Journal Series 4 Number 9

Note that this is a re-print of the original publication, based on a scanned copy. During the process of converting the original paper copy to this electronic version, the original formatting, page layout and page numbers have been lost. All diagrams and surveys have been scanned from the original and are consequently of poor quality.



Afton Red Rift Cave, Devon by EJ Watkins

Draining Sumps by Pipes and Dams by RD Craig

St. Cuthberts II, A Further Extension by RD Craig

Survey of Shatter Cave by MT Mills


Published by the Shepton Mallet Caving Club

The Mineries, Wells Road, Priddy, Wells, Somerset, BA5 3AU



It is very gratifying for a newly appointed editor to be supported 'fairly promptly' with articles for publication. Although this edition is somewhat late (for which I apologise), the tradition has been maintained; however, my hopeful intention is to see the situation reversed for the next issue. (Imagine being inundated with articles of quality two or three months prior to publication – ah 'tis but an editors dream.)

The reason for the delay has been the necessary wait for articles to be finished.

The suggestion has been put forward, that the Journal should be 'published' (two covers and an editorial) on time whether there are any articles to publish or not, and while this is a somewhat drastic step to take, it would perhaps be a timely reminder to those who suffer from that ancient industrial disease known in certain spheres as "extractum digitalis" (my apologies to all Latin scholars).

Had this action been taken, we would probably have lost the informative and interesting articles contained herein. So be it.

The extensive dissertation from Devon had me reaching for the dictionary and I have found more than one reading well worthwhile.

St.Cuthberts has yet again induced an article of interest culminating in a report of a new discovery and a high grade survey; I gather, after much mud, sweat and more mud!

Sumping, by traditional methods, will become a pleasure of the past, provided that you can carry several bags of cement and a quantity of drain piping to your destination, as well as the rest of your caving paraphernalia!

Lastly, a high grade survey of Shatter Cave being an article of quality and interest to all would be cave surveyors as well as the more experienced.


Afton Red Rift Cave, Afton

National Grid Reference SX8384 6321

Afton Red Rift cave is situated in Afton Tor Quarry, which lies in the C-shaped Afton Tor Wood, to the north of the minor road running from Red Post (on the Totnes – Newton Abbot Road, A381) to Afton. The entrance to the wood is about half a mile from Red Post; the quarry is along the right hand fork, (6" OS sheet SX 86 SW).

The cave entrance lies about 18 feet up the quarry face in the form of a narrow solutionised crack. It is situated almost in the centre of the quarry face.

The quarry was worked in the 19th century and the early part of the 20th century. The Devonian Limestone was highly prized as an ornamental building stone. A by-product of this quarry was the burning of the shattered limestone to obtain lime for the more acid lands locally and on the edge of Dartmoor. Many small quarries were used in this way and can be seen in the surrounding areas.

It is obvious that the cave entrance at one time extended further, but quarrying operations possibly destroyed the original entrance. This, however, has not been proved, as the entrance might have been buried under head deposits, as with many of the Devon caves which were discovered by quarrying operations.

History and Exploration

The entrance had been noted by AJ Fleming and W Joint in 1946, but was only investigated for the first recorded time on April 20th 1947, by W Hooper, J Hooper and W Joint of the Devon Speleological Society. They managed to 'push' into the cave for approximately 120 feet with great difficulty.

After this, little recorded work was done on the cave until 1961 when "Britain Underground", an official caving publication, backed by the British Speleological Association listed the cave as being 370 feet long. No official exploration by the Devon Speleological Society was undertaken from 1947 until 1961, when Horsley and Duckworth extended the cave from the official 370 feet to 1,075 feet in ten expeditions between the 2nd of February 1961 and the 3rd of August 1961. They broke through into the passage beyond the 'squeeze' by chipping the large protrusion which had excluded their progress. For reference purposes, they divided the extension into 'Lower Series', 'Main Series' and 'Upper Series'. They carried out a Grade 3 survey using a hand held compass, which was read to the nearest 5 degrees, and a 20 foot piece of string. Early in 1963, small discoveries were made by the Seale-Hayne Agricultural College Caving Club, namely the Lighthouse Series and the downward extension of the rift in the Flower Chamber. A small stream in Mud Hall was noted.

In March 1964 a 'dig' was started on the connection from the 'Lower Series' to the 'Upper Series' by L Maunder, JR Evan, P Mountain and EJ Watkins. In May the first verbal connection was made by Dr. O Lloyd and EJ Watkins. On the 4th of October there came the final breakthrough by Dr. D Thompson, WN Tolfree, H Steele-Perkins and EJ Watkins. The connection was named 'Watkins Squeeze'.

In 1965 a high level stream, in Mud Hall, was discovered by members of the DSS. Since this time spasmodic work has been carried out on the cave, ie., digging, blasting and some survey work.

Geology of the Cave

The cave is situated in Middle Devonian Limestone and surrounded by slates and shales of the same age. The Mid Devonian Limestone is best regarded as a thick group of slates or shales in which are set lenticular sheets of massed limestone which are very fossiliferous and are well known as building and ornamental stones. Dominantly grey in colour, the limestone is partially recrystallised, shattered and recemented. The limestone is of the massive type and has many useful index fossils from which the lithography and stratigraphy may be discerned. Small tabulate corals, Stromatoporoids and Bryozoa appear within the limestone. These fossils are in a state of decay where they are exposed to view in the cave and are difficult to extract.

Geomorphology of the Area Around Afton Red Rift Cave

The cave is situated beneath a twin topped hill at 402 feet and 409 feet respectively, both of which are relatively flat topped with a small col in between. The cave entrance is at 380 feet To the south of the cave flows the small Catcombe Brook which joins the River Hems to flow NNW to join the main River Dart, at 20 feet OD, which flows from Dartmoor.

The twin hill immediately to the north is capped by Permian Sandstone and as such represents the extremity of these beds in the area now remaining. The strata in this area is extensively faulted; the nearest observable dislocation being 300 yards east of the cave where the fault trends N 15° East and is part of an extensive system of N-S faulting in Torbay, (extending from Dartington through Afton to Marldon), associated with earth movements of the Calodonian period of disturbance. The uplift of this area and the minor faulting is of mid-Tertiary age associated with the Alpine Orogeny.

Due to an almost equally shared joint pattern, the cave trends due east, but this pattern is interrupted in two places. The higher stalagmite encrusted rift near Cascade Cavern and Mud Hall itself cuts across the jointing to lie approximately N-S and hence must be considered to be lines of weakness developed with the faulting. In view of this weakness, Mud Hall, which is the deepest and muddiest part of the cave, was probably a significant drainage point for the cave, though not necessarily the dominant one.

Five mapped springs issue within the area of the twin capped hill at:

 1) 280 feet 2) 275 feet 3) 250 feet 4) 230 feet 5) 225 feet.

which lie around the hill at regular intervals. If these indicate the base of the limestone, a junction with impervious shales, this base must lie at approximately 250 feet OD and dips ENE beneath the cave, giving a total thickness of limestone in the region of 150 feet. The lowest point in the cave is Mud Hall at 280 feet OD which is very near this junction; hence one cannot expect to extend this system much lower vertically. Cascade Cavern is in a similar situation at 285 feet OD where there is perhaps 40 feet of limestone left.

Several points in the cave approach very close to the field above and it is surprising that no surface depression exists. The Terminal Chamber of the 'Upper Series' is within 10 feet of the grass. These figures will remain approximate until the field surface is levelled along the line of the cave.

Evolutional Processes in the Formation of the Limestone and the Resulting Cave

Devonian rocks record part of a major geological phase in the evolution of NW Europe and lie within the 'type-area' of the Devonian system. It was Sedgewick and Murchison in 1839 that used the term "Devonian" in the stratigraphical column.

It has long been recognised that the Devonian rocks were deposited in a subsiding geosynclinal trough which continued southwards to Brittany and eastwards into Belgium and the Rhineland. Hendricks (1959), stressed that it was divided into secondary longitudinal basins by intervening ridges or axes. Monotonous grey muds were deposited in the basins; the ridges were the site of varied, often rapid, sedimentation with the deposit of sandy beds and limestone. The middle Devonian group of rocks, as mentioned before, consist of a thick group of slates or shales and lenticular sheets and masses of limestone, and possibly originated on discontinuous ridges or shallows within the developing geosyncline. Coarse clastic sedimentation, sand, silt, etc, did not occur to any appreciable extent though shallow water conditions are abundantly indicated in the limestone areas around Torbay. The major reef forming organisms in the Torquay area were probably Stromatoporoids and Algae; Tabulate Corals and Bryozoa occupying a secondary but not insignificant role.

Erosion Surfaces

Evidence of the former existence of marine surface may be found in flat topped hills and interfluves separating more or less angular V-shaped valleys around this coastal area.

The first recorded recognition of erosional surfaces in South West England appears to be Clement Reids description towards the end of the last century of the narrow shelf fringing the south and west coasts of Cornwall. This shelf being backed above 430 feet by a steep bluff suggesting a worn down cliff line clearly seemed to be of marine origin. Clement Reid in 1896 postulated an early Pliocene origin, but there now seems some doubt as to this dating. HB Milner (1922) suggested an Eocene origin in view of the mode of the occurrence and elevation of the deposits, whilst, more recently, SW Wooldridge (1950) has suggested that a late Pliocene or even later age is more likely.

In 1907 Jukes-Brown described the coastal plateaux in the neighbourhood of Torquay. All of these erosional surfaces were attributed to marine agencies. Green in 1936 drew attention to the difficulty of determining with any accuracy the heights of the sea level related to any particular marine erosional surface, locally complicated in the west because of the accumulation of "head" upon the platform (Head is usually regarded as rock debris which has moved downhill by solifluction in periglacial conditions although such hill creep has of course also taken place in recent times. In the field it is commonly impossible to distinguish clearly between rock waste which has moved and that which has weathered "in situ". Head in SW England is typically about 10 feet thick, but may be up to 100 feet.)

By 1938 the evidence was clearly pointing towards a widespread and substantial eustatic shift of sea level in post Alpine time. JFN Green in 1941 looked into the high level platforms of East Devon distinguishing at least six erosional surfaces in the high range between 440-920 feet. These stages were all considered to be of marine origin and Miocene to Pliocene in age. In a further paper in 1949 Green examined the River Dart area and traced at least eight levels between 430 feet and 1,150 feet. The 430 feet level was thought to have been a major standstill.

The late Tertiary landscape is thought to have been drowned to a height of 700 feet by an early Pleistocene marine transgression. The subsequent regression was punctuated by standstills at 690, 600, 550, 460, 400, 327, 150, 126, 25 and 14 feet.

At present the majority of the evidence points to a eustatic marine hypothesis with emergent surfaces being subjected to sub-aerial modification and rejuvenation affecting the rivers after each marine regression.

The decision of the International Geological Congress 1948 suggested that some, if not all, the "Pliocene" features of SW England should be regarded as early Pleistocene.

The Evolution of the Drainage Pattern

The effect of the Tertiary earth movements is not fully understood and the extent of modifications during the Pleistocene is not fully known.

The evolution of the surface which is now the more elevated in the landscape is thought to have started in early Tertiary times. It is generally thought that the cover of Cretaceous rocks extended over an extensive area, and that at the beginning of the Tertiary times, the Cretaceous cover was the land surface upon which a drainage form was instigated.

Detailed mappings of SE Devon have revealed the presence of numerous flat surfaces which record the history of the landscape below 690 feet. The evidence in the field is of four kinds:

  1. a series of flats bevelled across spurs; 

  2. well marked bluffs at the back of the spur flattenings; 

  3. bevelled hills; 

  4. terrace and floodplain fragments on the streams. 

The first three features are interpreted as evidence of a marine planation.

A problem however, remains unanswered. It is clear that there were two possible outlets for the River Dart:

  1. was eastwards towards Brixham across mid Devonian limestone; 

  2. essentially that of the present discordant course through the resistant lower Devonian rocks at Dartmouth.

How the present course was initiated is still a matter for discussion. There is no evidence for the guidance of the present steeply incised estuarine course by any line of weakness.

One possible hypothesis (Brunsdon 1963) is that the retreat of the 690 feet sea was a complete withdrawal and then the river was superimposed across the Dartmouth anticlinal ridge. A re-advance of the sea to the 430-445 feet level then took place, interrupting the drainage for a second time and sending two gulfs of this sea through the coastal ridge, one across the Brixham area and another at Dartmouth. This situation persisted during the 325, 270 and 230 feet stages but by the time of the cutting of the 180 feet platform the river was again flowing in its discordant course at Dartmouth. This hypothesis would clearly require the river to re-occupy a course which was cut after the retreat of the 690 feet sea.

Theory of Evolution of Afton Red Rift Cave System

A lengthy discussion on cave formation put forward by various schools of thought would be out of place in this dissertation. Therefore, only theories of cave formation relative to this type of cave will be considered.

Ford (1968) stated that "There are opposed general theories which place the bulk of cavern development in the vadose zone or along a water table, or at random depth beneath it" (see Warwick's review 1962). The weight of published opinion latterly favoured shallow phreatic development (Moore 1966). When studying a group of caves it is desirable to distinguish conditions of 'cave initiation' or 'origin' on the one hand and 'cave development' on the other (Warwick, 1967 page 72). The principle is that at the moment that 'cave development' commences, the ground waters have a widespread mesh of minutely open channels to choose from. This mesh constitutes the first phase of the cave history.

Many variables of the erosional environment may be quite different from those which develop the cave, eg. parts of the mesh may be diagenetic.

Some authorities hold that the initial course will be wholly waterfilled at the onset of cave development, (ie. all limestone caves have a phreatic origin). Others consider that a water table may already be established at a depth in the rock (Davies 1960 and Woodward 1961). From White and Longyear (1962, page 165) the lowest threshold value for the condition of development will be a diameter of width of some 5mm in a groundwater conduct.

In the Buckfastleigh Complex and the Afton area there would seem to be two distinct conditions of origin:

  1. the Buckfastleigh Complex appears to have been wholly phreatic in origin, and 

  2. the Afton area seems to have been shallow phreatic by mainly vadose in origin. 

The early development was created predominantly in bedding planes (the standard accepted evolutionary theory) by widespread anastomosis. When water first seeps along a bedding plane it commonly dissolves out an anastomosing pattern of small half tubes from 3mm upwards. A sequence of development has been worked out by CD Ollier and EK Tratman. These stages are shown diagrammatically (in Figure 1):

  1. at first all channels are entirely water filled and half tubes are formed above the bedding plane. The tubes are larger in the centre. At this stage there is little downward solution; 

  2. further downward solution or less water flow caused the large tubes to become partially air filled, while the smaller tubes remain water filled. In the still water filled tubes all round solution commences, while in the partially air filled ones solution only proceeds in the lower part which is still submerged producing a tapering pendant; 

  3. with further downward solution the water surface is lowered. The tapering pendants may already be above water but others will project beneath the surface where the tips are removed by solution leaving a flat or current marked (scalloped) surface. The smaller tubes at the margins become air filled relatively quickly and so retain their phreatic form; 

  4. the present day form is shown. 

More complicated forms than those shown can occur – but this example gives a model from which to work.

This course of development is unlike that postulated by Bretz. Roof form may change in a distance of only a few yards from half tubes to reduced pendants, fading out into current-produced scalloping and grading into a nearly smooth roof. Evidence produced is in definite favour of solution playing a dominant role in cave formation and corrosion playing an almost insignificant role.

Figure 1 – Stages of Anastomisal Development

Caves themselves are solutional features and their shapes reveal much about the nature of limestone solution. Most caves have a marked horizontal development, even in steeply dipping or folded limestone. This suggests that their position and solution of limestone follows the water table in Limestone Country. In plan many caves show a joint controlled pattern (as in Afton – see survey).

Originally the Devonian Limestone was covered unconformably by Permian Sandstone which was laid down some 270 million years ago in a desert-like basin and originally must have covered a considerable area to the south. A thickness of some 3,100 feet of this sediment is believed to underly the English Channel south east of Looe. The limestone may have been covered for a considerable period and possibly the Permian might have been in turn covered by succeeding deposits right up to the Cretaceous transgression. (In fact this is what did happen.)

Succeeding transgressions and regressions would have altered the original surface, both by marine and sub-aerial erosion, until the Pleistocene Glaciation when the sea level would have fallen drastically to expose an erosional surface. Mechanical and chemical agencies would have cut into this exposed surface. Between advance and retreat the sea level would have altered so that a eustatic marine situation would have occurred.

The late Tertiary landscape is thought to have been drowned to a height of 700 feet by Plio- Pleistocene marine transgression. The subsequent regression was punctuated by standstills at 690, 600, 550, 460, 400, 327, 150, 126, 25, and 14 feet. (Orme 1961). The majority of opinion suggests that the 430 feet standstill was a major occurrence with easily recognised features around the SW coast. It would appear that it was this standstill that finally eroded away the Permian cover so leaving the Devonian Limestone exposed to sub-aerial attack. One must remember that the remaining Permian cover would have been much larger in extent and the emerging erosional surface would have been theoretically level with a slight slope in the direction of withdrawal. This would leave the rivers free to flow in the main along the slight slope until lower structural features made themselves felt with the gradual process of river capture and associated hydrological processes. If the Permian cover had been eroded away during the marine transgression there would have been ample time for the development of a phreatic network of tubes and solutional passages forming along weaknesses within the limestone. With the lowering of the waters, the Phreatic system would have merged as a partial vadose system until the changeover would have been complete (thus, Phreatic to Vadose).

As with many caves it is clear that their entrances lie close to, or at, an impermeable boundary. There is little surface drainage on limestone although it is not entirely lacking. Where a stream leaves an impermeable or less porous strata than limestone, it almost immediately goes underground and blind valleys develop as surface denudation by stream action ceases beyond the sinkhole. The impermeable area acts as a collecting area for the water and the size of the catchment affects the stream size which in turn affects the cave size.

The 400 feet initiation of the development of Afton would have started during this standstill and the subsequent lowering of the waters would have played a vital part in the formation of this beautiful Swallet Cave system.

Bedding planes are most important lines of early phreatic development. Almost every known passage has evolved from initial development along a bedding plane. In the early stages it appears that a braided stream of branching and anastomosing streamlets flowed over the bedding planes and corresponding channels were incised. At this stage bedding plane passages may be present.

Ultimately the drainage becomes better integrated and one main, but meandering, channel will develop. When one well-defined channel exists from the early stages, a simple canyon or rift passage may develop without any oxbows. Downcutting is obviously the work of vadose streams, as phreatic action would produce all round solution. Afton, it seems, has had an almost entirely vadose history. The normal sequence in such cases of vadose evolution is:

  1. 75% / 25%  water flowing across limestone underground – phreatic; 

  2. 50% / 50%  gradual enlargement of joints and bedding planes – solutional; 

  3. 25% / 75%  gradual movement of water underground enlargement by streams own water until completely vadose

The vadose system acts only as channels / passages to carry water as quickly as possible to the water table.

The Afton Red Rift Cave system runs due east and it would seem probable that this was the direction of the original river system. Brunsdon (1965) postulated the probability of the Dart having at some time during the 430-455 feet standstill run due east across the middle Devonian limestone, even after the re-advance of the sea to the 430 feet level – this is supposing a complete withdrawal of the 690 feet sea, but this need not have, in fact, been necessary. If the drop in sea level from the 690 feet transgression took place as a gradual process with intermittent abrupt drops in level, the waters, as postulated before, would have flowed across the slightly sloping strata and this would fit in nicely with the theory that the Dart at a former period flowed eastwards towards Brixham. Even if this were not so it is very possible that streams did drain south-eastward across the emerging landforms

It is obvious from observation in the field that the waters that formed Afton did flow across the former Permian Capping. All the cave deposits are stained a beautiful bright red and fine clay and silt are deposited in the lower system. The speleotherms are a bright red in section.

The total quantity of sediments is relatively very small and very fine with no sands or rock fragments. Rock debris does occur in Cascade Cavern, but it appears that those fragments are from a later process and they show no signs of having been used in a corrosive process. It appears that corrosion has had little effect in the formation of this cave.

After the exposure of Afton associated with the lowering of sea level, climatic factors would have played a part in the modification of the original system. The climate of the SW changed frequently from Arctic cold glacial conditions to a climate at least as warm as today during the inter-glacial phases. This part of the country lay beyond the limits of the ice sheets, though there is clear evidence that the hills were snow capped as might be expected. The limits of the ice sheets, it is suggested, were roughly along the north coast of Devon with extensive outwash channels flowing in front of these ice sheets.

The south-west was a Periglacial area, that is, a region where the climate is one in which freeze and thaw alternate repeatedly. In such a region there are more favourable conditions for effective mass movement. Although melting of the surface may occur in the summer, the soil at depth remains frozen. Solifluction is a form of creep which refers to the creep of soil debris under frost action either in areas of permafrost or in nival areas subject to freezing and thawing. This action of frost caused a modification to the cave. In Cascade Chamber massive blocks have fallen onto the cavern floor so greatly modifying the final shape – this process of fallen blocks can be seen also in Flower Chamber. Both of these chambers are within the upper system fairly near the present day surface.

The mechanical action of ice on rock has been looked into by AK Wells and JF Kirkaldy (1937) and King and Embleton (pp. 449). It has been found that if water is entirely enclosed within rock, under a temperature of -22°C, the conversion into crystalline ice will produce a theoretical maximum force of about 2100 Kn/Cm2 – although in fact no rock would have water totally enclosed within, but the principle of ice shattering is amply illustrated by this example. The massive blocks in Afton, some in the region of 15 feet by 10 feet by 5 feet, would have been easy prey for such action. In the formation of the cave many anastomosing channels would have been interlaced throughout the various bedding planes. These channels would have weakened the roof materials, so that the action of frost would be a comparatively simple process in wedging off of these blocks. A Sutcliffe, in his research into Joint Mitnor Cave, Buckfastleigh, concluded that the frost shattering of stalagmite floors in Joint Mitnor Cave was of Post-Eemian origin. Many Devon caves exhibit frost shattering features.

It can be seen that this occurred in Afton Red Rift at probably the same time, and that after this shattering the redeposits of calcite were of a decidedly different colour. Whereas before shattering occurred the speleotherms were bright red; after, the deposits were clear to mud-brown in colour. It is probable that these clear deposits of milky calcite were deposited after the complete removal of the Permian so that the beautiful staining from water draining off, or percolating through, the Pernian cover would have ceased with its removal, and succeeding deposits would be uninfluenced by the staining.

After the retreat of the last ice sheets, the valleys would have emerged partially filled by the materials left by the action of solifluction. The sea level would have risen slightly to accommodate the melting ice, and the valleys would have been regraded to this situation.


No comment can be made on the hydrology of the cave and its region. One can only theorise on this most interesting point – since the cave system is in a state of fossilisation, that is, the active period of formation when water played a major part in the comparatively rapid evolution of this system, has ceased. The cave has been, as it were, left high and dry, and is now in a process of suspension before a gradual breakdown of its deposits.

The cave is now completely dry and one can only speculate that the water which formed the cave did, at some former period, flow across the Permian cover and sink into the limestone outcrop on its way to the water table. Within the cave all of the walls exhibit many fine fossils and colours. These fossils are in a state of decay on the outer 1/8 - 1/4 inch and will come away in ones hand to form a mudlike texture. This can be put down to atmospheric solution with the humidity playing a part in the gradual decay of the rock surface. Deposits of flowstone in the Flower Chamber exhibit this rotting process – the red stained flowstone can be clawed away to form a paste-like material. The flow looks something like 'moon-milk' and handles in the same way, but the formation is of flowstone – that is, it exhibits all the physical characteristics of flowstone, until one touches it, then it crumbles into this paste-like material. On looking into this phenomena, I have been unable to discover any references to the process. There must be a chemical explanation to this – possibly the breakdown of crystal structure through an impurity – flowstone being high on the list of impure chemical composition. Perhaps this is the first process in the cycle of cave formation?


The rate at which the process of solution is controlled is not only by equilibrium reactions, involving concentrations of CO2, pH, etc. but also by the kinetics of reactions. The faster the water speed and the motion, the faster the solution of limestone, (as with the wind erosion), so that with eddies and the pressure differences within, the solution along a water course will not be uniform. It is thought that this is the probable origin of the formation of scallops.

The process of scallop formation can be likened to the action of sea waves – shallow water moving over sands will be impeded by friction until the water moving at the surface will overtake the water beneath – this causes waves.

Formation of Scallops

The water in the turbulent and non-turbulent sections, by effecting the kinetics of reaction, will give rise to different rates of solution; and so periodic hollows will form. The turbulent water in the hollows is overshot by water with laminar flow, and the very existence of the hollows (scallops) becomes asymmetrical. A smooth surface first becomes slightly pock-marked, with quite separate small depressions like bullet marks. These then grow larger, and eventually coalesce into a common continuous mass of asymmetrical scallops. Maximum solution takes place in hollows where the water is turbulent; hollows migrate upstream. This form of solution causes more rapid formation than any other. There is little corrosion in the formation of scallops. Ollier and Tratman (1968) suggest that it is possible that this is significant in forming caves in a short time. They go on to tentatively suggest that scallops are rare at the top of canyons and common at the bottom. This is not so in Afton where it appears that scallops are of equal distribution throughout the section.

Figure 2 – Stages of Scallop Formation

The formation of scallops by cavitation and corrosion by sediments favoured by continental writers, eg. Groze and Mooreard Nicholas, is now held to be basically unsound.

The scallops in Afton are mostly in the region of 15-20 inches long and are generally deeply etched into the limestone, some of them having a depth of 6-9 inches.

Mud Deposits (Clastics)

Within the lower series, especially in Mud Hall, there is evidence of mud deposits. These fine formations fluted and drip sculptured, are susceptible to peoples hands. It would appear that Mud Hall at some time was half filled with stagnant water – this still water acted as a settling tank and a thick deposit of red mud was deposited. These fine details have now almost totally been spoiled. Along the passages which lead from Mud Hall there is a generous deposit of fine red clay.

Passage Formation

The development of the canyon passage is of interest. The passage which is high and narrow, exhibits many interesting features:

Scallop Formation

The base of a canyon is commonly semicircular in section as the water dissolves in all directions, (Ollier and Tratman 1969).

  1. with no change in water level, the canyon will be deepened while retaining a semi-circular U-section; 

  2. if the water rises the passage walls will be dissolved as well as the semi-circular base, and the passage will be widened until the water can again be discharged in a semi-circular channel; 

  3. if the water level falls, a passage of a smaller cross section will be incised within the older, wider one. 

Figure 3 – Stages of Canyon Passage Development

The profile of a canyon passage is dependant upon two factors which may act separately or jointly. One is the solubility of the rock, and the other the volume of water. In downcutting the cave will become wider if a more soluble band is reached, or, if the volume of water is increased, the cave passage will become narrower.


Within the passage of Afton there is little decoration. A few calcite veins have small stalagmites and flowstone formations, but generally the formations are mainly in Cascade Chamber and Flower Chamber. In the lower series, pure white helictites (eccentrics) occur. These are most beautiful. Within Flower Chamber there are beautiful calcite forms – called cave flowers – hence Flower Chamber. These formations are very delicate and are stained a bright red. The possible origin of these flowers is still uncertain, but it would appear that there are several factors which should be taken into consideration. The first is that these delicate formations could not have been laid down when there was an active streamway as their delicacy would inhibit any growth. One must suppose that they are of secondary nature to the original cave passage formation. Secondly that the deposits were laid down under still water as their form is at right angles to the rock surface; had there been moving water their direction would have indicated this fact. One could picture these formations growing from the process of "flow" down over the rock surface – but in all other cases this action produces flowstone. It appears from observations in the cave that these flowers have various stages of development. The form that these formations take is botryoidal – that is, the rounded growth of calcite one layer upon the other, in a spherical form. The only anomaly to this sequence is that these flowers are grown on stalk-like appendages.

Figure 4 – Stages of Speleotherm Development

How this formation develops is a partial mystery – observation of these formations in various stages of development (now fossilised) reveals that many of them are secondary growths upon former calcite crystals, and that the first step in the chain is the formation of rounded pimple-like calcite forms at the tips of the calcite crystals.

Figure 5 – Stages of Calcite Crystal Development

This would explain the stalk-like feature of such formations. Obviously more research into such formation is needed. These flower formations are the only example I have ever seen.

In Cascade Chamber there is a massive flowstone deposit entering from the ceiling and flowing down over a beehive-like rounded slope. This formation is still marginally growing although very slowly. Beautiful curtains of calcite hang from the ceiling. These growths are formed when CaCO3 laden water is induced to run / trickle down over a mass of rock, the resulting deposit gradually builds up.

These formations are usually fluted and delicate and those in Afton are stained a beautiful red.

Figure 6 – Stages of Calcite Curtain Development


It is the Plio-Pleistocene period that presents difficult questions.

There is a wealth of knowledge on this period, but, with all this, many questions still lie unanswered. Because of the nearness of this period, we are faced with the situation of being unable to 'see the wood from the trees'. Much extraneous material is available to cloud the main issue and structure of events.

It can be reasoned that Afton Red Rift Cave is an earlier feature in the landscape than the Buckfastleigh Complex and its wholly phreatic evolution, at a later date; and is probably the oldest cave system yet discovered in Devon. The cave lies at 380 feet OD at the highest altitude. This fact suggests that it was formed during and very soon after the 430 feet standstill of the Pleistocene sea, when the sea was gradually regressing. The sea has not since that period re-advanced to its former 430 feet.

The later solifluction and "head" deposits present an unanswered problem, that is, why the cave does not exhibit any major infilling which would be expected from a partial vadose system such as Afton. One can only surmise that either there was no infilling (which seems most unlikely), or that any infilling which did at some time partially fill the cave has since been carried away by later water action. What still remains unsolved is the fact that the surrounding cover is of shales and limestone, but there is no evidence of any of this broken down material having entered the cave. From this situation it points to the fact that Afton was wholly evolved during the rapid marine planation and the following sub-aerial erosion that followed, and that the Rift entrance, wherever it may have been, was blocked by infilling during the succeeding phase of time; only to be discovered by the quarrying action during the 19th Century.

Therefore, my conclusion is that Afton Red Rift is of probable Plio-Pleistocene origin.


The present survey follows the standard of accuracy of the Cave Research Group as laid down in their Transactions Vol. 8. No. 2, July, 1966, (Butcher and Railton).

The present survey was undertaken as a Grade 6 survey and was completed in six underground excursions totalling thirty hours, by myself, P Cousins (WCC), P Cornelius (CCC) and J Steere, and other occasional assistants. The instruments were borrowed from the Wessex Cave Club. The 'leap frog' technique was used throughout with mounted instruments and candle flame targets. Since the cave is an open rift and often 'bottomless', a conventional tripod could rarely be used, and the instrument mounting employed was an adjustable stemple. This is believed to be the first time that a stemple has been so extensively used in a cave survey. The instruments – an ex-Army prismatic compass coupled with a Japanese-built Abney-level – were mounted on a combined head made of brass. Headings were made to the nearest half degree with the aid of built-in illumination. A fifty foot Fibron Tape graduated in feet and twentieths of a foot was read to the nearest tenth of a foot.

Around the line framework thus surveyed, the cave outline was located by station co-ordinates – the distances right, left, above and below at each survey station. The measurements thus obtained underground were calculated by the use of log tables to five decimal places.

The cave contains a loop of approximately six hundred feet due to the connection 'Watkins Squeeze' from Mud Hall back to the entrance rift. When calculated, the survey misclosed by 3.8 feet vertically and 6.4 feet horizontally. Expressed as an absolute, ie. three dimensional closure, this is 1.2% of the traverse length. The discrepancy has been distributed in equal parts to each leg of the traverse for the published survey. In view of the use of mounted instruments throughout, and the precision with which they were read, the survey should be classified as CRG Grade 6. However, reference to Warburton’s theoretical treatise (Wessex Cave Club Journal No.8.) shows that misclosure of 1.2% over 600 feet is more likely to occur in a Grade 5 survey. Hence the grading of 5 has been adopted as being a more accurate representation of the precision of the present survey. Being the first people to make extensive use of expanding 'Stemple' bars for holding the survey head, and bearing in mind the difficult terrain, Grade 5 is felt to be an adequate survey.

Figure 7 – Survey of Afton Red Rift Cave

So far only a small amount of detail has been recorded, and to clarify this presentation, very little has been reproduced here. A number of known side passages have not yet been surveyed through lack of time, but these and any more that may be discovered will be added before any large scale final survey is printed in mass.

In the meantime myself and P Cousins will be pleased to hear from anyone making additional discoveries.

EJ Watkins

January 1970


Draining Sumps by Pipes and Dams

Sumps have always proved a problem to explorers underground. Compressed air breathing apparatus has alleviated the problem somewhat, but with an ever increasing number of sumps to pass in order to visit the working face, a great strain is put on even this sophisticated equipment. The result is that such trips are becoming major expeditions to support one man who may push a few more feet into the terminal sump. When this point is reached it may be better to try a different approach.

If sumps are considered as chokes (they are in fact choked with water), and compared with silt chokes which are usually excavated before anyone ventures through, then why not excavate the water before going through? It should be possible to redirect the water to the far side of the sump so that it runs away downhill. The manner by which this may be achieved is as follows.

Firstly, a stout concrete dam must be constructed on the upstream side of the sump. The dam should preferably be sealed to bed rock, but if the bed rock cannot be located, as may be the case, then the base of the dam must be at least 2-3 feet below the stream bed. This is best done by digging a trench and filling it with rocks and rapid hardening concrete made from 'cement fondu' or rapid hardening Portland Cement. Above the level of the streambed ordinary Portland Cement may be used. A pipe should be incorporated in the dam with the entry at least 1 foot above the streambed: fibre glass is recommended as it is light and inexpensive. The purpose of this is to raise the water by 1 foot, thus causing a small pool which will prevent stones from being washed into the pipe under these more static conditions.

The dam should be built a few feet above the pipe to allow both for bailing the sump, and water backing up under wet conditions. Enough pipes to take the water from the dam should be assembled on site (the length of the pipes will depend upon the various constrictions in the cave). Rubber seals are best and if car or lorry inner tubes will fit over the pipes, they make a very good seal. Each seal must be firmly clamped otherwise water pressure will force the seals off.

There are two ways of conveying the assembled pipes through the sump; either by baling, which is long and laborious and may require extra dams, or by sending a diver through with one end of the assembly. The pipes are connected to the pipe in the dam and the water allowed to flow through. To drain the sump and keep it permanently drained, enough garden hose should be obtained to start a syphon, putting one end in the sump and the other in the stream. Care must be taken not to fully drain the sump, otherwise the syphon will break. In other words, the end in the stream must not be lower than the end in the sump.

Figure 8 – Diagram Showing Dam and Pipe System

Having drained one sump in this manner, there is no reason why the next sump should not be tackled in the same way, and so on.

The sump in St. Cuthbert's Swallet was tackled in this manner and at present the system appears to be operating satisfactorily.

RD Craig

June 1970


St. Cuthberts II – A Further Extension

During the exploration of St. Cuthbert's II on 31st October and 1st November 1969, several members of the party noticed a wide opening on the left at stream level about 50 feet down stream from the sump. (1). A passage continued for about 10 feet before becoming choked with silt, with a couple of inches of air space visible above. Excavation of the passage was started immediately, but further progress was prevented a few weeks later, and for most of the time during the winter months, when St. Cuthbert's II was cut off by winter rains. During this period work was commenced on a project to permanently pipe the stream through the sump, making access to St. Cuthbert's II possible in most conditions. This had been done crudely soon after the discovery, but the pipes were joined with inadequate seals and were connected to a leaky mud and boulder dam.

On 11th April 1970, after many weeks of preparation, the system was ready for assembly. Fibre glass pipes five feet long and eight inch diameter, were to be joined with lorry inner tubes and connected to a 3 feet high concrete dam through which a pipe had previously been sealed. Using this dam and another, 5 feet high in Gour Hall, the stream was blocked and the sump bailed. Carefully the pipes were threaded through the sump and one end connected to the dam. Slowly water was allowed to enter the pipes and a visual inspection indicated no leaks. The system was working.

Work at first concentrated on a 'soakaway' in the sump itself, but after a few weeks the digging had disturbed the balance so that the water could not drain away, and the sump slowly filled up.

On Tuesday 5th May 1970, Roy Bennet, Tim Large, Ray Mansfield, Martin Webster, Pete Rose and Bob Craig entered St. Cuthbert's II with the intention of restarting the dry dig, 50 feet downstream from the Sump. It was hoped that if the soakaway was left alone it might start talking water again. It did not, however, and on Friday 8th May Martin Mills, Tim Large and Bob Craig arrived at the sump to find it impassable. The following day, the same party with the addition of Bob Mehew, entered the cave armed with 120 feet of garden hose. The plan was to put one end in the soakaway and one end in the stream to start a syphon that would operate continuously. When installed everything appeared to be working well with water issuing from the stream end of the syphon but it would be a few days before it would be known if it would operate continuously. A little more work was done on the dig the same day and 2 feet of progress was made.

On Tuesday 19th May, Ray Mansfield, Tim Large, Pete Rose, Roy Bennett and Bob Craig continued forcing the end of the dig which was blocked to the roof with thick mud. After an hours digging the excitement grew as air space was struck once more and a strong draught emerged. Ahead the passage could be seen to grow higher, but it was impossible to tell whether or not it was high enough to crawl along.

On Friday 22nd May, Martin Mills, Bob Craig and Bob Mehew were back at the dig. The open passage they were chasing was proving a little elusive, but they were convinced that a new discovery lay just around the next corner. The following day Tim Large joined them on a mammoth digging session. The passage was pushed a further 6-8 feet in conditions that were turning everything into a slurry, making it difficult to recognise one's neighbour. At this point they thought that they were about to enter a passage at least 8 feet high running at right angles. Their hopes were shattered then they found themselves in a 20 feet high aven with no way on. On closer inspectionhowever, a mud choked passage was seen to continue straight ahead. Excavation continued for 4 feet or so, with air space visible after burrowing upwards. At this point, with prospects looking very good, they decided to leave the dig until the next morning when a renewed attack could be made, after a much needed rest.

On the next day, Sunday, a slightly depleted team consisting of Tim Large, Bob Mehew and Bob Craig renewed the digging. Four feet of passage was quickly dug out revealing a 3 inch high air space above the mud, but to the left the roof seemed to rise indicating that they might be approaching open passage, (they were becoming a little cynical about these signs, since they had dug about 25 feet chasing 'open' passage up to that point that weekend). After digging for another couple of hours the great moment came; Tim was able to squeeze through and found himself in a sizeable chamber, perhaps 30 feet high. A few moments later they were all in the chamber examining the way on. To the left of the chamber a large inclined passage continued at right angles to the dig. After an awkward climb of 20 feet, progress could be made along the top of the incline which varied between 20-30 feet deep. Here and there, at the top of the incline where it was possible to stand, 'dead' and crumbly speleotherms could be seen. After 150 feet the incline disappeared and the passage became more boulder strewn, but 20 feet ahead a tight squeeze over a stalagmite floor could be seen. Bob Mehew and Bob Craig got through with relative ease, but Tim had to chip away a little of the floor to make progress. Just beyond the squeeze a wedged boulder was pushed aside to gain access to a very beautiful grotto decorated with a pure white column, 'streaky bacon' curtains and a grotesquely eroded stalagmite boss, but the whole had a rather dead appearance. In the roof considerable fault breccia and calcite spar was visible. Due to the danger of damaging the formations if they passed this point, it was agreed to leave exploration for the day and photograph the grotto before continuing.

The following day, Monday, Tim Large and Bob Craig went to the grotto to take photographs whilst Martin Mills and Pete Bowler started а CRG grade 6 survey of the new section. On the way up the incline Tim and Bob heard an odd noise; rather like the sound of water pouring into a bottle. The noise was traced to a couple of small holes at the top of the incline about 100 feet from the dig. They thought at the time that it might be water entering the pipe in the Gour Hall Dam, or possibly the Sump Passage Dam. On a later trip a gallon of water was poured down each hole and both emerged from the roof just upstream from the Sump Passage Dam.

After completing the photographs they decided to have a quick look ahead. Just beyond the grotto the passage veered to the right, and on the left a very fine white fluted flowstone wall was revealed. The floor consisted of white crystalline rimstone pools which no longer contained water. A little further along a straw at least 2 feet long hung from the roof. The passage turned sharp left at this point and a short steep slope led to what appeared to be a 20 feet pitch into a rift with open passage going both ways. They left further exploration until the next day when the pitch could be laddered.

The following day, Tuesday 26th May, a large party (Roy Bennett, Ray Mansfield, Bob Mehew, Bob Craig and Tim Large, followed by Dave Turner, Pete Rose and Nick Chipchase), arrived at the 20 feet drop having optimistically brought 50 feet of ladder. The pitch proved to be 21 feet and entered a 3 feet wide rift passage with fault breccia in the roof. To the left the passage was followed downhill for 20 feet where it ended in a mud choke. To the right the passage ascended steeply at least to the level of the top pitch, but ended abruptly after 65 feet at a sandy choke. At this point the passage cross section was very reminiscent of the far end of the Dining Room Dig and the survey would seem to indicate a connection (2). One side passage on the right, 20 feet from the end, was examined, but it became too tight after 12 feet.

With the exploration completed the total length of the new section, including side passages, amounted to some 400 feet. At the highest point the passage was about 60 feet above the level of the start of the dig in the streamway. Where oral connection was made with the sump passage, the roof of the sump passage was only 20 feet below the inclined passage. At the far end of the extension fault breccia was seen in the roof which appeared similar to that found in the Dining Room Dig, but the two passages were separated by over 100 feet. There would seem to be little to gain however, by making this connection. Water seepage appeared to be very aggressive as evidenced by the fact that the speleotherms were dissolving rather than growing.


  1. SMCC Journal Series 4 no. 8 pp3-11
  2. SMCC Journal Series 4 no. 7 pp3-6

RD Craig

June 1970

Figure 9 – Plan Survey of New Extension to St. Cuthberts II

Figure 10 – Elevation Survey of New Extension to St. Cuthberts II

Survey Note

The accompanying plan and extended section have been prepared from survey notes taken during two trips by R Bowler, T Large, R Craig, A Butcher, S Summerhayes and myself. The instruments used were a tripod mounted survey head with compass and clinometer read to ½ degree, 50 ft fibron tape read to 0.05 ft, and passage detail was taken at every station. The survey is therefore to CRG grade 6c, although the nearer section of the new extension has conditions which are not conducive to surveying to a high accuracy. The survey data was reduced to co-ordinates, from which the survey was plotted and drawn by myself.

The plan and extended section of the new extension are shown in relation to Gour Hall / Sump 1 section of the cave, the detail of which survey has been taken from BEC Caving Report No 13, part F (April 1969) and this source hereby acknowledged.

MT Mills


A Survey of Shatter Cave


This report has been prepared as far as possible in accordance with the recommendations of the Cave Research Group (1) and Mendip Cave Survey Colloquium (2). The recommendations of the latter have been followed only where they are additional to, and not at variance with the former.

Shatter Cave is a notable example in a long history of caves discovered in Fairy Cave Quarry, almost without exception as a result of quarrying activities. The Cerberus Speleological Society (CSS) who control access to all caves in the quarry, have officially numbered this cave as Fairy Cave Quarry Cave No. 15, but doubts have been expressed as to whether this number is correct.

The cave was first entered on the evening of 8th April 1969 by CSS members Messrs. Vaughan, Conway and Saxton, after several hours digging. They explored as far as the beginning of Erratic Rift including the chambers to the west of Diesel Chamber (3 & 4). The following evening a larger CSS party continued the exploration as far as the boulder choke at the end of Erratic Rift. During the morning of 15th April, representatives of the Bristol Water Works, Bristol Avon River Authority and the Quarry owners visited the cave under an agreed procedure for whenever a new cave is quarried into. Dr. WI Stanton, who accompanied the party, and T Frost found the route through the boulder choke into Tor Chamber.

The following evening a large party from several clubs visited the cave and explored as far as Pillar Chamber including the Ring Road alternative to the Z Squeeze. While this exploration was continuing CSS members RJ Whitaker and Miss S Bax commenced a preliminary survey from the entrance to the end of the shattered chambers parallel to the working face, so that a blasting limit could be imposed to allow exploration to continue.

On the next trip on 22nd April a hammer and chisel were taken to Pillar Chamber to enlarge a squeeze which allowed Messrs. Lavis, Prewer and others to continue exploration as far as Plughole Chamber. This trip was also the first visit to the cave of the writer and Alan Butcher, who ran a line survey using a 50 ft fibron tape and prismatic compass (read to the nearest foot and one degree respectively) only, from Pillar Chamber to the entrance (5). This survey was continued on the following evening when members of the BEC and SMCC, Messrs Turner, Irwin and Craig opened up a squeeze at the bottom of Plughole Chamber which gave access to the length of passage leading to the Terminal Choke and its offshoots (6). The survey had to be abandoned at the entrance to Plughole Chamber due to photographers at work, but it was felt that the amount of cave beyond was small compared with that surveyed and thus the purpose of the survey ie. to find the general direction of the cave, had been achieved.

Since that date a few small extensions have been made to the cave, and digs undertaken, but the sequence of discovery of the greater part of the cave remains as above.

Other Surveys

To date several surveys have appeared in various caving publications, mostly of CRG grade 1.

The first was, as mentioned above, carried out on 16th April 1969 by CSS members RJ Whitaker and Miss S Bax and was from the entrance to the end of the shattered chambers parallel to the working face. It is not known what instruments were used, but the information obtained was presumably the basis for (7) and later for (3). This survey which had no stated scale was reasonably accurate as to direction as far as the beginning of Erratic Rift, but beyond this point was based presumably on directional impressions and geological considerations, and shows the cave going east; whereas in fact the cave does not return east of a line due south from the entrance.

The above was followed by (9) which was similar to the above except that it was to a scale of about 1" = 180', CRG grade 1, showed the descending passage off, the top of Tor Chamber, and the cave beyond Fourways Chamber as running north from Plughole Chamber and not east as in (7) above. However in (9) the north point is in fact the direction of south, but the survey shows Shatter Cave in relation to Balch W/L Series, Conning Tower Cave and the remains of Balch Cave.

Because of differences of opinion between various people as to the true direction of the cave, (10) was prepared and circulated privately to interested parties. As it was thought that the cave might only have a short life, the author and Alan Butcher, who had already commenced a grade 6 survey, surveyed the main line of the cave (via Z Squeeze). If in the meanwhile, before the completion of the survey, the entrance was quarried away, then at least the whereabouts of the further reaches of the cave could possibly be located on the surface and an alternative entrance made. This survey to a scale of 1" = 30' was based on CRG grade 6 survey data of the main line plotted with protractor and scale only, and with CRG grade 1 detail of the main passage sketched thereon.

The next survey to appear was (11) a CRG grade 1 survey with no stated scale. This was largely similar to (7) but incorporated the additions made in (9) except for the direction of the cave beyond Fourways Chamber, and also a few further minor passage additions. As a result of (10) a note was added to the survey corresponding to the portion of the cave beyond Tor Chamber to the effect that "This section is now known to head South".

The above was followed by (12) which is a CRG grade 1 survey based on (7) at a scale of 1" = 120', but with the addition of the descending passage off the top of Tor Chamber. Next to appear were (13 & 14) which were based on (8) to scales of about 1" = 65' and 1" = 87.5' respectively, and are both similar except that the former has no north point.

The latest survey to appear (15) shows Shatter Cave, based on (10) with all known extensions to date, at a scale of 1" = 60' but no CRG grade is claimed. This survey also shows the cave in relation to the remains of Balch Cave, Conning Tower Cave and Balch W/L Series with the new extension.

Instruments Used

A survey head being a brass plate with two levelling bubbles at right angles, and compass graduated in degrees, read through a mounted prism column and an Abney level graduated in degrees with a vernier reading to ten minutes. The Abney level is arranged to swivel vertically parallel to the sighting plane of the compass for which it acted as a telescope on steep sights. It is possible that there is a small built in, presumed negligible, position error as a result of this. The survey head rests on a three screw brass levelling plate which in turn was mounted on a telescopic photographer's tripod with a ball and socket head. Distances between stations etc. were measured using a 50 ft Fibron tape graduated in feet and decimals. Offsets and station details were measured with a 10 ft steel pocket rule.

The compass was calibrated at one of two known points on the surface either before or after each survey trip. The bearings were compared with those obtained from the 6 inch and 25 inch Ordnance Survey maps for these two points. The Fibron tape was checked along its length with a steel tape, and no error of any consequence was found. Both compass and clinometer had previously been checked for scale errors. The clinometer was also checked taking fore and back sights between two fixed points as it became apparent from another field surveying experiment that there was probably an error. It was found that the clinometer was reading 1¾ degrees too low in elevation (too high in depression). This was corrected by adjusting the instrument, and all readings taken prior to adjustment werecorrected accordingly. The compass and clinometer were read to the nearest ½ degree, and the tapes were read to the nearest 0.05 ft.

The 'leap-frog' technique was used throughout and as mentioned earlier the main line through the cave starting from the entrance was measured first. Thus almost all stations, or at least all those stations on the main line adjoining side passages, had to be marked. For a few awkward minor side passages the instruments were hand held and read to the nearest degree.

The survey was made to the required standards as detailed by the Cave Research Group and Mendip Cave Survey Colloquium, and a grade 6 is claimed. Passage dimensions were recorded at every survey station and between stations where relevant to a substantial change in shape or direction. The detail is therefore classified as grade D. Where hand held instruments were used in a few awkward minor side passages the grade 6 standard is not maintained, but it is felt that the overall grading for the survey is unaffected.

Traverse Closures and Errors

Four underground closures were obtained. These were:

  1. Traverse 1:  Just south of the Tor in Tor Chamber, entrance to side     passage west of south end of Tor Chamber, to main passage en route to Pisa Chamber.
  2. Traverse 2:  Circuit of side passages and chambers west of south end of Tor Chamber.
  3. Traverse 3:  Higher and lower routes into Pisa Chamber.
  4. Traverse 4:  The Ring Road and Z Squeeze circuit.

Resulting in the following closures:


Traverse 1: Distributed along that part of the traverse that does not lie in the main passage.

Traverse 2, 3 & 4: Distributed throughout traverse.

The higher than expected values can perhaps be explained by the relatively short traverses. Closure errors were distributed proportional to the slope distance between stations. One further closed traverse was possible, this was the normal route into the cave and the low level route from the small chamber just beyond the Entrance to the far end of the crawl which leads to Canopy Chamber. However, because of the objective dangers from loose boulders in this low level route (undoubtedly more unstable than even the normal route into the cave) the writer was the only one of the surveyors able to pass the squeeze into the low level route from the entrance end; the grade 6 accuracy for this portion was not maintained, and this traverse was only approximately closed.

Correlation of Surface and Underground Surveys

As the extent of the quarry is constantly changing, even the most recent Ordnance Survey map, the 5 inch sheet ST 64 NE (1961 edition) was not unnaturally out of date in this respect. Therefore to determine the position of the entrance a traverse line was run to this to the same accuracy as the cave survey from a known point on the surface, namely the CSS cottage. Thus the position and the Map Reference (ST 6568 4752) of the entrance, have been accurately determined. The 'leap-frog' technique was used for the traverse line. Upon reflection it would probably have been better to take both fore and backsights between all stations in the traverse line to endeavour to minimise any magnetic anomalies resultant from the large amount of metal in the form of quarry plant in the vicinity.

The present extent of the quarry for the Location Plan, was checked on site by walking around the top of the quarry taking such measurements as are possible to remaining field boundaries etc. The results were compared with some aerial photographs taken in 1969 and kindly loaned to us by the CSS for this purpose.

Originally the altitude of the entrance was determined using a barometric altimeter reading to ±1 metre. Subsequently however, the opportunity arose to run a line of levels to the entrance from the Ordnance Survey Bench Mark on the NE corner of Phippen's Farm on the Withybrook to Stoke St. Michael road. The floor level of the entrance of the cave is 561 feet above Ordnance Datum. At the same time, and for future reference by others who may be involved in surveys of any future cave discoveries in the area, a Temporary Bench Mark has been established on the upper concrete base at the bottom of the concrete block wall at the SE corner of the tar boilers in the quarry. This Temporary Bench Mark is 557 feet above Ordnance Datum.

During the course of levelling from the Bench Mark to the cave entrance, as the route of the cave below ground had already been related to the surface, it was possible to take levels of the surface of the ground above certain parts of the cave, and thus discover the cover above the cave system. The surface ground level falls almost uniformly, but the levels at the following 3 points were obtained:

* less 4 ft of quarry top soil

Calculations and Plotting

The calculations of all 14 stations, with an average leg length of 20.3 ft were done using 5 figure log tables and were calculated to 2 decimal places of a foot, (rounded to one decimal place) and final results were expressed as co-ordinates related to magnetic north. The calculations were repeated a second time and any anomalies checked and corrected. The co-ordinates and data were plotted on good quality cartridge paper, from which the final drawing on Ethulon plastic tracing material was taken. The plan, extended, and cross sections were drawn to an original scale of 1/360 (30 ft to 1 inch). The grid on the plan has been related to grid north. Copies of the plotted survey and extended section were taken into the cave and the passage detail checked. Details of the floor covering were also taken at this time but due to the profusion of stal and boulders and the difficulty of differentiating between the two in places, and the prospect of a cluttered survey, it was decided to omit these details as it was felt that an outline survey only should be provided for specialists, who would then be able to plot on details of their own specific interests (eg. geology, etc.).

Symbols were used are in accordance with CRG recommendations and permanent survey stations, as detailed below, are marked on the survey with a triangle symbol, thus △.

Permanent Survey Stations

A number of these have been marked in the cave, by a hole about 1/8" deep made by a 3/8" stardrill. Note: co-ordinates given below are related to magnetic north; grid on plan is related to grid north.

The positions of other survey stations, though most are not marked, have been recorded in the survey notes and can be obtained from the writer.

Availability of Survey

Copies of the sheet comprising plan, extended and cross sections can be purchased through the cave survey scheme from BM Ellis, 7 School Lane, Combwich, Bridgwater, Somerset.


The surveyors wish to acknowledge the assistance given to them by the CSS and their arrangement with the quarry owners, particularly Bob Whitaker and Brian Prewer who have always made keys to the cave available; and Nick Chipchase and Pete Rose who assisted the writer in the survey of one of the more tighter parts of the cave. Without this help the making of the survey would not have been possible.


The survey was carried out between June 1969 and May 1970 by Alan Butcher and the writer, and involved 24 trips into the cave, plus 3 trips to take surface details and totalling in all 84 hours. In addition, at least twice this amount of time has been spent in calculating, plotting and drawing the results. The total length of passage in the cave is about 4000 ft, and the lowest and highest parts of the cave vary by about 94 ft.

All the survey and calculation sheets have been bound together and deposited in the club library for future reference.

Whilst the writer has assisted in cave surveys before, this is the first time he has conducted a survey to this grade. Both Alan Butcher and the writer have learned a great deal from carrying out the survey (which is the only way to really learn cave surveying), and trust that those who have had to wait while we completed our labours now find the wait worthwhile.


  1. BUTCHER, AL & RAILTON, CL, 'Cave Surveying'. CRG Trans. 8(2). June 1966
  2. Mendip Cave Survey Colloquium. 'Report of meeting', 31st August 1963
  3. WHITAKER, RJ. 'The Discovery and Exploration of Shatter Cave' CSS Nltr. (19), pp3-7
  4. WHITAKER, RJ, 'The Discovery and Exploration of Shatter Cave, (Fairy Cave No. 15) Fairy Cave Quarry', Mendip Caver 5(2), pp7-9 (July 1969)
  5. SMCC Hut Log 6, pp67-68. Entry dated 22/4/69
  6. SMCC Hut Log 6, pp68&70. Entries dated 23/4/69
  7. WHITAKER, RJ, Shatter Cave, CRG grade 1 survey, dated 20th May 1969
  8. ibid. Mendip Caver 5(2), 9. (July 1969)
  9. CHIPCHASE, NR, Shatter Cave, CRG grade 1 survey. Descent (5), pp18-19, (July 1969)
  10. BUTCHER, AJ & MILLS MT, Shatter Cave, CRG grade 6 main line survey, dated 9th July 1969
  11. Shatter Cave, CRG grade 1 survey, dated June 1969. CSS Nltr. (19), 21
  12. WHITAKER, RJ, Shatter Cave, CRG grade 1 survey, drawn by JD Hanwell. WCC Jnl. (124), p344. (August 1969)
  13. BUTCHER, AJ, Shatter Cave. SMCC Jnl. 4(7), p13 (June 1969)
  14. Belfry Bulletin, (261), p220. (December 1969)
  15. CHIPCHASE, NR 'A Plan of the South East Corner of Fairy Cave Quarry'. CSS Nltr. (21), pp8-9. (Autumn 1969)

MT Mills

May 1970



 Journal Series 04 Number 9