro, but it still works.

BRIAN JOHN said...

Agree, Dave. Simple methods used in the field are essential, and still need to be taught to students. I used dilute HCl every day during my D Phil years when I was in the field and wanted to find out which erratics were made if Carb Limestone. And the good old hand lens still works, as far as I am aware......

Manuel of Nagoya said...

Is 1950/60s material vintage or retro? When did you do your PH.D???

Oh course, in the field, for those little fg dark rocks a little dil HCl is in order although a scratch test would also serve but never into the lab.

Were good provenancing simple everyone would do/could it. It is not. The archy literature is overwhelmed with dreadful 'definitive' provenancing done by destructive amateurs, that work takes decades to undo. Read the tripe written about the Amesbury Archer's bracers a classical example of the 'oh both rocks are red so must be the same'. (the detailed lithology and geochemistry are fine it is the disturbingly naïve interpretation, just this side of Sailors of Stonehenge).
It has been called the granny 'Mills and Boon school of artefact provenance' and rightly so.

BRIAN JOHN said...

Ah, the good old days! I submitted my P Phil thesis in 1965. The methods I used in the field were incredibly primitive by comparison with those used by research students these days. But I still got almost everything right..... and my provenancing of erratics wasn't too bad either. You learn how to recognize rock types by looking first at the outcrops.

Moron of Monrovia said...

Is that P.Phil as in Philistine???

I remember a great Feldpar mineralogist say that his many feldspar determinations that took-up most of his Ph.D -it took him half a day for each determination, using, I think, a universal stage,(I do not know anyone who can do that now)- could all be done in an afternoon in the late 60s. Now it would take the same time but be done automatically and all the data plotted, whilst the end user is on twitter.

It is the fg Jean Muir rocks that are buggers, can be anything.

For good/excellent stone provenancing read chapters in Peacock's The Stone of Life or read between the lines of the BA bracer volume or Allen's excellent whetstone volume.
There is good provenancing out there floating above the perm(eating)snake oil.

BRIAN JOHN said...

It was something to do with philosophy, as I recall. But it was a VERY long time ago, so my memory might be at fault.....

Jon Morris said...

"Provenancing clean sst is mainly done by magnetostratigraphy these days (no good for loose blocks of course)."

Can anyone explain what Myron means by this?

BRIAN JOHN said...

Because the magnetic pole migrates around all the time, and has done thorough geological history, there is now a good record of where it was at any particular period in that history. So if you look at the geomagnetism in a particular stratum, you should be able to date it and identify its place in the stratigraphic column by recording within it magnetic north. But you can of course only do that if your bedrock is in situ -- no good at all for erratic blocks and pebbles. For those you have to fall back on other tried and tested methods of geoochemistry, petography etc.....

Jon Morris said...

How would that be of any use to provenance Brian? I understand how it might help date bedrock.. but don't get the use of the word 'provenance'.

BRIAN JOHN said...

No, it wouldn't be much use for provenancing, except as a method of supporting or confirming an hypothesis based on petrology or geochemistry. If the fabric in a thin slide of a sample from an erratic stone allowed you to tell approx which orientation it had in the parent rock (for example because of water current flow direction or wind direction in aeolian sediments) you might then be able to use geomagnetism to support your case. But a very long shot.

Jon Morris said...

Thanks Brian. That's what I thought when I read it.

Thanks for all the info on the other thread Dave. It's disappeared from the listing now, but very helpful. Much appreciated.