Distinctive altered dolerite

(and a basket of volcanic tuff) from Dyfed, southwest Wales

[471 kb]

Figure 1. A curious souvenir, clearly a paperweight, composed of an attractive ornamental stone with a medium polish, green baize beneath, and a sculpture of a stone age structure made out of a pale metal, perhaps tin. There is a story here --- but what is the rock, where is it from, and what does the little sculpture represent? I figured out it was probably from Dyfed (formerly Pembrokeshire) in southwest Wales. In that I was right, it seems. But I identified the rock as a porphyritic volcanic tuff, a surficial product of an ancient volcano. Wrong! Clue: the groundmass of the piece appears holocrystalline, not a mix of fine ash and glass.

"Rock of the Month #160, posted for October 2014" ---

A Little Mystery, and the Value of a Reference Collection

The paperweight belonged to the late Wilma Wybourn (née Roberts), a dear friend, to whom this little tale is dedicated with affection. Wilma was of Welsh descent, and had numerous memorabilia from that fair country. That, and some personal history with the Welsh countryside, proved enough to craft an explanation which seems plausible. There are three parts to this mystery.... Here is how it unfolded, followed by a learned correction received in early 2021, offering new insights into the lithology, locality, and historical context.

1. The rock itself appears (I thought) to be a coarsely feldspar-porphyritic fragmental volcanic, a striking unit of volcanic tuff. Each of the creamy-white feldspars appears to be an aggregate of several smaller crystals, so the term glomerocryst may be used (however: see under "lithology" in the concluding update).

The complex Welsh coastline includes a wealth of geological interest, around the coast from Pembroke past Haverfordwest, St. David's and Fishguard to Cardigan (Ordnance Survey, 1964, 1965). These include volcanic and volcaniclastic rocks of early Ordovician age in the Fishguard area (Bloxam, 1971). There are fossiliferous shales with graptolites, volcanic and intrusive igneous rocks, and more. Arenig volcanism is represented by the Trefgarn (Treffgarne), Sealyham and Skomer (Skomer Island) volcanic series (Neville George, 1970, pp.23-25).

2. The origin of the rock might, I thought, be the Treffgarne quarry, a source of green, purple and black tuff of Arenig (lowest Ordovician) age. Treffgarne is near the coast in what was northern Pembrokeshire and Cardigan, now combined with Carmarthen in the region of Dyfed. Samples were collected at the quarry in summer 1973, and the ornament appears similar to the green variant, albeit more spectacular due to the especially coarse phenocrysts. The Treffgarne rocks show evidence of waterlain bedding, ripple marks and mud cracks, all evidence of a subaqueous environment, whether marine or in a volcanic crater lake (Bloxam, 1971).

Back to the paperweight for a minute: the metal is slightly magnetic, but its magnetic susceptibility cannot be measured by my trusty SM-30 meter, because the metal is evidently a reasonable conductor of electricity (the test leads to a very large negative value, which is not a meaningful result!). The tuff gives a value of 0.3x10-3 SI units, which is not appreciably magnetic.

As an aside, three mines in the Dolgellau gold belt account for the bulk of historical production in the UK, roughly five tonnes of the metal, mostly mined between 1870 and 1910. The gold was in quartz veins cutting Cambrian to Silurian sediments. Curiously, elevated gold values are reported in argillites at Treffgarne (Cameron, 1992).

3. The Stone Age structure is a dolmen, the original constructed of four large pieces of rock. A famous example is located within a few km of the quarry, albeit with 3 "legs" rather than four. The Pentre-ifan Neolithic chambered barrow or portal dolmen lies on a hill north of the Preseli Mountains, overlooking the sea at Cardigan Bay (Ordnance Survey, 1965). Another local example is the Carreg Coetan portal-dolmen monument (Taylor, 2005, p.112). Morphologically similar Neolithic structures are widespread in the world, including Switzerland (Sauter, 1976); south India (Wyckoff, 1951); and the ancient metallurgical districts of southwest Spain, which includes the Iberian Pyrite Belt (Rothenberg and Blanco-Freijeiro, 1981).

The region has many scattered occurrences of volcanic rocks, such as the Asaphus Ash (as at Meidrim), named for a local trilobite, as well as shallow and deeper intrusive igneous rocks (e.g., dolerite [diabase] from Solva Harbour and the Llanwnda gabbro in the Henner Cross area. The Turnstone reference collections include examples of all these, and recognition of the dolmen enabled rapid comparison with local rocks. The paperweight rock may not be from Treffgarne, and the dolmen may or not be a plausible copy of Pentre Ifan, but it seems very likely that the original stone structure, the rock and the paperweight all come from within 20 km of Treffgarne. A nice souvenir of ancient Welsh history!

[483 kb]

Figure 2. Four samples from Treffgarne quarry, representing green, purple and black tuffs. All these rocks are fragmental, as opposed to lava flows of more coherent, crystallized and/or vitrified magma. Selected from samples 112.3-9, collected 22 August 1973.

Update: an inside view of the local rocks and their greater significance:

Some six and a half years after this story was posted, I received a fascinating correction and update to the attempted provenance of the little ornament, presented above. For this, my thanks go out to regional experts Rob Ixer and Richard Bevins, and to Brian John, who spotted my obscure post and brought it to the others' attention.

  • Location: most probably about 20 miles (32 km) ENE of Treffgarne, perhaps a locality known as Carn Goedog, on the northern flank of the Preseli Hills (Mynydd Prescelly on the old O.S. maps).
  • Lithology: though the ornament is presumably an inexpensive - though nicely crafted - tourist souvenir from maybe fifty years ago, I was not about to slice it up for microscopic analysis. Nor did I have the means to non-destructively test its bulk chemistry. However, the green spotty rock looks awfully like a spotted dolerite (diabase in N.America) from the Preseli Hills. The spots resemble feldspar phenocrysts, but apparently are fine-grained clots of secondary minerals: chlorite, epidote and albite (Ixer and Bevins, 2017). Dolerite / diabase, the shallow sub-surface equivalent of basalt lavas erupted above ground, are typically fine-grained, black and massive, often with prominent cooling joints, and may or may not have coarser phenocrysts in a fine-grained matrix. They tend to be very tough, and so can serve as excellent aggregate. This dolerite is pretty altered, judging from the colour: a local name is preselite.
  • Context: The significance of this unusual dolerite goes far beyond the craggy localities where it is found in southwest Wales! As has been known or suspected for almost a century, the Preseli Hills are the location of the bluestones, smaller igneous architectural elements of the world-famous Stonehenge Neolithic stone circle in Wiltshire, some 230 km to the ESE. See their article (Ixer and Bevins, 2017; also Ixer et al., 2020) for more details of provenance of the Stonehenge building materials (which also include a more-local silcrete) and for references to earlier studies on the origins of the remarkable monument.


Bloxam,TW (1971) Haverfordwest, Strumble Head and Abereiddy Bay. In `Geological Excursions in South Wales & The Forest of Dean' (Bassett,DA and Bassett,MG, editors), Geologists' Association, South Wales Group, Cardiff, 267pp., 199-205.

Cameron,DG (1992) United Kingdom's gold potential. Northern Miner 78 no.8, 12, 27 April.

Ixer,R and Bevins,R (2017) The bluestones of Stonehenge. Geology Today 33 no.5, 183-187, September.

Ixer,R, Bevins,R and Pirrie,D (2020) Provenancing the Stones. Mapping the Stonehenge bluestones with mineralogy. Current Archaeology 366, 35-41, September.

Neville George,T (1970) British Regional Geology: South Wales. IGS / HMSO, 3rd edition, 154pp.

Ordnance Survey (1964) Cardigan. Ordnance Survey one-inch map 139, 1:63,360 scale.

Ordnance Survey (1965) Fishguard & Pembroke. Ordnance Survey one-inch map 138/151, 1:63,360 scale.

Rothenberg,B and Blanco-Freijeiro,A (1981) Studies in Ancient Mining and Metallurgy in South-West Spain. Institute for Archaeo-Metallurgical Studies, Institute of Archaeology, University of London, 320pp.

Sauter,M-R (1976) Switzerland: From Earliest Times to the Roman Conquest. Thames and Hudson Ltd, 208pp.

Taylor,T (2005) The Time Team Guide to the Archaeological Sites of Britain & Ireland. Transworld Publishers / Random House Group Ltd, 320pp.

Wyckoff,CC (1951) Kodaikanal: 1845-1945. London Mission Press, Nagercoil, 3rd revised and enlarged edition, 78pp. plus 2 maps.

Graham Wilson, 13-14 September 2014, updates and corrections 06-08,10 May 2021.

Visit the Turnstone "Rock of the Month" Archives!

Other samples from this region may be seen here: see Rocks of the Month 53 (halleflinta from St. David's) and 111 (graptolites from Abereiddy Bay).

or browse the Rock of the Month Index, with (2021) over 800 line items.