The Rhosyfelin sediment sequence

The new article by Dyfed Elis-Gruffydd, John Downes and myself, entitled "Quaternary Events at Craig Rhosyfelin, Pembrokeshire" is now published in Quaternary Newsletter No 137 (October 2015).   Articles in this journal are not published online, and so it is reproduced -- with different format -- on the Scribd web site.  Link:
https://www.scribd.com/doc/289210389/Quaternary-Events-at-Craig-Rhosyfelin-Pembrokeshire
 A facsimile version will shortly be published on Researchgate.

Above is a stratigraphic diagram used in the new paper.  It shows the sedimentary sequence at three locations within the dig site.  As indicated in other posts, there are quite wide variations in the thicknesses of these layers, but their relative positioning is consistent.

The lowest discernible horizon (1) seems to be a layer of broken rock debris resting in hollows on an undulating rock surface, parts of which appear to be ice-smoothed.

We have shown here the till horizon (2a) as stratigraphically equivalent to the fluvioglacial horizon (2b).  This is because the archaeologists did not dig deep enough for us to claim with certainty that the till UNDERLIES the water-lain deposits.  We would fully expect this relationship to be confirmed if future digs go deeper, or if a borehole is put down as part of some future research project.

In places, near the rock face, there is rockfall material (3) on top of the till and apparently intercalated with it.  In our view, rockfalls have occurred from the overlooking crags and from the rock face intermittently ever since the onset of ice wastage at the end of the Devensian ice incursion across this site.

At the base of the post-glacial sequence there is a band of fine-grained colluvial sediment (4) with traces of involutions -- this suggests a periglacial environment.  (We do not interpret these features as loading features as found in some saturated sediments, because we cannot see how hydrological circumstances can have been suitable in this precise location.)  This band, including some clasts, was quite well exposed in the vicinity of the big stone in the first couple of years of the dig.

The stratified slope deposits (5), containing many clasts, are very prominent, and are seen in many of the photos posted on this blog.  Within this band there are four or five subtle changes in colour, texture, and stone concentrations, suggesting environmental changes. These sediments are over 2m thick in places, and they grade through to the colluvial layers seen near the end of the spur and out onto the valley floor.  We haven't had time to examine the internal variations in this colluvium, but it does contain bits of charcoal either from settlement at this site or from natural or man-made woodland fires.  There are also colour variations that seem to be associated with pedological processes.  This is a very wet and acid environment in which podsolization is normal.

Finally, close to the ground surface there is a layer of humus-rich colluvial material and modern soil (6), also with included clasts.

Note that the top of the iron-stained band does NOT mark a worked quarrying surface, "floor" or "platform."  It transgresses sedimentary boundaries and is within a metre of the ground surface in the lower part of this site.  It is in our view a pedological feature related primarily to the position of the water table.  As such, it is of no archaeological significance.

The sediments at Rhosyfelin owe nothing to human interference, and are entirely consistent with the sequences displayed at many sites around the Pembrokeshire coast.  What we see revealed here are the deposits from the deglacial / postglacial part of a glacial cycle, with sediments typical of those found during and after the wastage of a Late Devensian ice mass.  In many places (for example, Aber-mawr) there are thick head or slope deposits BENEATH the Devensian till; and indeed it is possible that such sediments do exist at Rhosyfelin and might have been revealed if the archaeologists had dug any deeper on the lower part of the site.

We have examined many features which have been described in quarrying or engineering terms by the archaeologists.  We consider all of them to be entirely natural, and our reasons will be enumerated in a forthcoming paper accepted for publication in "Archaeology in Wales."

In short, even if there are abundant radiocarbon dates from this site and even if there is evidence of a prehistoric hearth close to the rock face, we see no sign of quarrying activity for monolith removal either in the Neolithic or later in the Bronze Age or Iron Age.