The sites of speleological interest in the valley fall into four main categories, which are dealt with separately, as follows:
Springs And Sinks
Apart from on the hillside itself, no surface water flows across the exposed limestone outcrops. On the east side of the study area, water sinks underground by percolation into fissures on the limestone pavements and also into the many unconformity-type caves (see section 3) and shakeholes at the Millstone Grit boundary. Some, such as Pwll Derw, have become major stream sinks. In this particular case it seems probable that a stream, which once flowed further south-westwards to sink in shakeholes on the limestone, has been captured by the formation of a shakehole in the gritstone.
The number of springs on the east side of the valley is relatively small, the majority of the percolation water, apparently, joining the flow of the Mellte on its underground course beneath the east bank of the valley. South-east of Porth-yr-Ogof, however, is a series of springs, which appear to be associated with water sinking into unconformity caves in Cwm Forth Woods. To the north of the area is the important resurgence of Ffynnon Garreg Fawr, which feeds the Nant Garreg Fawr, the only major tributary of the Mellte within the limestone outcrop. This strong perennial spring does not appear to be associated with any well defined sinks. The reason for its being perched on the valley side, well above the river, as are so many of the springs in the area, is not known; but it may be due to a shale band creating a perched water-table, or it may mark the level of a fossil valley floor which must have existed when the valley was choked with boulder clay. Ffynnon Garreg Fawr is the only spring in the area which is associated with an extensive bedding cave development.
On the west of the valley water percolates into numerous shakeholes in the Shar Wlad, an area which displays many dolines and a number of complex closed depressions, or uvalas. Water sinking in the largest of these depressions at Hole by the Wall, is reported to reappear at the large spring just upstream of Porth-yr-Ogof. Several other important sinks exist to the south-west of this area.
Elsewhere, the west side of the valley is dotted with numerous springs, ranging between river level and quite high elevations, and from insignificant seepages to strong perennial springs. Those at higher elevations appear to occur mainly at the boundary of the boulder clay, the presence of which apparently produces a perched water-table within the limestone. The lower springs rise within the boulder clay and, although no limestone can be seen, the limestone origin of the water is indicated by the deposition of tufa at the spring. An important spring at Pant-y-Llwyn is the main source of the Nant-y-Caran, which flows across the gritstone shales to join the Mellte further south.
Evidence of other surface streams having once existed within the study area is given by a number of dry ravines which intersect the sides of the valley, these occur on the east side of the valley above Porth-yr-Ogof Farm and Cwm Forth Farm, and on the west side of the valley near Maes Forth. They were probably formed by glacial melt waters at the end of the ice age, in the period when the main dry valley was being re-excavated from its glacial drift filling and before the underground drainage pattern was fully developed.
Caves Along The Afon Mellte
Because the Afon Mellte is elevated above the level of the main River Neath by the staircase of waterfalls below Porth-yr-Ogof, the thalweg of the stream is above its potential limit of downward progression. The water table therefore tends to be slightly below the river bed and the water is inclined to corrode the joints and bedding planes under the river bed so that it can flow at a lower level The largest and most obvious of these developments below the river bed is Porth-yr-Ogof where the' whole river has migrated underground in spectacular fashion leaving a dry gorge above the cave. The presence of phreatic features in the system indicates that the cave must have originated below the water table. Valley rejuvenation and subsequent lowering of the water-table later, initiated the development of a vadose system and the enlargement of the main bedding cave. The process of cavern enlargement at Porth-yr-Ogof obviously proceeded to such an extent that the cave roof collapsed causing the wholesale diversion of the river below ground.
The absence of drift deposits from the dry bed of the river above the cave indicates that the river must have been diverted sometime after the end of the ice-age, at a time when all the drift filling was removed from this part of the valley. The numerous entrances to the cave are mostly formed in enlarged joints and bedding planes where they intersect the surface, but the right hand series entrances are probably formed by aven development which has reached the surface, and are not ancient swallow holes as suggested by North in his River Scenery at the Head of the Vale of Neath. The downstream entrances are evidently formed by roof collapse on a prominent fault which traverses the system at that point. The impressive entrance gorge is probably formed by the progressive collapse of the first chamber.
It would appear that the underground course of the River Mellte is well established and that Forth-yr-Ogof merely represents a breach and extensive vadose development at the lower end of a much larger system of passages which extends well upstream. The river water first sinks into the river bed in small quantities more than one kilometre upstream of Porth-yr-Ogof at Morgan's Hole and later all of the remaining flow normally sinks at Church Sink. Although Church Sink itself consists of a number of impenetrable fissures it is likely that larger conduits channel the water flow towards Porth-yr-Ogof below the east bank of the river.
The gradient of the river between Church Sink and Porth-yr-Ogof is small and it is likely that much of the passage would be at, or below, water level. However, assuming that the process of cavern formation below the river bed has been continuing for a long time there may well be deserted phreatic passages left at higher level as the water table has been progressively lowered whilst the river has cut down its bed. In addition, water sinking underground, on the east of the valley, may have produced some higher level vadose cave development on its course to join the underground Mellte.
Caves Formed In The Post-Dinantian Unconformity
This type of cave is a feature peculiar to the Northern Outcrop of the South Wales Limestone and, particularly, of the Ystradfellte area. They have formed in the unconformity immediately below the Millstone Grit caprock, and occur along the limestone boundary where this capping is thin. Typically they are fairly small consisting, in the mature state, of a series of blind shafts on the limestone, connected by bedding development in the unconformity itself. The roof of the caves is the underside of the Millstone Grit. This is slightly permeable and the shafts are formed by water dripping from its underside and corroding the limestone into sharp fluted pits. These tend to coalesce to form the caverns and are usually blind at the bottom. In most cases water finds its way out by percolation, although some caves have been modified by flowing water. Entry to these caves is usually via a collapse doline in the Millstone Grit, and boulder ruckles are a feature of this type of cave. In general, caves of this type are fairly young features and are very unlikely to give access to major cave systems. Their development, and the part they play in the regression of the Millstone Grit boundary, are described in detail by Burke - Br.Speleo.Soc.Proc.No.s 1967.
Fossil Cave Systems
On the Shar Wlad, to the west of the river, a number of small caves exists that they bear little relation to the present topography being neither sinks nor resurgences and being situated high on the hill tops. They are, typically, phreatic in nature with some vadose development and are choked with red cave earths and dormant stalactite formations. It would appear that these caves are the truncated remnants of one or more larger systems which formed in the area before the stripping of the overlying rock during the glaciations.