The South Coast—St David’s to Carmarthen

Fig. 13. St. David’s to Carmarthen.
ST. DAVID'S PENINSULA
Fig. 14. Geologic map of St. David's Peninsula.
The north coast of St. David's Peninsula has Ordovician slate and intrusive igneous rock beneath the plateau. The south coast cliffs are of Cambrian sandstone lying on Precambrian volcanic rocks.
The
coastal path north from the car park at
Whitesand Bay and St. David's Head (173 228)

Fig. 15. Cambrian sandstone with ripple marks. North end of Whitesand Bay. View is about one metre wide.
The path crosses steeply dipping pale grey and green sandstone, the Cambrian Lingula Flags, to the first headland. The prominent sandstone layers are only a few centimetres thick. Thinner layers of mudstone separate them. Look carefully for the fossil Lingula; they are about the size and shape of fingernails, and even have a similar lustre. Although they look a bit like little clams, Lingula is actually a very primitive brachiopod.
Lingula appeared in Cambrian time, and survives today apparently unscathed after almost 600 million of years. It is the ultimate survivor. Modern lingulas live in shallow mud flats, so it seems reasonable to suppose that their Cambrian ancestors may have preferred similar habitats. The beds become thicker near the top of the flagstones. Careful study of the fossils shows the uppermost sandstones date from Ordovician time.
The Welsh Basin deepened in Ordovician time and collected mud. Rocks around the bay north of the headland show these black muds with thin bands of green volcanic ash. The beds are nearly vertical and the mudstone is now slate. A prominent stack in the northern half of the bay is made up of ash beds that resist wave attack better than the slate. North of this stack are many thin bands of ash and intrusive igneous rocks, each jutting out somewhat from the surrounding slate.
The prominent hill east of the bay is Carn Llidi. It is the highest point on the peninsula, 180 metres, capped by a thick band of gabbro similar to the one on St. David's Head. The gabbro intruded into the Ordovician mud and cooled slowly so its grains of pale greenish feldspar and black pyroxene grew to about three millimetres across. You can see them without a lens.

Fig. 16. Massive intrusive gabbro in the foreground and slate in the valley. Carn Llidi, on the skyline, is another gabbro intrusion. View from St. David's Head.
The impressive headland of St. David's Head is another large igneous intrusion, a black gabbro. The contact between the gabbro and the Ordovician mudstone lies just below the path, near the Iron Age wall. The igneous rock baked the muddy sediments to a flinty rock called hornfels, and in turn, the cold sediments chilled the hot melt. The absence of slaty cleavage in the baked rocks shows that the magma intruded the mud before the Caledonian collision. The baked mud resisted the squeezing, while unbaked mud compressed to slate. The chilling effect of the mud on the magma caused the very fine grain size of the gabbro near the contact. The big crystals in the middle of the gabbro grew far from the contact where the rock cooled and crystallized much more slowly. The gabbro on the north side of the headland is layered, conceivably because it intruded in a series of pulses.

Magnificent exposures of Cambrian and Precambrian rocks appear in the cliffs around Caerfai Bay, south of St. David's. The coastal path along the west side of the bay shows the lowest Cambrian rocks in the cliffs. Thick sandstone layers near the head of the bay contain internal crossbeds, filled channels, and a few gravel layers. These are clearly shallow water sediments laid down when the Welsh Basin was just opening and still shallow.
Fig. 17. The Cambrian rocks of Caerfai Bay.
Red mudstone lies on the sandstone, indicating somewhat deeper water. The waves carve out the soft mud to cut deep notches in the cliffs. The mud became slate as it was squeezed during the Caledonian collision. The slaty cleavage dips between 50 and 70 degrees down to the south, across the bedding. The red mudstone passes upward into purple sandstone in beds a few metres thick.
About halfway out from the head of the bay, a fault separates the purple mudstone and sandstone from banded sandstone of the younger Cambrian Solva Beds. The headland across the bay shows one of the many igneous intrusions of dark dolerite.
St. Davids (175 225) – Broad Haven and Little Haven (185 212). 15 miles (24 km)
Waves crashing along the north side of St. Bride's Bay cut cliffs into a low plateau underlain by Precambrian volcanic rocks and Cambrian sedimentary rocks. Few bedrock exposures exist on the plateau. The coastal path passes excellent cliff exposures.
Solva Harbour (180 224)
The Royal Air Force took advantage of the flat plateau northeast of Solva to build a large airfield north of Newgale. Solva Harbour (Fig. 14) is built around a valley that cuts into this plateau.
The coastal path at Solva passes an excellent section of Cambrian rocks, but the beds at Caerfai are easier to see. Sandstone and conglomerate like those at Newgale (see Fig. 18) are exposed in a small quarry just north of the village car park. The path along the north side of the harbour passes a mass of Ordovician dolerite that intruded into the Cambrian sandstone. Along the southern side of the harbour the rocks are sandstone of the Cambrian Solva beds. These sandstones are pale greenish grey, and are in layers ten to twenty centimetres thick. Turbidity currents carried these sediments into the deepening Welsh Basin. The Caledonian collision folded these beds, which are almost vertical south of the kilns.
The cliff path that follows the southern side of the harbour to the headland passes through some large sills that intrude the Cambrian sedimentary rocks. Some are dark dolerite; others are a pale igneous rock. Towards the headland, the path crosses thinly bedded sandstones separated by siltstone and mudstone. These are the Lingula Flags, like those at Abereiddy and St. David's Head. Look for the ripples on the upper surfaces of the beds, and the scour marks on their under surfaces.
Fig. 18.
Geologic map and
section across St. Bride's Bay between Newgale and Little Haven. Note
North is to the left.
Newgale (185 222)
Newgale stands on a wave-cut terrace cut across the soft Coal Measures. A fault brings the folded Cambrian rocks against the Coal Measures at the north end of the village. Cliffs north of the village, at the end of the shingle beach, are grey Cambrian sandstone and siltstone of the Solva Beds, along with a few conglomerate beds that contain pebbles of igneous rocks eroded from the Precambrian rocks. The beds show rhythmic layering. Some grade from coarse sandstone at the bottom to fine sandstone above; some have ripples on the top. These are typical turbidites, and are the oldest of the Welsh Basin.
A
small quarry just beyond the sharp bend at the north end of the village
is in
sandstone of the Coal Measures that tilt gently to the south. The
Cambrian
rocks are much harder, and dip more steeply to the south. The
Caledonian
collision folded the Cambrian rocks, which were then raised and eroded.
Carboniferous
rocks blanketed their eroded surface producing an angular unconformity.
Look for the fine exposures of glacial sediments above the beach. Brownish Irish Sea till overlies about three metres of layered gravel. Glacial melt water laid down the gravel, and then the advancing ice plastered it with till.
The view south of Newgale shows gently dipping layers of Coal Measures sandstone in the sea cliffs. The Coal Measure sandstone and shale continue in the cliffs between here and Little Haven to the south. The original sediments were deposited in freshwater deltas, floodplains, and lakes.
The A487 passes few road cuttings as it crosses the low plateau between Newgale and Haverfordwest. A low crag of flinty rhyolite of Precambrian age was used as the foundation for Roch Castle (188 221) (Fig. 18). Southeast of Roch are a few cuttings in soft Ordovician slate.
The back roads between Newgale and Broad and Little Havens cross the Carboniferous Coal Measures, though the only exposures are in the cliffs.
The Carboniferous Coal Measures in the cliffs at Nolton Haven show a fine example of the change from delta to flood plain to coal. The beds nearest the shore are sandstone beds that dip about 20 degrees to the west. Individual beds are about one metre thick, and show excellent cross-bedding where sand tumbled down the face of shifting deltas. Filled stream channels cross some layers, and ripples appear on the tops of thick layers of sandstone. These features are typical of deltas. Thin bedded sandstone lie on the deltaic sands -- these are fresh-water flood plain sediments deposited on top of the delta. Some of the sand was exposed to the air and a leached tropical soil developed on it. Such soils are commonly bleached, and have patches of ironstone in them. If the tide is out, you can see a coal bed lying on a bleached sandstone. The coal is from forest vegetation changed when buried deeply beneath later sediments.
Druidston Haven (186 217)
Druidston Haven shows Ordovician rocks that lie beneath the Carboniferous Coal Measures. The rocks in the bay are unique because they are the only well-exposed Ordovician rocks that show the effects of both the Caledonian and Hercynian collisions.
The rocks on the north cliffs are slate showing the very steep inclined cleavage typical of the Caledonian structures. The Hercynian collision folded the earlier Caledonian slaty cleavage, and rotated it to nearly horizontal on the southern side of the bay
The top of the cliff has about 15 metres of Irish Sea till beneath a layer of head.
Broad Haven (186 214)
South of Druidston you approach the Hercynian mountain front and the rocks are more strongly folded. Spectacular small folds and thrust faults in the Coal Measures are exposed along the five-kilometre stretch of Pembrokeshire Coast Path between Druidston Haven and Broad Haven.
The cliffs at the north end of the beach at Broad Haven provide a good glimpse of these structures. The rocks are all sandstone and mudstone of the Coal Measures, and all were deposited during Carboniferous time. Careful study of these rocks shows they contain several packets of different kinds of rocks laying one on top of the other in predictable order. Each packet starts with black mudstone collected on the sea floor. Then a layer of fine sandstone sits on top of the mudstone as the water shallowed. Thick layers of sandstone, with obvious cross beds, lie on the fine sandstone--a delta filled in the bay. The tops of the thick sandstone layers are commonly bleached, the result of leaching in a tropical environment, and are called podzol. Some have brown ironstone nodules in the bleached part, which is exactly the sort of thing you may see in modern soils of tropical regions. In a few cases, a coal seam lies on the thick sandstone beds; it must have been deposited as peat, probably in coastal swamps. You can see one of these sequences in the first cliff outcrops at the north end of the beach. Why are there so many repetitions of these sequences of drowning and filling? Did the land rise and fall or was it due to changes in sea level? Some geologists suggest that repeated glaciation in the southern hemisphere caused sea level to fall when ice stored water on the land, and caused it to rise when the ice melted. Other geologists suggest advancing thrust sheets repeatedly depressed the crust during the early phases of the collision of Gondwana with the Avalon/North America block.
Look for the tight folds
in the cliffs at the north end of the bay. The Hercynian crustal
movements
pushed rocks from the south, making steep north sides and gently
dipping
south sides to the folds. In many places the coal beds slip easily
because coal contains graphite, the form of carbon used as a lubricant,
and
faults
are often localized along the coal beds.
Fig. 19.
(Left photo) Faults cut some
folds where the folding could not accommodate all the pushing from the
south.
North
is to the left.
North end of Broad Haven beach.
Fig. 20.
(Right photo) Farther south from the previous photo, folding dies out
as the effects of the
Hercynian earth
movements are less pronounced. About half a kilometre north of
Broadhaven
beach.
The cliffs at the south end of Broad Haven beach show folds similar to those at the north end, though they are not as spectacular. The first headland is in a large fold in one of the thick sandstone beds. The steep north limb forms the headland and the gentle south limb continues into the next bay. A similar fold forms the next headland.
Little Haven (185 212)
You
can see a fine anticline on the east side of the small cove at Little
Haven.
The fold is in a thick sandstone, bleached pale, and full of reddish
brown
ironstone nodules. This association is typical of the tropical
weathering of
sandy delta beds built up above the flood plain. Coal lies on the
sandstone,
but the coal, being weak, sheared from the sandy substrate and was
thrust over
the anticline.
(Mileage for the Marloes and Pembroke peninsula depend on the routes taken. Allow for about 20 miles travel for each peninsula.)

Fig. 21. Geologic map of Pembroke and the Marloes Peninsula.
The Roadside Geologist will have a frustrating time driving along the narrow winding roads of Marloes Peninsula southwest of Haverfordwest. Banks built up on either side of the road prevent your seeing anything of the countryside, except tantalizing glimpses at farm entrances. The only rocks exposed are in the cliffs below the Pembrokeshire Coast Path.
Fig. 22. Marloes Peninsula shows Silurian rocks thrust onto Devonian Old Red Sandstone.
The
minor road west from Little Haven to Tallbeny
(184 212) intersects the
Pembrokeshire Coast Path, where the road comes close to the cliff.
Outcrops in
the cliff and along the path expose sandstone of the Carboniferous Coal
Measures. The headland to the west is a Precambrian intrusive igneous
rock
similar to the diorite in the Bolton Hill quarry west of Johnston. The Precambrian igneous rock
was thrust over the much
younger Carboniferous rocks during the Hercynian collision.
St. Bride's Haven (180 211)

Fig. 23. Horizontal Upper Carboniferous rocks in the cliff at St. Bride's Haven lie on folded rocks on the foreshore. The view is to the west. The cliffs are about ten metres high.
Red mudstone along the shore at St. Bride's Haven is part of the Old Red Sandstone. The many faults and small folds visible along the shore trend from east to west. A large nearly horizontal fault is exposed in the low cliffs west of the beach (Fig. 23). Away from the coast the land is flat, and the road crosses a raised platform about 60 metres above sea level, the remains of an old erosion surface.

Fig. 24. Silurian sandstones dip south in the rocks in the foreground. The 60-metre terrace continues to Skomer Island, once part of the mainland.
Martin's Haven is at the tip of the Marloes Peninsula. The rocks of the headland are Silurian volcanic and sedimentary rocks, thrust up to the surface during the Hercynian movements.
The National Trust owns
the headland, and several paths fan out from the National Trust car
park. The
one to the south goes to a deep cleft in the cliffs known as Renny
Slip. From
the cliff-top you can see the volcanic rocks of Skomer Island (173 209) to the west, and
grey Silurian sandstone to the
east, toward Gateholm Island (177 207).
Fig. 25. Silurian sandstone dips south toward Gateholm Island. South side of the headland at St. Martin's Haven.
Rocks around the headland, west of Renny Slip, are conglomerate and volcanic rocks of Silurian age. These volcanic rocks are the youngest of the Welsh Basin and represent the dying stages of the destruction of the Iapetus plate. The volcanic rocks withstand wave attack leaving minor headlands, while softer sandstone erodes more easily to become the coves. Near the lookout at Wooltack Point are lava flows of andesite and basalt containing many gas bubbles.
The beach at Marloes Sands is reached from the National Trust car park near Marloes.
Beneath the glacial till left behind by the Irish Sea ice, the rocks at Marloes Sands are evenly bedded Silurian sandstone. These sandstones lie directly above the Skomer Island volcanic rocks exposed along the path to the beach. The Hercynian collision folded the rocks and fractured them to give a weak cleavage.
Some of the sandstone beds are full of fossil worm burrows. They look like giant macaroni, about one centimeter in diameter and as much as ten centimeters long. The worms that dug them probably lived in shallow water.
Farther west along the
beach, the cliffs expose a fault that places this burrowed sandstone
next to
the grey sandstone that makes the cliffs east of Renny Slip. The
sandstones are
in layers between five and fifty centimetres thick.
Fig. 26. Devonian Old Red Sandstone conglomerate between Gateholm Island and Marloes sands.
Along the cliffs east of Gateholm Island, the grey sandstone pass up into the red and purple sandstone, mudstone, and pebbly conglomerate of the Old Red Sandstone. The most obvious rocks are coarse conglomerate full of pebbles as much as three centimeters across. All were laid down in a broad valley to which occasional storms brought sand, mud, and gravel from the surrounding hills.
The conglomerate opposite
Gateholm contains many small veins of white quartz cutting through the
rocks (Fig.
27).
The rocks were compressed during the Hercynian movements; the
compression
direction is along the veins, from south to north. The dark cracks
crossing the
veins are a very weak cleavage, where water dissolved away some of the
rock.
Note the cleavage is nearly at right angles to the veins.
Fig.27. Quartz veins in Devonian sandstone at the west end of Marloes sands.
The B4327 crosses the old
coastal terrace now 60 metres above sea level before dropping to the
shore at
Dale (181 206). The road west
of Dale follows an old glacial melt water
channel.
Follow the signs to St. Ann's coastguard station.
The coast path west of the coastguard compound passes cliffs of purple mudstone, part of the Old Red Sandstone, here thrown into large folds. A good cleavage cuts across the beds. There are impressive folds, cleaved red mudstone, and a few greenish sandstone beds along the coast path to Little Castle Point.
Fig. 28. Devonian mudstone. St. Ann's Head. The view is to the west.
The
road north and east of Dale to Milford
Haven (191 205) shows few accessible
outcrops of
interest, and the roadside geologist will do well to head directly to
the Pembroke Peninsula south of Pembroke.
RETURN TO TABLE OF CONTENTS
or go on to the Pembroke Peninsula