On September 29th 2014 I had a very jolly day on Flat Holm in the company of geologists Sid Howells and Chris Lee, and the Flat Holm tour guides, hunting for erratics and other traces of glaciation. We were blessed with perfect weather and a lack of gulls, so we were able to examine the island quite carefully. The island is important to Quaternary (Ice Age) studies because of its location at the junction between the Bristol Channel and the Severn Estuary, and because there is still great uncertainty about the location and timing of ice limits in this area. We know from studies of glacial deposits that the Irish Sea Glacier has extended at least once well into Somerset. I am a geomorphologist, and I hoped in the company of two geologists we might be able to sort a few things out, and so it proved.
First the shape of the island: it looks like a classic roche moutonnée shaped by overriding ice, with a gentle up-glacier slope on its western flank, and steeper slopes down-glacier to the east. We have to be rather careful about this, because the dipping Carboniferous Limestone strata have obviously played a role, but other evidence also supports the idea that ice has overwhelmed the island, travelling more or less from the west towards the east.
Next, the big erratics: the famous pink erratic at the north end of West Beach is spectacular and easy to spot because of its colour. We have sampled it, but we still do not know where it has come from. Could it have come from Lundy Island? That’s a possibility, but there are a number of other sources of pink granite erratic in the UK, some in Wales and some in Ireland, but the main possible source is in northern England where the Shap granite outcrops. Currently the best professional opinion is that the erratic has come from the Late Pre-Cambrian Coedana Granite outcrop on Anglesey. We cannot tell as yet whether the Flat Holm pink erratic comes from the same source as another famous pink erratic at Saunton in Devon. Chris spotted another large erratic, about 14” across on the north west foreshore. It is a brownish conglomerate, and again we have no idea where it came from. There may well be other erratic boulders with greyish colouring and fine textures, but because of the coincidence with the colouring of local bedrock boulders they are difficult to spot. A more time-consuming survey is clearly needed in the future.
There appear to be two groups of pebbles: relatively local, and far-travelled. The former include limestone, flints, dolomites, purple-brown quarzitic sandstones and shales – the latter are highly variable in lithology. Some, like vein quartz and quartzites, might have come from northern England, for example Carrock Fell. There are also many pebbles made of feldspar porphyry, feldspar-rich lava and silicified welded crystal lithic tuff. In hand samples, Sid already thinks some of these pebbles have come from north west Pembrokeshire, and it might even be possible to identify provenances on Ramsey island. Before we get too excited, thin section work has to be done. But all three of us looking at the pebbles agreed we have not yet seen anything which can be traced with confidence to the eastern part of Preseli, where spitted dolerites and rhyolitic "bluestones" demonstrate a strong link with the bluestone settings at Stonehenge. So a preliminary conclusion might be that the ice stream which crossed Flat Holm might have travelled via western rather than eastern Pembrokeshire.
When the pebble analysis is complete we will report back, but in the meantime we can say the following. The island has been glaciated at least once by ice travelling from the west rather than from the north. The ice of the Devensian Glaciation (20,000 years ago) probably did not affect the island, although the ice front of Welsh ice was not very far away. It may be that the last glaciation of the island was by the huge Irish Sea Glacier in the Anglian Glaciation, about 450,000 years ago.
I hope to return to do some more work, but in the meantime, thank you for your very kind hospitality and a grand day out.