I was struck by the second model on the site, in which Henry reduced temperatures by 2 degrees C in order to see what happened. Lo and behold, a single phase ice expansion was replaced by a double phase glaciation, with extensive ice cover occurring around 29,000 yrs BP and around 23,000 yrs BP. (Correction 23.2.2012; Henry points out that I have misunderstood this. Apologies -- if I had examined his graph properly I would have seen that the 2 degree reduction was applied to both models, but that on the second model he applied a further temperature reduction of 0.2 deg over an extended period prior to the glacial maximum.)
Now that's rather interesting because I have always thought that in West Wales the Early Devensian was a time of intense cold -- with prolonged periglacial activity and with long periods when the ground was affected by permafrost. That is shown in all of the coastal stratigraphy which I examined back in the good old days when I was a research student. Stratigraphy doesn't lie. The classic site is Abermawr (see below), where there is a variable periglacial deposit (4) with a weathered upper surface, underlying the Irish Sea till (5). So if the Irish Sea till was deposited during the ice advance that occurred in the Late Devensian, it seemed entirely reasonable to me (and it still does) that the periglacial materials accumulated during the Early Devensian and that they were subjected to weathering during a slightly warmer Middle Devensian "warm episode." See also:
That interpretation of events was not at all widely accepted in 1965, but it accords very well indeed with what we now know of the Late Pleistocene climatic oscillations. According to the chart below, the cold episode of the Early Devensian lasted for more than 50,000 years, starting with O18 stage 5d and ending with O18 stage 4.
This is not just an academic point. At New Quay there is a strange deposit beneath the Irish Sea till which I took in 1965 to be a soliflucted local till, laid down by an earlier advance of Welsh ice out from the mid-Wales uplands and extending westwards to a position beyond the present coastline. I don't think this Welsh ice expansion reached Cardigan.
Along the Cardigan coast between New Quay and Aberystwyth there are many places where thick "head" deposits are sometimes more than 40m thick. These have been interpreted as accumulated layers of solifluxion deposits showing that these coasts were not glaciated during the Devensian -- but that story can now be discounted, and recent research suggests that the "Blue Head" is a soliflucted or rearranged Welsh till, related to ice flowing down across the coast from a Welsh Ice Cap. The favoured interpretation seems to be that the "rearrangement" of this material occurred after the peak of the Late Devensian glaciation. I think the deposits are too thick for that explanation to be tenable -- and I suggest that the glacial deposits were emplaced in the Early Devensian (let's call this the New Quay glacial episode) and later rearranged by solifluxion processes under the coastal slope. Later on still, during the Late Devensian, these materials were all frozen solid, and the ice from the Welsh Ice Cap passed over them once and maybe twice, without having any great effect on them. (These are the episodes illustrated in Henry's models.) Then there was some further rearrangement of the surface materials in the Late-glacial, giving rise to the "Brown head" which we see close to the surface today.
For me, that explanation accords most accurately with the evidence on the ground, and it will be interesting to see whether Henry's most recent modelling experiments support or contradict my proposed sequence of events..........