South Pembrokeshire really is a tough nut to crack. I just wish that I had been around to keep an eye on all those elongated trenches that were dug across the area when they were installing the oil and LNG pipelines between the Milford Haven installations and the English Midlands. Much will have been revealed.......... ah well, too late now......
When I looked at all the info from West Angle Bay over the past few weeks, I was reminded forcefully that there is abundant evidence of Devensian ice pressing into the mouth of Milford Haven from the west or north-west. Along the coast to the north there are glacial and fluvioglacial deposits at Westdale Bay, Mullock Bridge, Fopston (801096) and Fold near Marloes. The Geological Survey field workers identified a sheet of "boulder clay" on the western side of the Dale Peninsula, and we are familiar with the evidence of Devensian ice affecting all of the coasts of St Bride's Bay. We have mentioned Druidston and Sleek Stone (Broad Haven) on more than one occasion. The surveyors identified strong evidence of glaciation on both Skokholm and Skomer Islands. HH Thomas thought that there was a considerable expanse of "yellow loamy drift" with many foreign erratics to the west of Roch -- but very few patches to the east of the castle. Cantrill said that the sheet of boulder clay on the coast between Nolton and Druidston extends "some little way inland." He said there are good exposures of till at Madoc's Haven and Druidston, but that the drift keeps to the top of the cliff and thins out at Druidston Villa. He said that to the east and north of the Villa there is a "small sheet of gravel" -- could this be a pro-glacial fluvioglacial deposit, laid down just beyond a static ice front?
Another spread of till with igneous erratics is described from the Talbenny area, and there is a giant erratic near the cliff edge at Mill Haven. Small patches of gravel and sand occur at Rickeston Bridge and in the Walwyn's Castle Valley, and the surveyors seem to have thought that there was some glacial river diversion associated with a feature called "The Rock" c 300 yds NE of Rickeston Bridge. Might this represent the maximum inland extent of Devensian ice in the SE corner of St Bride's Bay? This might also be supported by the presence of a "conspicuous mound" of fine yellow sand, about 200 yds in diameter, to the NW of Orlandon, about 6 km to the west; this is linked to another smaller mound, and to exposures of gravel containing much ORS material observed to be about 15 ft thick. OT Jones thought that these deposits might be linked with the kame terrace at Mullock Bridge (which I studied intensively during my doctorate fieldwork in 1962-65). There is another patch of gravel on the edge of the cliff west of Ripperston Farm, with rounded ORS pebbles "probably picked up on the sea bottom".
The surveyors noted various small patches of reddish-yellow boulder clay (with striated pebbles) and also gravelly patches around Milford, Thornton and St Botolph's; but from the published descriptions it appears that these patches are well weathered and eroded, and that they lie stratigraphically beneath head. Might they be the last remnants of a pre-Devensian cover of glacial and fluvioglacial deposits? Much fresher ice-related deposits appear around the Dale Estuary, especially on the western flank. On the eastern side of Gann Flat, almost a metre of gravelly and sandy head is seen in the cliff; this material seems to have originated in fluvioglacial sands and gravels like those at Mullock Bridge, but there has been "paraglacial rearrangement" similar to that at Westdale Bay.
My current impression is that the ice that came in from the NW across St Bride's Bay was not very thick, and that it found it very difficult to surmount the rampart of cliffs (which were of course at that time not sea cliffs, since sea-level was around 120m lower than it is now) between Newgale and the mouth of Milford Haven. In many places these old cliffs or steep coastal slopes were in excess of 60 m high, and in places up to 80 m above present OD. Beneath present sea-level the coastal slope continues to drop quite sharply down to -15m along most of the southern shore of the bay -- so incoming ice had to surmount an obstacle generally between 75 m and 100 m high between Little Haven and the western end of Skomer Island. The stretch of territory to the north of Talbenny must have been a particularly prominent obstacle. Lobes of ice may well have pushed inland both to the east of this rampart (towards Walwyn's Castle) and to the west (towards Orlandon and Mullock Bridge). My impression is now that all of the land to the west of the St Bride's - Dale Estuary through valley was inundated by Devensian Irish Sea ice.
The coast between Tenby and Pendine appears not to have been affected by Devensian ice, and DQ Bowen has presented evidence from Marros to support this contention. In my own research at Marros I did not observe any till or related deposits either.