Journal Series 5 Number 4

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.



A Preliminary Investigation of Víðgelmir, Iceland: A Case for Cave Conservation

by MT Mills and C Wood

A Preliminary Exploration and Survey of an Ice / Snow Cave Kerlingjarfjöll, Iceland

by TA Prior

A Preliminary Exploration and Survey of Grettishellir, Iceland

by TA Prior

The Banpfyld Copper Mines

by T Hayman

A Survey of Dallimore's Cave

by MT Mills

Review: The Ulysses Factor


Published by the Shepton Mallet Caving Club

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



Following the prompt publication of the last issue, this issue seems to have reverted to the usual pattern.

It has been another year of visits to Iceland, and this issue seems to be almost entirely devoted to lava caves but not quite!

Difficulties in producing this issue are due in part to the present inoperable state of the Club duplicator but a little thing like that doesn't seem to stop us for long.


A Preliminary Investigation of Víðgelmir Lava Cave, Mid-West Iceland: A Case for Cave Conservation

The lava tube is situated at latitude 64° 44' 55" N and longitude 20° 47' 40" W, at an altitude of 250m (720 feet); in the northerly arm of the Hallmundarhraun, a post-glacial basaltic lava flow lying to the west of the Langjokull in the county of Myrasysla. The narrow part of the flow in which the cave is found in known as the Grahraun. The cave entrance is easily located and is best approached from the Gilsbakki-Kalmanstunga road, from which a well signposted track leads to the cave. Approach on foot, from the Surtshellir track in the north-east, which involves the fording of the Nordlingafljot river. The cave entrance is roughly located on the 1:100,000 and 1:50,000 maps of the Geodetic Institute of Copenhagen. Víðgelmir lies 5 km (3 miles) to the south-west of the famous lava tube cave Surtshellir / Stephenshellir, which occupies a higher part of the same lava flow (Mills & Wood, 1971).


Víðgelmir was visited by six members of the Shepton Mallet Caving Club in the afternoon of Wednesday, 16th August 1972, in order that a preliminary examination of the cave could be made with a view to assessing its suitability for detailed scientific investigation in the near future. The cave entrance was located with little difficulty and proved to be a roof collapse of large dimensions but subsequent exploration of the cave was halted in the downstream direction before daylight was lost, because the tube closed rapidly to two impenetrable squeezes. Recollection of a description of the cave and a quoted length of 1 km (0.62 miles) seemed to us to be a good reason for an attempt to force the squeezes. The left hand (south) squeeze proved to be far too narrow, though a more diminutive member of the party believed that he could squeeze through the right hand (north) passage. The ice at the entrance to the squeeze was cut with a geological hammer to allow a pool of water in the passage to be drained. The verdict upon now closer examination was that smaller members of the party would be able to squeeze between the lava walls, but should anyone get caught up on the jagged surfaces we would have great difficulty in freeing him. It was decided to cut the ice in the floor of the gap to enlarge the passage. In fact, after about 45 minutes digging with the geological hammer, a small trench was excavated which allowed ones shoulders and hips to be lowered and the narrower top of the passage was partially avoided. Even so, only three members of the party were able to pass through and continue the exploration of the cave.

Beyond the squeeze the lava tube cave was found to be extremely impressive with regard to its dimensions, its remarkable ice formations, and the profusion and beauty of its lava decorations. So impressive was the cave that it was decided upon rejoining the other members of the party some two hours later, to postpone our departure from the Hallmundarhraun for another day (Thursday 17th August), in order that the entrance squeeze could be enlarged and photographs and measurements of the cave be taken. There was in good deal of 'litter' found inside the cave beyond the squeeze, from which it was concluded that in the past the entrance to the cave was much larger, and that ice had been steadily accumulating each year in the squeeze. Indeed it was selfishly regarded as fortuitous that the entrance to Víðgelmir would soon be sealed completely by ice, thereby saving this impressive cave from the disgusting despoliation which has taken place in Surtshellir over the past years.

Although we had no research permit we felt that a cursory examination of the cave was therefore justified on two points:

  1. It was regarded by us that the entrance to the cave was in the process of 'freezing-up’, and that perhaps within a year ice would be blocking the entrance squeeze completely, closing the cave to future visitors.
  2. This lava tube cave was the finest example yet investigated by us in Iceland and possessed great wealth and beauty of ice and lava formations, some of which had been vandalised. It was hoped that a preliminary survey of Víðgelmir would be presented to the Icelandic Research Council with recommendations that a more detailed examination be carried out and some sort of conservation order to placed upon the cave.

Earlier Accounts of the Cave

The earliest mention of Víðgelmir is in Olafsson & Palsson (1774), which states the cave formed in a great lava field along with Fornurettur, the latter because of its notable width has a capacity for several thousand sheep. This reference is quoted by Preyer & Zirkel (1862). The account of the cave by Thorsteinsson (1963) is one of the most detailed and, referring to Olafsson’s work, Thorsteinsson mentions a cave called Fornurettir to the south of Arnarvatnsheide which holds several thousand sheep. This cave was a short distance from Víðgeymi or Víðgelmi. Olafsson also apparently mentioned another cave which was longer and, the story has it, swallowed a tributary of the Nordlingafljot river. According to Thorsteinsson this was our only previous knowledge of Víðgelmir, but he adds in addition that the cave has vastness, broad and tall below its roof and is nearly as long as Surtshellir. He mentions the occurrence of fine stalactites, but he has heard that many have been broken off by visitors, and he concludes that people need educating on how to treat antiquities and natural treasures such as these, in order to prevent damage to this notable cave. One concludes from this account that Fornurettir is perhaps a separate cave to Víðgelmir, though the only known cave which is likely to act as a sheepfold for several thousand sheep might be the entrance passage to Víðgelmir on either side of the roof collapse.

Miscellaneous references to the cave are many. Thoroddsen (1908) stated that the cove is in the lava field above FIjotstunga farm. Rothery (1952, reprinted in British Caver 1955) mentioned both Surtshellir and Víðgelmir, saying that they are both often visited, are about a mile long and contain ice formations. Corbel (1957 – summarized by Michowska, 1958, and Cascade Caver, 1966) states that the cave is situated in the lower part of the Hallmundarhraun, on the lower side of the Nordlingafljot river, 5km from Surtshellir. A map of the lava field shows the location of the cave, and it is also shown in an area of part stratified and block lava on the map of the Geodetic Institute. A black and white photograph of a view looking back towards the collapse hole from the bottom of the boulder slope leading to the squeeze may be found in Bardarsson (1965). The book by Samivel (1967) also contains a similar but different black and white photograph and states that not far from Surtshellir this huge chasm gives access to a vast passage filled with ice fossils, but does not actually give the name of the cave. Kjartansson (pers. comm. BM Ellis, 1969) pointed out that this was one of the three great lava caves of the Hallmundarhraun lava flow formed about 1,200 years. Atkins (1971) merely noted the existence of the cave, but it was not examined.

The Cave Survey

The survey took about four hours to complete, with Geoff Dodding as target lamp and tape man, Alan Butcher taking compass and clinometer readings, and Martin Mills booking the notes, which included sketch sections and as much detail as time would permit. Chris Wood also accompanied the surveyors making geological notes and assisting by taking some cross-sections. Dave Wilkinston and Mike Illingworth followed the survey party through the cave taking black and white and colour photographs.

The instruments used were a 100 feet 'Fibron' tape read to 0.05 foot, a Suunto compass (K8-14/360) and clinometer (PM-5/360 PC), the former read to the nearest degree and the latter to 0.05°. Distances and roof heights were estimated at every station, and even with the combined lights of three carbide lamps with large reflectors, there was at times considerable doubt as to the actual height of the roof, and to take account of this the roof line on the extended section has been indicated by a broken line accordingly. Likewise there was doubt about possible recesses at high levels along the walls of the cave and these have been shown in a similar manner. The 'leap frogging' technique was used throughout the trip. The survey figures were reduced to co-ordinates and a survey at scale of 1:2000 plotted from these, the accompanying survey (Fig. 1) at a smaller scale has been prepared from the larger survey.

It will be appreciated that due to magnetic anomalies which are contained in the basalt rock and will vary throughout the cave, the line of the cave will vary slightly from that we ascertained by our magnetic survey. However, as far as is known no other previous survey of this cave has been published and with the limited time at our disposal, it would not have been possible to complete a theodolite traverse. For these reasons, and also the possibility that the squeeze near the entrance may ice up so as not to readily permit future entry, it was felt that this survey was justified. According to our finding the total length of the cave is about 1585m (5200 feet) and the lowest point about 39m (128 feet) below ground level at the entrance.

Figure 1Survey of Víðgelmir


Víðgelmir is formed in lava of basaltic composition which has been dated by Saemundsson (1966) with C14 methods. Unfortunately, the two sources which most readily come to hand referring to Saemundsson’s dating of the Hallmundarhraun are contradictory (Atkins states the age of the flow to 775 ± 100 years B.P. and G Kjartansson in pers comm to C Wood, 1971, states the age of the flow to be 1190 ± 100 years B.P.) and the matter needs further checking. Atkins (1971) believes the Hallmundarhraun lavas were erupted from craters situated beneath, and immediately to the west of, Langjokull, and that these craters were active contemporaneously. Pahoehoe surfaces are typical near the cave and surface relief forms are represented by tumuli, pressure ridges, contraction fissures and collapse depressions.

The Hallmundarhraun has a total length of almost 48 km (30 miles) and for much of this distance is rarely more than 7 km (4½ miles) wide. The two important lava tube caves Víðgelmir and Surtshellir / Stephenshellir are located where the flow occupies the relatively constricted Nordlingafljot valley, at a distance of 33 km (21 miles) and 28 km (17½ miles) respectively from the vent. It is perhaps significant that in this region the flow is never more than 2 km (1¼ miles) wide and at its narrowest point between the two caves is less than 1 km (0.62 miles) wide. Of further significance is the gradients of the lava flow in the region of the caves (see figure 2). Speleogenesis appears to have occurred in regions of almost negligible gradient, there being a marked increase in slope between the two lava tubes and below the area occupied by Víðgelmir. It was shown elsewhere by one of the authors (Wood, 1971) that the primary reason for the occurrence of the lava tube cave Raufarholshellir in the Leita lava of S.W. Iceland was the existence of a marked increase in gradient of the pre-lava surface, which facilitated the draining of the tube after activity had ceased at the vent. The gradients of the lower section of the Hallmundarhraun may have played a similar role in the draining of the two large lava tubes in this region. Different flow conditions are shown to some extent by the nature of the surface forms, for in low slope areas around the caves relief features are subdued and relatively few in occurrence, while between the two caves the surface of the flow is a chaotic assemblage of collapse depressions and pressure ridges.

Figure 2Contour Map of Hallmundarhraun

The caves formed, therefore, in regions of relatively tranquil flow. The enormous size of Víðgelmir in particular may be partly attributed to the extreme narrowness of the valley at this point, and in the later stages of the emplacement of the flow this tube may well have conveyed the majority of the lava, which passed into the Nordlingafljot valley.

At the entrance to Víðgelmir collapse of the cave walls has provided a remarkable longitudinal exposure of the flow. Unfortunately, there was very little time in which an examination of the flow structure could be carried out, but on cursory inspection multiple units were noted and the flow structure bore a similar pattern to that found at the roof collapse above Surtshellir / Stephanshellir. Similarly it was discovered on numerous occasions within the cave that where the roof had collapsed ropey oxidized surfaces were observed on the upper surface of fallen blocks. Often this feature was extremely marked and it may prove to be an easy task of some later expedition to correlate the various surfaces at each breakdown pile and distinguish several unit forms and the direction they lie in relation to the cave.

The General Form of the Cave

Víðgelmir lava tube cave any be described as unitary (long, sinuous, unbranched and unilevel), but for the fact that loops appear in the main tube in three places and recesses in the walls at high level may prove to be blocked side tubes. At the lower end the tube splits and closes down to terminate in lava seals. The overall impression of the cave is one of vastness. Below are listed some dimensions:

Length, 1585 m (5200 feet or 1 mile)
Mean height, 9.15 m (29 feet)
Max height, 15.9 m (53 feet)
Mean width, 20.2 m (33½ feet)
Max width, 16.5 m (55 feet)
Max cross-sectional area, 10.7 x 15 m – 160.5 m2 (35 x 55 feet – 1750 square feet)

Entrance to the cave is made through the roof collapse which occurs in the east (upper) end of the cave. The extent of the collapse is enormous, being some 75 m (250 feet) in length and in places as much as 15 m (50 feet) deep and partly filled with breakdown. In one place where the roof remains standing it forms an impressive natural arch.

The tube in the eastern section is floored with a great amount of breakdown, is some 10 m (33 feet) in height and closes down after a short distance of 75 m (250 feet) in a breakdown pile. The thickness of the roof above this part of the cave is 12 m (40 feet).

To the west and downstream of the roof collapse, the cave closes down after 45 m (150 feet) at the base of the entrance boulder pile. Two small passages lie either side of what appears to be a massive blockage of fresh unglazed lava which almost entirely fills the tube. The south passage is far too narrow to allow human body to pass through, but it is vertically elongated and its continuation may be the blocked side tube observed once past the restrictive section of the cave. The passage by the north wall of the main tube is slightly larger, but is partly filled with ice. The shape of the passage is anomalous (see figure 3) and not the shape one would expect to be formed of flowing lava. This narrow passage gradually widens over a distance of 15 m (50 feet) and possesses a cross-sectional form for the whole of this distance which is roughly triangular in shape. This peculiar form of the cross-section also suggests that a mass of lava fills the original tube; the way past the blockage being a tight squeeze between it and the original cave wall. If on later closer examination of this feature this conclusion is verified, it may be well to remember that some Hawaiian lava tube caves are constricted by accretionary lava balls.

Figure 3 – Form and Possible Origin for the Entrance Squeeze

Beyond the restrictive section of the tube the roof rises to 13 m (45 feet) above the floor and an extremely beautiful part of the cave in traversed which is filled with ice formations, until a point is reached which is about 168 m (650 feet) from the squeeze where the roof quickly lowers to a height of 3.5 m (12 feet). Just before this low, 100 m (330 feet) of cave is reached, marked lateral benches are developed at a height of 5-6.5 m (16-21 feet) which give the impression that the low tube section a little farther on was part of a formerly double tiered structure. For some reason, in this part of the cave, only the lower tier remained open. A similar cross-sectional form to that above the low tube section is found beyond it and these have been shown in figure 4. From one part of the roof here, near the south wall of the cave, a steady downpour of water passes into the cave.

Figure 4 – Superimposed Cross Section Profiles

Beyond the low cave section and before the first loop is encountered, several interesting forms are discovered. Lateral benches are well developed at a height of 3m (10 feet) on either wall and these pass into large recesses which lie opposite one another across the tube. From both of these recesses extremely impressive glazed flow features are recognisable. Just below the recesses there in an abrupt increase in the gradient in the floor of the tube for a distance of 4.5m (15 feet). Before the loop is reached a step of 0.6m (2 feet) high is seen in the floor. Flow over this feature has resulted in contorted flow patterns and benches extend from the step for a considerable distance on both walls. Near the centre of the tube a channel some 1.6m (5 feet) wide has been eroded in the step by the lava.

The entrances to the larger loop tube in the south wall of Víðgelmir are situated 2.2m (7 feet) above the floor of the main tube and they exhibit a highly contorted and fractured flow pattern. The loop is just over 72m (235 feet) in length and about 4.5m (15 feet) high, and strangely for most of its length has a gradient towards the upper entrance. This loop tube is particularly noted for its wealth of speleothems, many delicate straw stalactites being over 60cm (2 feet) in length.

Speleothems are more numerous and are better developed towards the farther end of the cave. From the large loop tube to the termination of the cave, crusts of lava have pulled away from the walls and resulting small recesses have become glazed and filled with straw stalactites. In this part of the cave also, deep longitudinal recesses in the walls are found which are highly glazed, with smaller flow grooves superimposed upon their surface, many recesses holding lava speleothems. The smaller loop tube situated 180m (600 feet) beyond the first is relatively featureless and is little glazed.

Near the termination of the tube a small shelf-like feature is developed near the floor on both sides of the main tube, but because breakdown of the roof is more extensive here the lateral extent of this feature cannot be ascertained. Just beyond the smaller loop the tube loses height and remains 4.5-7m (15-23 feet) high, although the width of the main passage is maintained. Little of the lava floor is seen here for it is overlain with boulders which are mainly of an orange colour. At the edges of the tube, beneath the boulders, are found many fine speleothems. The lava tube terminates in lava seals in two separate passages, both of which are similar in appearance each with blistered roof and filled with lava speleothems.

The Ice Cave

Ice in Víðgelmir is found partly filling the tube for almost 168m (650 feet) and it commences at the base of the entrance boulder pile. The occurrence of ice in the entrance to lava tube caves in Iceland is a common feature as Watkins (1971) has shown in Raufarholshellir. Ice formation is generally attributed to the fact that the colder air of winter, being denser, will sink into the cave and remain there through the summer months. During the summer some remelting of ice stalagmites is commonly observed to take place by the dripping of warm water from the surface and many of these ice features have a distinctive summit crater-like form which results from much erosion. What makes Víðgelmir a special case from the other lava tube caves we have visited is the size and variety of ice forms met with in this cave. We believe the cave to remain colder than other caves through the summer because of the very small entrance squeeze (called appropriately by one member of the party 'deep freeze squeeze') which means that warm air has very little chance of invading the cave. Sadly, should at any time in the future the entrance passage be widened to allow more people to visit the cave, this unique micro-climate will be destroyed and, ironically, .so also will be the ice formations.

The part of the cave filled with ice is generally of large dimensions, glazed and is no different in form than the rest of the cave. Sheets of hard ice normally floors the tube, and this ice in places is covered by water of 10cm (4 inches) depth. Where there is no water the ice is seen to be of an uneven, course crystalline form, which makes walking across it safe. Occasionally, where warmer water dripping from the roof has attacked the ice of the floor, pits or deep holes have been formed. One such hole found by this party had depth of at least 60cm (2 feet) and showed that the floor was composed of solid sheets of ice some 5cm (2 inches) thick, each separated by a layer of water of similar thickness. Ice overlying water was not at all an uncommon occurrence in the cave. It was soon found to be extremely hazardous crossing pools of water which overlay sheets of ice, for more often than not the surface ice would collapse and one would be dropped into still deeper cold water. It may well be that the water which has dripped through the roof during the summer is frozen during the winter, but freezing of the water is not total because after a certain thickness of ice has been formed, the water below it is protected from the cold air that fills the cave. Thus alternating layers of ice and water may be built up.

Sometimes freezing of surface water was seen to take different forms. Near the exit from the squeeze (i.e. lower, inner most exit) the floor of the cave was seen to be constructed of coarse crystalline ice and was not covered with water. Overlying this ice is a loose veneer of radial, dendritic ice crystals, each cluster measuring up to 10cm (4 inches) in diameter, and linked to other such crystal clusters. Upon closer inspection these crystals formed an extremely beautiful regular pattern which suggests that they formed in completely quiet and undisturbed environment. This crystal veneer, it is thought formed either by the freezing of a thin film of water, or by direct deposition of water vapour carried in the air which passes into the tube from the squeeze. Crystals of ice formed by the freezing of bulk water have faces which are usually poorly developed and it is generally held that the best ice crystals appear when water vapour is deposited directly upon an ice surface. This latter principle may account for the formation of the crystals in Víðgelmir only near the entrance to the cave. A similar pattern of dendritic ice crystals to that found in Víðgelmir is represented by Figure 5. It is unfortunate that we had to destroy part of this veneer of crystal to gain access to the rest of the cave.

Figure 5 – A Pattern of Dendritic Ice Crystals

Ice stalagmites are common and generally reach heights of 1m (3 feet), though quite a few features are shoulder height and up to 50 cm (18 inches) wide at the base. Very few ice stalactites were observed. Cascades of ice were found to be common. Just beyond the exit from the squeeze, after a treacherous pool of water is crossed, thick sheet ice has formed a large cascade the total width of the passage, down the steep slope of a boulder pile. The cascade reaches the roof of the cave near the south wall and it was found to be extremely hazardous ascending and descending this slope. It is recommended that future visitors ascend the cascade near the north wall, where it is easier and where no damage can be done to this impressive ice formation. Some way beyond this cascade, near the south wall of the cave is found on equally impressive ice curtain. This fluted sheet of hard translucent ice in 5cm (2 inches) thick and 3.3m (10 feet) wide, and lies far enough from the wall to allow a man to walk behind it.

Of final note in the ice cave is evidence that frost shattering weathering processes are active on the roof and walls. In some areas overlying the white ice, flakes of basalt are commonly found with lengths ranging from 1-15cm (½-6 inches). These have obviously fallen from the roof and were not the result of normal breakdown, which occurs as larger blocks and results from the loosening of the jointed lava.

The Lava Decorations

The walls of Víðgelmir are glazed to a grey-red colour and show a wide diversity of decorative forms which developed because of the superficial remelting of the lava by hot gasses during the draining of the tube. Normal glaze features such as a rippled surface, blister, vertical ridges, flow grooves, etc., are well developed, and although not unique to this cave are particularly noted for their extensive development and variety of form.

Decorations are found throughout the whole length of the cave, but are most especially well developed in the lower third of the main passage and in the larger loop tube. Delicate lava stalactites and stalagmites are as large, and so found in much greater profusion, in Víðgelmir than in any other Icelandic lava tube cave we have visited. In fact, one passes from cluster to cluster of speleothems in a totally awe inspired manner and only photographs can illustrate these impressive forms adequately.

The occurrence of stalactitic forms are generally restricted to the walls of the cave and edges of the floor where the walls overhang. Stalagmites are mostly located in large clusters on the floor of the main tube, which in some places in almost completely covered with these forms from wall to well. In the larger loop tube where spleothems are best developed, stalactites occur mostly on the south wall and stalagmites litter the floor. Great care must be taken when walking through the cave not to inadvertently knock over tall stalagmites.

Lava stalactites in Víðgelmir are principally of the straw variety, and the many other forms appear to be derived from this basic typo. Clusters of straws which cover an area of 2 or 3 square feet are numerous and generally the individual stalactite has attained a length of over 30cm (1 foot). The most common form is the simple, straw stalactite, but these are rarely as long the erratic type shown in Figure 6a which may reach extreme lengths of over 1m (3 feet). It is obvious that the erratic form is a development of the straw variety, for on close inspection some straws show bulbous terminations (sometimes called 'tear-drop' stalactites) where a bulb of glaze passing through the straw has collected at the tip and solidified. In the erratic form, when drops formed at the tip of the stalactite, a skin developed around the droplet and this skin burst in one place to allow the still liquid glaze lava to drip downward. One may envisage, therefore, that in order for a skin to form around a growing drop at the tip of the straw, flow of glaze through the upper part of the speleothem was for a time diminished. The downward growth of erratic straw stalactites was thus a process whereby successive droplets were formed and enclosed in skin which subsequently burst. The erratic form is extremely common in Víðgelmir and many individual stalactites interlink to form a more complex structure. Numerous compound forms such as these are joined to globular stalagmites and form a beautiful column like structure. Another variety of straws of larger diameter are formed in similar manner to the erratic forms. These decorative stalactites (Figure 6c) appear to have developed from a succession of droplets which have subsequently drained and collapsed, thereby providing intricate permutation to the straw. Of more uncommon occurrence are 'pipe-stem stalactites' (Figure 6d) which result from the partial crushing of the straw on draining of the fluid interior.

Figure 6 – Lava Speleothems from Víðgelmir

Stalagmites are exclusively of the 'globular' variety and many reach a height of over 30 cm (1 foot). These speleothems litter the floor of the cave in certain places, with broken straw stalactites welded to the glazed patches which lie between individual forms. These speleothems are also commonly found bordering the walls of the cave and it is in such localities that the larger specimens are found. Globular stalagmites are known (unhappily) in the US as 'lavacicles'. They grow by the accumulation of glaze droplets which fall from the tips of straw stalactites.

A Case for Cave Conservation

Víðgelmir lava tube cave is the outstanding example of Icelandic caverns which we have so far examined. It possesses many unique features and is particularly notable for its impressive size, its wealth of lava formations and its distinctive ice section.

We are particularly concerned that these features should not be destroyed. Already there is a large amount of litter to be found in the cave and although some speleothems were undoubtedly broken by blocks falling from the roof of the cave, many broken formations are found for which there is no apparent explanation and one must reluctantly conclude that they have been broken by past visitors. Furthermore, it is the custom of the Icelandic caver to illuminate his way by means of a petrol soaked rag tied to a long pole. Petrol cans and torch's litter the cave and, worst of all, dirty petrol and charred remains of the torches have become frozen into the ice formations. Once out of the cave and into daylight we found that our hands were black with burned petrol which we had picked up from the ice. Chocolate paper, flash bulbs, polythene bags and other objects litter the cave.

Unfortunately it is a simple matter to find the cave and many people visit Víðgelmir and Surtshellir / Stephanshellir during the summer months. Indeed, whole coach parties have been observed at the caves, but while one would not wish to discourage people from seeing these impressive natural features it is hoped that some form of supervision is provided to safeguard the caves from despoliation. In the case of Surtshellir / Stephenshellir there is little that can now be done to save its natural beauty, for it must rank highest in the amount of garbage deposited in it and this cave is now almost totally devoid of lava speleothems. We are luckier with Víðgelmir for, fortunately, the entrance is now almost totally inaccessible to the general visitor. There may come a time, however, when the entrance ice may diminish or is dug out and the cave may be vandalised in the same way as its much larger neighbours.

Although we recommend that some sort of conservation order be placed upon this unique cave, it is apparent that only through locking up the entrance squeeze would access be limited to properly supervised caving parties. We have seen the destruction that has taken place in caves in the Snaefellsnes region which were protected by a conservation order. It is extremely fortunate, therefore, that Víðgelmir is itself locking way its inner treasures until a more suitable time when, perhaps, an Icelandic caving society can look after them.


  1. Atkins, FB "British Schools Exploring Society Report", 109p, London 1971.

  2. Bardarsson, HR "Island:Iceland:Islande", 108p, Reykjavik 1965.

  3. British Caver 26, p26, 1955.

  4. Cascade Caver 5(1), p2, Seattle 1966.

  5. Corbel, J "Les Karsts du Nord-Ouest de L’Europe", pp261-2, Lyon 1957.

  6. Michowska, A "Psuedokrasove zjevy na Islandu", Ceskoslovensky Kras 2, pp208-9, Praha 1958.

  7. Mills, MT & Wood C "A preliminary investigation of Surtshellir", SMCC Journal 5(1) Spring 1971.

  8. Olafsson & Palsson "Reise gienem Island", p353, Leipzig 1774

  9. Preyer, W & Zirkel, F "Reich nach Island im Sommer 1860", p106, Leipzig 1862.

  10. Rothery A "Iceland: Bastion of the North", p74, London 1952.

  11. Saemundsson, K "Zwei neue C14 Daterungen Islandischer Vulkanausbruche", Bandt 17, pp85-6, 1966.

  12. Samivel "Golden Iceland", Almenna Bokafelagid, p201&289, Reykjavik 1967.

  13. Thoroddsen, Th "Lysing Islands", p118, Copenhagen 1908.

  14. Thorsteinsson, Th Arbok 1962, p38, Reykjavik 1963.

  15. Watkins, EJ "The Micrometeorology of Raufarholshellir" CRG volume 13 number 4 pp257-260 1971.

  16. Wood, C "The nature and origin of Raufarholshellir" CRG volume 13 number 4 pp245-256 1971.

MT Mills & C Wood, December 1972.


A Preliminary Exploration and Survey of an Ice / Snow Cave Kerlingjarfjöll, West Central Iceland

Whilst recently on an expedition in Iceland, part of the programme was to discover, explore and survey a snow / ice cave in the Kerlingjarfjöll. There are, however, several reported sightings of these caves in the area although only one was visited.


Kerlingjarfjöll: Latitude 64° 37' 30" N; Longitude 19° 15' W; Altitude 1000 m (approximately).

The entrance to the cave may be found by following the gorge "Innri Asgardsa" high up on its eastern side, from the 'skidaskali’ on the north eastern slopes of "Asgardsfjall", taking the obvious track almost due south until it peters out, climbing all the while for a distance of approximately 6km, passing many steam vents on the way. Eventually one arrives on broken ground with a small cliff high up on the left, an outcrop of "Fannborg", and a gully running east-west across one's path. Almost opposite in the snow field a resurgence can be seen from a hole in the snow. This is the cave entrance; it's size may vary with weather conditions as was found when the entrance hole increased in size while inside the cave! One descends the gully to the river at the bottom, crossing this to ascend the resurging stream of warm water.

The General Form of the Snow / Ice Cave

From the entrance one enters a large chamber, in natural reflected light, which is perhaps best described as 'beehived' in shape with scalloping on roof and walls. The floor, however, was composed entirely of small scree. There was, at times, a considerable amount of steam filling the chamber. There were several smaller passages leading off as well as many alcoves. The main passage closes down fairly soon beyond the resurgence inside the cave, but a passage leading off to the left gave way eventually into a snow cave proper, ie. snow and ice covering the floor. This led into a crevasse-shaped rift where natural light could just be made out some indeterminable distance above.

There are three aspects concerning this cave which are unlike anything one is likely to encounter in other non-snow / ice caves. The first is that one ascends from the entrance when entering the cave, travelling up-slope rather steeply, unlike that of 'normal' caving where one usually descends on entering. The second point worth mentioning is that, in this case, the internal form of the cave appears to be kept relatively constant by the warmth of the resurging stream and that if one assumes that the snowfield in which the cave is situated is permanent, being added to from above by fresh snow periodically and therefore keeping the pressure of the snow fairly constant, a balance is maintained and probably only minor fluctuations in size and form actually take place (the author has at this stage no evidence to support this theory – yet it is worth considering). As stated earlier the entrance change shape rather rapidly once the group were inside the cave and one is tempted to consider that the group's presence may have had something to do with the fact that there were three snow collapses around the entrance in fairly quick succession, after the group had been in the cave for some considerable time, ie. hours. However one can only speculate.

The third point concerns the presence of some ice formations on the floor of the cave seen on a previous visit a week earlier, for on return the formations had all disappeared!

The Survey

The survey was carried out using a Suunto KB 14/360/R compass and a PM 5/360/PC clinometer; both were graduated in degree divisions and owing to the nature of the passage and surveying conditions, the instruments, being hand-held, were only read to the nearest half degree. Two 30 metre 'Fibron' tapes were used to measure the slope distance, passage width, ceiling height and eye height, and were read to the nearest centimetre. Passage details were taken at every station and at intermediate points where necessary owing to any significant change in passage detail or form.

The total length of the surveyed passage was approximately 100 metres.

Figure 7 – Survey of Kerlingjarfjöll Snow / Ice Cave


The survey of the Kerlingjarfjöll snow / ice cave was not completed due to a combination of inexperience of the party and a lack of lighting and time. However, a future visit by a more experienced group of speleologists would no doubt achieve completion and possibly carry out a more detailed study of the cave.


The author wishes to acknowledge the assistance of BM Ellis who has drawn the survey of the Kerlingjarfjöll snow / ice cave.


  1. Prior, TA (et al) "Iceland '72 Expedition" Report. Bridgwater Area Youth Committee, Somerset Youth Service. Pub Somerset Education Authority November 1972.

  2. Kidson, P "Climbing and Skiing in Iceland". Pamphlet Pub Icelandair, 2nd edition.

  3. Anon "Iceland". Pamphlet Pub Icelandair.

  4. Map 1:100,000 Uppdrattur Islands. Blad 55 Hveravellir and Blad 56 Kerlingarfjoll. Pub Landmaelingar Islands, Copenhagen 1956.

  5. Barth, Tom FW "Volcanic Geology, Hot Springs and Geysirs of Iceland". Pub Carnegie Institution of Washington, 1950.

  6. Mills, MT "Non-Lava Caves of Iceland" SMCC Journal 5(1) Spring 1971.

TA Prior, October 1972.


A Preliminary Exploration and Survey of Grettishellir,
West Central Iceland

Whilst recently on an expedition in Iceland part of the programme was to explore and survey the lava tunnel of Grettishellir.


Grettishellir: Latitude 64° 48' N; Longitude 19° 27' 30"W; Altitude 660 m.

The entrance to Grettishellir may be located in the Kjalhraun, a lava field of 'block lava' midway between the Langjokull to the west and Hofsjokull to the east. If one starts from the south-western tip of "Porisvatn" and walks on a true bearing of 228° for approximately 1.5km and looks for a cairn now situated on a rise adjacent to the collapse and faces west, the entrance should be located immediately in front. There are many other cairns in the area, some marking a track running north/south while other are just marking high ground on the lava flow. The plethora of cairns appears to be an old Icelandic custom! However with careful use of the map, one can locate and identify one’s position. When the summit of "Rjupnafell" is directly to the north and "Strytur" is directly to the west one should be in a position to see the cairn marking the high ground above the entrance collapse.

The Lava Flow

The Kjalhraun, a post-glacial basaltic lava flow, probably has as its source the volcanic crater "Strytur" immediately west of the entrance. The lava is 'block lava' (Icelandic "Apalhraun"; basalt lava which supports a little vegetation in the form of lichen and heather). The whole of the lava field is monotonously undulating with its surface forms being largely represented by tumuli (mounds formed by pressure from liquid lava under the surface) most of which were cracked open at their summits. Other features to be seen were 'depressions' and 'contraction cracks'.

The General Form of Grettishellir

The entrance is undistinguished being a small, narrow tunnel collapse partially filled with debris from the roof collapse and very approximately in the middle of the length of the tunnel, with passages going off in opposite directions. For convenience, the 'down flow' section was termed the East Series and the 'up flow' section the West Series, although these are only approximate directions from the entrance.

For the most part the tunnel's inner surface was entire, being covered with glaze as well as a profusion of stalactites and stalagmites. The stalactites either formed as drip pendants, straw, rods or helictites and varied in length from a few centimetres to a third of a metre. The stalagmites were either conical, globular, flanged or a 'mass of entwined and ramified narrow lengths of rods twisted together' ranging in height from a few centimetres to a metre.

The tunnels tended to be low and wide with a flat floor and arched roof giving very little standing space. In the East Series was found an area of 'sand' which reflected a yellow colour underground yet black when brought to the surface. Also a section of smaller tunnels was found containing clear ice which choked the extremities of these tunnels, although one could see further into larger chambers.

The West Series was considerably larger in size and shape which contained several reasonably sized chambers, and at one extreme, ice was found which was crystalline in nature, also giving a restricted view into a larger chamber. There were no ice formations as seen in Raufarholshellir in 1970.

Lateral benches and drainage channels were also seen in various parts of the tunnels yet not one appears to be continuous over any appreciable distance.

The Survey

The survey was carried out using a Suunto KB 14/360/R compass and a PM 5/360/PC clinometer; both were graduated in degree divisions and owing to the nature of the passage and surveying conditions, the instruments, being hand-held, were only read to the nearest half degree. Two 30 metre 'Fibron' tapes were used to measure the slope distance, passage width, ceiling height and eye height, and were read to the nearest 100th of a metre. Passage details were taken at every station and at intermediate points where necessary owing to any significant change in passage detail or form.

The total length of surveyed passage was approximately 400 metres. The total length of time spend surveying was approximately thirteen hours, made up of three sessions.

Problems of magnetic disturbance were encountered when surveying using the compass although this phenomenon was expected as similar problems were experienced when surveying Raufarholshellir. A paper was written concerning these magnetic anomalies.

Figure 8 – Plan Survey of Grettishellir

Figure 9 – Elevation Survey of Grettishellir


The survey of Grettishellir was not completed due to a combination of inexperience of the party and a lack of lighting and time, however, a future visit by a more experienced group of speleologists would no doubt achieve completion and possibly carry out a more detailed study of the lava tunnel.


The author wishes to acknowledge the assistance of BM Ellis who has drawn the survey of Grettishellir.


  1. Prior, TA (et al) "Iceland '72 Expedition" Report. Bridgwater Area Youth Committee, Somerset Youth Service. Pub Somerset Education Authority November 1972.

  2. Map 1:100,000 Uppdrattur Islands. Blad 55 Hveravellir. Pub Landmaelingar Islands, Copenhagen 1956.

  3. Watkins, EJ "The Micrometeorology of Raufarholshellir" CRG volume 13 number 4 pp257-260 November 1971.

  4. Ellis, BM "The Survey of Raufarholshellir" CRG volume 13 number 4 pp235-244 November 1971.

  5. Bowler, PGB "Magnetic Anomalies around Raufarholshellir" CRG volume 13 number 4 pp261-264 November 1971.

  6. Corbel, J "Les Karsts du Nord-Ouest de L'Europe" Pub Lyon 1957.

TA Prior, October 1972


The Banpfyld Copper Mines

These mines are situated three miles north of North Moulton near the village of Heasley Mill

The mine consists of four shafts, two of which are situated on the west if the road, reached by climbing the waste scree slope. The other two are in the trees on the east side of the road (not examined for this article).

The farthest shaft (map ref 734322) was first visited last year when I noted that a crane had within recent months re-opened the shaft.

On Sunday November 5th a group comprising of John Blasdale, John Elliot and myself revisited the mine. We needed 106 feet of ladder to reach the bottom. From the bottom, two passages diverged in opposite directions. The larger was followed in knee deep water for about 75 feet, when the passage opened out and a shaft was seen disappearing vertically in the then clear water. We traversed our way around the shaft and followed the old worked passage for about 200-300 feet to an abrupt end.

We returned to the main shaft and then followed the smaller passage, At first we walked almost upright but soon the passage got much smaller and we were soon crawling on all fours in the thick mud and water. We noted that the passage led steadily down hill and contained quite a little stream of water. Eventually after 300-400 feet we met still water which gradually got deeper until there was only a few inches of airspace between the water and the roof. It may have been possible to "push" the passage further but floating on my back my head bumped into a decomposed hedgehog which rather put me off (to say the least).

It looked to me as if there has been a roof fall at some time which is holding back the water, however the blockage must be small as the water does not rise to roof level. As the water obviously has an outlet I would think it highly probable that this particular passage is probably a drainage adit with possibly a connection to the larger mine shaft nearby.

The nearest shaft (map ref 734326) was visited a week later on Sunday 12th November with John Blasdale, John Elliot, Bob Burns and myself. This shaft is supposedly 672 feet deep according to the book "Days of Renown". 112 feet of ladder were suitably belayed and Bob volunteered to be first down. He descended the ladder at a very rapid pace and soon reached the end and called for more ladder. The last 35 feet were lowered to him. He had a very tricky job of fixing it on as the ladder was free hanging all the way. Bob soon reached the ladder end again and told us that a recessed passage left the main shaft a few feet below him. However the main shaft appeared to be another 50 feet deeper. Having no more ladder Bob climber out, being very relieved to stand on solid ground again.

It would appear that the main shaft is 200 feet deep; however it could have been a false bottom formed from joined rails, dead sheep, rubbish, etc. It is also possible that the apparent bottom is only a landing stage with a further descent staggered from there.

Tim Haymen, October 1972


A Survey of Dallimore's Cave, Central Mendip

Dallimore’s is one of those smaller Mendip caves that everyone has heard of but relatively few appear to have actually visited. The entrance is in fact situated in a shakehole (one of a group of five) in the field SW of the Hillgrove junction (Priddy turning) on the A39, and permission to visit the cave should be obtained from the nearby Ores Close Farm.

Previous writings about the cave are somewhat limited, and in fact only three items contain any useful information. Low (ref 1) states the entrance at the bottom of the 18 feet shaft was opened on 3rd October 1948 after only four weekends work, a long low horizontal squeeze prevented further progress until the following day when two persons passed this to reach the lower of two bedding planes. Prospects for further progress were judged as not very promising. Wells (ref 2) reports that in February 1955 the cave was open again, the farmer having tipped some rubbish down the shaft some time previously but did not object to the cave being dug open again. Towndrow (ref 3) records that during Whitsun 1966 a party from Blackmoor Caving Group found the shaft only 12 feet deep, and after 2 feet of excavation the small rift on the west side of the shaft was uncovered and for a time digging was concentrated here with no success. At a depth of 18 feet in the shaft the squeeze on the east side was revealed, Aven Chamber was reached but only on the next trip was the final rift located.

Previous surveys of the cave are likewise limited, it is understood there is an ungraded section without scale by CD Smith dated 23rd July 1953 in Sidcot School Speleological Society Logbook 6 and an ungraded section of the entrance shaft in Wessex Cave Club Log for 1949. However it is interesting that in all three of the above accounts surveys are mentioned. Low (ref 1) states that the cave of 150 feet of passage and it is hoped to publish a survey at a future date, and Towndrow (ref 3) says it is proposed to make a survey – nothing more is known of these projects. Wells (ref 2) reports that the cave has about 250 feet of passage and according to the survey is just 93 feet deep. Since our own survey gives a depth that agrees exactly with this figure it is fair to say that this survey was completed but unfortunately appears never to have been published.

The writer first visited the cave with Martin Webster on 23rd August 1969 when the crawl through the muddy pool approaching Aven Chamber was found to be blocked by flood debris (presumably from the Great Flood of July 1968) and was dug out. The accompanying survey was made during four trips to the cave with Bob Mehew on 26th February, 13th May, 8th July and 9th September 1972 and on the latter occasion we were also assisted by Steve Summerhayes. The survey was carried out by 'leap-frogging' using a Suunto KB 14/360 compass and PM 5/360 clinometer. Both were read to the half degree and were hand held, although care was taken to minimise errors due to station movement – the use of a tripod mounted survey unit had been discounted due to the constricted nature of much of the cave for if used this would have greatly taxed the time and patience of the surveyors. The 50 feet Fibron tape was read to the nearest 0.1 feet and the passage details were taken at every station. The survey figures were reduced to co-ordinates using four figure logarithm tables, and the survey plotted from these co-ordinates. Rough plots of the survey were taken into the cave and passage detail plotted on these. The total length of the passage is about 720 feet, and the total depth 93 feet. In addition, various surface measurements were taken to accurately pinpoint the entrance and the relation of this to the adjoining shakeholes.

Due to the complex plan of the cave resulting from the various levels in the vicinity of the final rift certain passages have been omitted from the extended section, together with a small side passage near the entrance in order to preserve clarity. It is intended in due course that this survey will be available through the Mendip Cave Survey Scheme and further copies may be purchased from BM Ellis. During the making of this survey no evidence was found of CO2 concentrations, said to sometimes occur in the lower passages mentioned by Barrington & Stanton (ref 4). The description of the cave in the latter work is necessarily brief and, in the opinion of the surveyors, difficult in parts to reconcile with the features of the cave and thus a more adequate description is given below:

"Entrance shaft 18 feet deep followed by a squeeze (occasionally becomes blocked and may need digging out), a descending crawl with avens in the roof leads to inclined rift passage and crawl through muddy pool. Beyond on left opening in roof connects to top of the rift, straight ahead is climb down over boulder to Aven Chamber. Above ledge in roof, aven 15 feet high leads to short horizontal passage, continuing down chamber, in right wall crawl leads to high rift and crawl beyond, a crawl with squeeze leads to small chamber and passage continues in opposite wall as flat out crawl descending to become too tight.

From Aven Chamber, back under boulder leads via step down to wide flat out crawl ending in rift, this can be climbed at right end via chockstone boulders about 20 feet to ledge with boulders and connection back to approach to Aven Chamber, and steep uphill tight crawl in opposite wall over rift. The rift also goes down about 30 feet, and short passage on left contains the best of the few formations in the cave, to horizontal mud floored rift extending in both directions at the farther end is the dig previously excavated by WCC & MNRC".

Figure 10 – Plan Survey of Dallimore's Cave

Figure 11 – Elevation Survey of Dallimore's Cave


  1. Low, CC Wessex Caving Club Circular. New Series 15. January 1949.

  2. Wells, CC Mendip Notes Wessex Caving Club Journal 49 p9 February 1955.

  3. Towndrow, P Cave Digs: Dallimore’s Cave, Mendip Caver volume 2 number 9, p76 July 1966.

  4. Barrington, N & Stanton W, The Complete Caves of Mendip 1972 2nd Edition p52.

MT Mills, October 1972


Review: The Ulysses Factor
By JRL Anderson, pub Hodder and Stoughton (1970) @ £2.50

This work has no direct connection with caving, but its subtitle will perhaps explain its relevance – The Exploring Instinct in Man. Occasionally one may half-heartedly endeavour to analyse the various facets behind, as the song says, "the reasons I go caving" etc., and it is these impulses that the author calls the Ulysses Factor.

In an endeavour to discover, explain and define the factor the author has taken the more notable adventures of three mountaineers: Herzog, Tilman and Shipton and nine sailors: Reyerdahl, Chichester, Ridgway, Lewis, Manry, Knox-Johnston etc. and analyses the decisions and actions made by them with frequent references back to mythology and the type factor. The experience of the author as a sailor following the route of the Vinland Sagas makes him perfectly able to appreciate and comment on the exploits of the sailors, but he is obviously less at home with his choice of mountaineers and everyone will challenge he names of certain of the twentieth century adventurers included in, or excluded from, the appendix to the book.

There are chapters dealing with the type figure, historical evidence, mutation, the factor in war, nationalism and in addition current trends, commercialism and sponsorship are examined and crude statistics of lone Atlantic sailors produce interesting national figures of adventurers and the significance of the increase in their numbers since the second World War.

The nature of the Ulysses Factor is complex and the authors observations can be considered throughout as generalisation, he fails to analyse it to the extent of specific definition. Personally having searched its contents I cannot explain the nature of the factor any better now than when I started, on the other hand it could equally well be called the urge to explore rather than the Ulysses Factor. Many of his adventurers are loners and these he obviously admires particularly the more elderly, and does not intend the new-age adventurers to invade this image when he says of astronauts "Men have got to the moon, which is an obvious triumph, but they have neither walked, or swum, nor sailed there: the huge industry of technology which got them there is as emotionally meaningless as a cement works".

However, notwithstanding my doubts this book contains much of interest and note. The only other work that I knew dealing with this subject, although approaching it from a different aspect, is Wilfred Noyce's 'The Springs of Adventure' which I would recommend to be read as a companion volume. Although I have not discovered through this book exactly why I go caving etc., I have found many indications of other facets that require exploration and self examination. As far as the written word is concerned I can but close by quoting Tilman, as is quoted in the book, as an indication of the appropriateness of some of the text to caving "Yearly … so many calls are made upon the various rescue organisations that by now the average man should be ashamed to think of adding to their number. The confidence that is placed, and successfully placed, in being rescued fosters carelessness or even foolishness, and condones ignorance ..."


 Journal Series 05 Number 4