BSA Cave Science Vol. 4 No. 29, July 1959, pp 206-228

Report. on the Cambridge University Caving Club
Expedition to SVARTISEN. Norway 1958.

by D.A.Jenkins


The following is a report of the third successive expedition to the Svartisen Ice cap, by the Cambridge University Caving Club, in the summer of 1958. The Ice cap borders the North sea and lies virtually on the Arctic circle (Map 1). In the five weeks spent there, the general exploring, surveying and photographing of the caves of this region, carried out by the two preceding expeditions in 1956 and 1957, was continued. In 1958, however, instead of combining a study of several caving areas, glaciology and mountaineering, work was limited to one narrow band of limestone.

The limestone band involved follows a small valley, Pikhaugene, which is cut into the shoulder of Kamplitind, the latter forming the eastern side of the main Hlaamaaga valley (Map 2). To the North towards the Svartisen peaks, the limestone band passes along the western shores of a small un-named lake and, 2km farther North, Pikhaugvand, before dropping down towards the Flatisen glacier. To the South, it leaves Pikhaugene, curving over the shoulder of Glomvasstind and then dropping over 300m to the lake Glomvand below.

Towards the end of the 1957 Expedition, Pikhauggrottene - two caves within this limestone band in Pikhaugene, were explored. These caves were briefly described by Horn (5) and considerable extensions beyond the survey he shows, were found. A new system was also found farther south, draining the small un-named lake. By means of fluorescein tests, it was shown that the water entering this system reappeared ten hours later at a resurgence on the shore of Glomvand 400m lower, and roughly 3.5km away. Little time was left in 1957 to explore this system, named Fosshullet on account of the waterfall at its entrance, or Pikhauggrottene, and their full exploration and surveying were therefore the main aims of the 1958 expedition.

Fosshullet was the first to be explored, but unfortunately it was found to be only 400m long, ending at an impassable sump. Some time was spent in an attempt to dig through some small silted passages leading south but these were not very promising. A grade 4 survey of the system was also made. Pikhauggrottene were then explored and further extensions beyond the 1957 survey found. Unfortunately, about this time the weather deteriorated unpleasantly, limiting activities. Exploration in Pikhauggrottene continued however, and still further discoveries were made whilst it was being surveyed and photographed. On the way back to Glomvand, attempts were made to gain access to the inaccessible Fosshuller-Glomvand system. Several small caves, and one large one, named Storbekkengrotte were found, explored and surveyed incompletely. Considerable possibilities therefore remain for further discoveries in the lower half of this drainage system and to a lesser extent in Pikhauggrottene, although the possibilities in Fosshullet were fully exhausted.


The Svartisen area falls within the main caledonian Orogenic belt, the rocks being deformed and metamorphosed. The Pikhaugene-Glomvand area forms the western rim of a regional basin structure and the strata therefore dip down to the S or S.E. at angles varying between 30° at Glomvand to 70° in North Pikhaugene. The rocks involved are those of the North-Norwegian Schist series, together with Caledonian igneous masses intruding them. The schists vary from calcitic to pelitic types and are generally of low-moderate metamorphic grade. Thus occasional thin bands of marmorised limestone (1 - 15 in. thick) are interstratified with - Garnet, Biotite-Muscovite, or Biotite-Tremolite schists. Intermediate semi-calcareous types are also found, as in the section along the Glomvand shore. An intrusion of granite is exposed just north of Pikhaugene, well exposed on the flanks of Sniptind and in the Flatisen valley. It is normal Biotite granite but with local concentrations of ilmenite and magnetite, and rare crystals of molybdenite. Associated with it are thermal metamorphic assemblages and also a series of Hornblende-epidote-plagioclase Epidiorites. A highly porphyritic sill caps Sniptind, and provides distinctive glacial erratics to Pikhaugene and Glomvand.

The limestone band in Pikhaugene consists of fairly pure calcite, weathering a light grey colour. Under the microscope, it shows a marmorised texture of large (0.5cm) interlocking grains. Towards the edges there is a small zone (1 in. wide) in which thin bands of schist and limestone are interstratified. Occasional thin schist bands and discontinuous lenticles of Quartz also occur within the main mass of limestone and small lenses of secondary quartz fill some of the joints. In one section, between Pikhauggrottene and Fosshullet, the limestone band is acutely folded into an Z shape, in which the limbs have been attenuated, implying plastic deformation. Apart from this, the structure of the band is simple.

As a topographical feature the limestone band is conspicuous. In Pikhaugene, where its dip is 60° - 70° and it outcrops along the valley bottom, it protrudes above the adjacent schist as a series of "whale backs" smoothly rounded hummocks of bare rock, about 50m along and 5 - 10m high. Jointing is patchily developed, chopping up the limestone into regular blocks in some places, whilst others, notably the deformed section, are relatively immune from jointing.

Weathering has resulted in a number of different features. On many of the "whale backs" long longitudinal grooves are present, although they are often obscured by other forms of erosion. They are continuous over long distances - 10m or more, and roughly horizontal parallel to the strike of the limestone. They are not however, controlled structurally by the limestone, as in places, they can be seen to transgress the thin schist bands within it. It is most likely that they, and the "whale-back" forms are relics of glacial erosion. Most prominent are the vertical runnels or lapies, which cut across the grooves described above. These are clean and sharp and very well developed - obviously active under present day subaerial erosion. Superimposed on both these larger scale features in near horizontal areas, are small sharp flutes, closely spaced - about 2 cms apart and 1 cm deep. These are not common in English Karst and are probably due to nivation, with the slow prolonged supply of melt water from snow-patches

To the south at Glomvand, the topography of the limestone outcrop differs. This portion is below the tree line, and thickly vegetated with birch and bilberry bushes, unlike Pikhaugene, where the limestone is bare except for patches of lichen or moss. As the dip of the limestone is less (about 30°) scarry faces and dip slopes are developed as obvious features and the belt of the outcrop is broader. Lapies are again well developed, draining into extensive deep grikes, but the long grooves and the small scale fluting are absent, and, in many places the superficial layer of limestone has been reduced to a fine rubble.


The Pikhauggrottene system lies within a marked rise that separates Pikhaugvand from the smaller un-named lake farther south. The single limestone band, to which it is confined, varies both in dip and width along its length; to the south the dip is 62°, increasing to 68° in the north, and the width fluctuates in an arbitrary manner between the limits of 17.6 m and 22.2m in surface exposure (i.e. stratigraphically 17.2m and 20m). Over the length of the cave, it is generally thicker in the north. The following observations on the cave were made during some 60 hours spent exploring, surveying and photographing it.


As shown by the plan, those passages explored generally run straight along the strike, and, are superimposed vertically one above the other in successive inter-connected, "series". The dip influences the plan in that the passages tend to move towards the "hanging" schist face, rather than the other, and hence lower passages lie farther to the N.E,. than those overlying them. When a passage comes up to the schist-limestone interface, it often continues into the solid schist for a metre or so, which is reduced to a soft, rotten condition - as in the "coal-mine". The passage is then deflected back into the main limestone band, though often moving diagonally upwards along the interface, for a few yards, before leaving it. The passages can be roughly divided into four main types:

The first type predominates in the northern section of the cave and in the smaller passages, well developed in the Aven series. These passages are governed by one of a set of joints orthogonal to the bedding plane. This joint dips NW at 30° and is developed into a bedding-plane type cave by central enlargement (fig. 1).

To the south, this type of passage grades into a larger, more open type - the "Trykkledning" of Horn (5). This has an elliptical cross-section the major axis of which coincides generally with the bedding or occasionally with a joint plane. Good examples of such passages are Horn's level, the final passage in Midnight series, and the middle level below Coffee Pot (fig. 2). The near vertical shafts or slits are related to the vertical, component of the same orthogonal set of joints. They are developed on a lesser scale and are seen in the Aven series, where they form connections between series at different levels. The keyhole is an excellent example, tight and rather exhausting (fig. 3). The fourth type - that of steep angle rifts, is determined by another joint system, dipping down the strike of the limestone, ie. approximately 30° to the S.W. Such joints can be seen as steps in passages of type (a), the passages changing in level across them by up to a metre or more. They are developed into connections between series at different levels, especially within the mud choke area (fig. 4). Here, one is also enlarged till it has a roughly elliptical profile similar to that of type (b) - (fig. 2).

[figs 1-4 and 6-8 here (there appears to be no fig 5)]

These passages are inactive with the walls stained a light brown, and containing sediments where the angle of slope is sufficiently low for it to accumulate. A noticeable feature is the scalloping of the walls. The scallops vary in size, but are generally of the order of 15cm - 30cm. In the Aven series they are up to 60cm across and markedly assymetrical ie. cross sections parallel to direction of passage as shown in fig. 6.

The cave system inevitably accepts the local surface drainage, which is only of significance in wet weather. This enters through a number of small inaccessible inlets which usually come down the S.W. limestone/schist interface. They pass below the main series of the system and can occasionally be heard running underneath. The limestone where such water is moving is a clean, banded grey-white, and typically eroded into small scallops (2cm - 5cm) and sharp fluting.

The walls of the 17m Aven are very sharply fluted and ridged where the water enters the Aven series, and the underlying series are clean washed and sparsly scalloped. "Petty Cury" is similar in places, and at one point a small vadose trench is cut into the floor of a "bedding plane" passage. Such modification has also occurred in the entrance of Horn's level with the incision of a small, deep meandering trench into the floor of the elliptical cross sectioned passage (fig. 7).


The sediments of most grades are represented in Pikhauggrottene. In the Aven series, the sediment is usually a dry, fine, light buff sand. The grade is <0.5mm the grains subangular to rounded and monomineralic. The dominant minerals represented are Calcite, Quartz, a pink Garnet, Kyanite, Clinozoisite, Biotite, green Hornblendes and occasionally plagioclase Felspar. This can be contrasted with the degradation product of the schist which forms a mound of sand in the Middle level below Coffee Pot. In this, the grade covers a larger range. The grains are more angular and there are occasional polymineralic rock fragments. The mineral composition also varies in that Rutile, Quartz, Plagioclase, Zircon, Muscovite, Biotite, Cordierite and a green Hornblende are present whilst both Calcite, Garnet, Clinozoisite and Kyanite are absent. This contrast shows a non-local source and modification by transport on the Aven series sediments. The latter correspond to a relatively high grade contact metamorphic assemblages such as occur around the granite intrusion to the north. For example, a clinozoisite garnet rock was identified on the shoulder of Sniptind, and this locality was probably the source for the clinozoisite found in the sediments. Originally it was hoped to connect sampling of the various sediments in Pikhauggrottene with a geological surface survey in an attempt to identify and locate the corresponding sources. Unfortunately, time and weather prevented the surface survey being done.

In the part of the Petty Cury series adjacent to the mud chokes, the deposits show interesting minor structures neatly exposed where a bank of sediments has been eroded by a small stream at some stage, to give a small scarp (fig. 8). There is a crude grading of the bedding from coarse sands to fine laminated silts, overlying gravels. Current bedding is well developed, and in places small funnels of collapsed sediment show where there has probably been a drainage pipe down through the sediments.

Rounded stones and unsorted gravel are present throughout Petty Cury. The pebbles and stones have the usual dark grey-black surface characteristic of stream deposits, which is probably of organic origin.


Loosely included in amongst the pebbles of Petty Cury, the peculiarly shaped concretions first noted in Nordre Hamarnesgrotte, were again found. They are complex lobed shapes, essentially planar and roughly 0.5cm thick and 10cm x 5cm in area (figs. 11-12). The composition may be one of two types: the material is either a fine grained sand showing no internal structures or it is a fine silt that is banded parallel to the planar surfaces, the banding showing no modification to the external shape at the edges. Both show limited reaction with hydrochloric acid, suggesting calcareous cement. The surfaces have no calcite covering and the two sides are dissimilar. One is undulous, approximating to planar; the other, in the coarser variety, is more lobed and ridged, and in the finer variety is "contoured" along the banding, showing more delicate detail. In some there is also an elongation of the lobes, indicating a directional factor operative during their formation. Examples of the finer varieties; found in Nordre Hamarnesgrotte the previous year, were of compound nature, with several fragments of the concretions cemented together. On one, a hemispherical nodule, of the type described below, was also attached (fig. 13).

[ fig 11-13 here]

Other concretions were found in situ on the clean scalloped limestone of periodically active passages, notably in Petty Cury. They occur like a row of mud splatters, some 1cm in diameter and protruding 0.5cm, sometimes coalescing into a streak. These again have a calcareous cement and are of two types similar to the larger concretions. The coarser type forms a blob flush with the wall, whilst the fine silt type has a stream-lined external shape and is banded parallel to the wall surface to which it is attached. In one dry inactive passage off the Aven series, these small nodules have a thin white matte coating of very fine grained calcite. Thin sections show that these concretions are unrelated to the crystal structure of the marmorised limestone to which they are attached. A similar, though larger (diameter - 2.5cm) hemispherical nodule was found in Hamarnesgrotten, which was of the coarser variety and had a thicker calcite coating. These nodules are quite distinct from the "helictites" of Nordre Hamarnesgrotte stream passage, which show concretic growth stages of small calcite prisms, including only occasional mineral particles.

The larger concretionary bodies cannot have moved far from their original position of formation since they are too fragile to have survived transport with the gravel and pebbles amongst which they are found. They must therefore have formed within the cave. The general planar nature suggests that they developed in association with a surface - Rock/air, rock/water, rock/sediment or even sediment/air. The slightly undulose surface often present is similar to the scalloped limestone of the walls. The other side, with its lobed directional forms could develop on an inclined rock/air or rock/water surface, so resembling, on a more extensive scale, the small concretions which are found under such conditions. The finer grained type, being softer, would suffer more from corrosion by moving water and hence assume a streamlined shape. The banding would correspond to successive periods of deposition.

There is little indication of the process by which the fine sand and silt have been cemented to form concretions. There is no internal structure apart from the banding in the finer varieties, and there is no trace of residual carbon to suggest accreting organisms such as algae, although this may not have been fixed within the concretion or, it may not have been detected. However the sediment was fixed, it postdates the scalloping, though requiring passages completely filled with water, moving at a velocity sufficient to streamline the finer grained varieties.

Lewis Railton (12) mentions that similar concretions were found by Mr. Bergenson of the Norsk Polar Institute, in Hamarnesgrotten.

Another type of formation is present in the North series, south of the keyhole, though only one was seen. It is a small trumpet-shaped mud stalamite standing about 4 cm high from the limestone of the passage floor (fig. 14). Such a shape would result from a slow regular dripping of water onto mud, scattering the sediment radially on impact and then cementing it by deposition of calcite on loss of CO2 from the water. Such formations have been described and figured by Mallot and Shrock (7) in America.

[fig 14]

Actual calcite formations are rare in Pikhauggrottene. The only ones seen were small straw stalactites and normal solid stalactites up to 20cm long and attached to limestone near the limestone-schist interface in the Aven and Midnight Series. Some show a two stage development, a central straw and successive zones of radial calcite fibres, as in many British formations.

The sparsity of formations has been commented upon by Horn (5 p.17) and L. Railton (12). It could be due to one or more of the variables which govern the growth of formations, ie. -

The amount of water draining into the system from direct rainfall and from melting snow is no limiting factor. The limestone, initially of low temperature origin, has been subjected to regional metamorphism, involving high pressures. Its primary porosity and permeability are therefore small, and seepage is confined to jointing and fractures.

The solubility of CO2 in water increases with lowering of temperature and this is enhanced by the fact that a lot of the water is derived from snow patches. These tend to concentrate CO2, since it is liberated in the freezing of water, but is retained within the snow patch due to its relatively high density. Both this and the inverse relationship between temperature and the solublity of CO2 in water, have been shown experimentally by J.E.Williams (15). The effect of temperature on deposition of calcite has not been studied. Deposition is governed by the equilibrium of CO2 over a convex surface of its own solution in water, and it is doubtful whether the 5°C, observed in Fosshullet, would be prohibitive.

Marmorisation has produced a relatively large grain size in the limestone. The calcite grains are disorientated and equidimensional, and roughly 0.5 - 1.0cm across. This would result in a lesser degree of solubility, compared to the normal D.I fragmental limestone of Craven for example. In the latter it is the fine grained matrix which is preferentially attacked by solution, and larger calcite crystals, as in crinoid ossicles, remain uncorroded - and often project from the walls of phreatic passages. The marmorised limestone lacks the more readily soluble matrix.

The admittedly thin thickness of overlying limestone does not seem to be significant when many profusely decorated caverns in Britain, such as the Reindeer section of Stump Cross Caverns, have equally thin coverings. It may, though, be operative under different conditions, where the limestone is covered by an impervious schist overburden, as suggested by Lewis Railton for Hamarnesgrotten. This solution, however, does not arise at Pikhauggrottene.

The time factor is one that was selected by Horn (5) who thought that lack of formations was due to the relative youth of the cave systems, or rather the "youth" of the necessary inactive conditions. As discussed below, in the case of the Pikhauggrottene system, this may well have been a limitation.

In general, the caves in the area are poorly decorated although Railton mentions one cave at Fausk reputed to have well developed formations. This tendency is more marked than in British caves, and probably reflects the different lithology of the limestone (ie. less permeable, and larger grain size) and the relatively shorter period over which growth of formations has been possible.


No living fauna was noticed within the cave, although remains were common in certain places. Included in amongst the gravel and pebbles of the lower Aven series and Petty Cury, for example, numerous rodent skeletons were found. The genera present, kindly identified by Dr Joysey of the Zoology Department, Cambridge, were

and in the lowest part of Midnight Series, a disintegrated skeleton of either (rabbit) or juvenile Lepus (hare) was found.

The only living fauna seen on the surface were specimens of Lemmings.

During the study of the sands mounted on slides, no evidences of flora - pollen, and algae, or fauna - diatoms, were observed.

The fauna recorded is not known to have any significance in relation to the postglacial history of the area, though the present fauna is not known, and in its detailed history might show some useful variations.


a) Cave criteria.

Of those who have worked in the South Svartisen Caving Area, Marstrander (3), Oxaal (11), Horn (5), Railton (12), O.C.Wells (14) and Corbel (2) have commented on the possible methods by which the caves developed. None, however, make particular reference to Pikhauggrottene, as previously, the cave system was only of small known extent - (ie. Horn (1947) - 230m; C.U.C.C. (1957) - 1000m; C.U.C.C. (1958) - 2000m). Of these workers, Marstrander, Railton, and in certain cases Corbel, favour major development by the outflow from lakes impounded by glacial barrage. Oxaal, Horn, and in other cases Corbel, favour actual subglacial development.

The most characteristic features of Pikhauggrottene are the tendency towards development of "tube" like passages, and the scalloped surface of the limestone.

The "Trykkledning" (Type b, fig. 2) have an elliptical to circular cross-section and therefore possess a bilateral symmetry plane which usually coincided with an oblique bedding or joint plane. Erosion must have occured symmetrically, and it can therefore be inferred that the major development took place when the passages were completely water filled.

Perhaps because the walls are generally unornamented, have no mud covering, and are white where not stained a light brown, the scalloping is more striking and seems more regularly developed than in British caves. The size of the scallops varies considerably. In some parts of the Aven series they are particularly well developed with a breadth of up to 40cm and showing a marked assymmetry (fig. 6). In other passages, especially those lower levels showing signs of recent vadose drainage, the scallops are less regularly distributed and are sharper and smaller - about 1-2 cm in breadth. An interesting variation seen in some passages is a decrease in the size of the scallop from ceiling (20cm) to floor (5cm).

Scallops have been described and discussed by several workers. Marcel Lugeon (6) compared the scallops of waterworn limestone to the smaller scale scalloping on rock surfaces in arid regions; he thought that they were similarly formed by horizontal blasts of small particles, though in the fluid medium of water instead of air.

Maxson and Campbell (9 - 10) describe them in more detail under the American term "flutes". According to their observations in the "Grand Canyon" "flutes" are developed prominently on the containing walls of stream moving at relatively high velocities and carrying a considerable load of suspended matter. This, together with the fact that they were also present as granites and gneisses, indicates corrasion as well as corrosion as the dominent agents in their formation. In certain places however, thin schist blades project from an otherwise regularly scalloped surface which emphasises the role of corrosion. They conclude that the "flutes" are formed by water in turbulent, as distinct from laminar flow, and that each concavity corresponds to a vortex in the water adjacent to the limestone face (fig. 15). On this explanation, the transition from an irregular pattern of small scallops to a more uniform pattern of larger regular scallops can be interpreted as one from a condition of inconstant turbulence, to one in which the turbulence became more uniform with the velocity flow and pressure gradient maintained constant over long periods. In the latter case large vortices, with breadth of the order of 40 cm could be stable under the corresponding range of velocity flow and pressure conditions. Once established, a scallop would tend to govern the conditions of turbulence and distribution of vortices, thus increasing its stability range.

[fig 15]

The decreasing scallop size, from ceiling to floor, would therefore imply less regular conditions of turbulence. This could be due to the heavier grade of the streams load saltating or rolling along the passage floor so preventing large vortices being established.

Theoretically, from hydrodynamics, the minimum velocity for turbulence can be calculated for the specific pressure gradient, but this requires the evaluation of Reynold's Number for the particular passage section, which would be difficult to do. Maxson, however, reckons that this minimum velocity would be of the order of 1 - 1.5 m/sec. It is therefore likely that actual velocities exceeded this.

Bretz (1) includes "fluting" amongst his criteria for vadose development, but states that "flutes cannot be used to prove vadose conditions existed at the time of their formation. Their value is in the record of a definite current and direction of flow". In the case of Pikhauggrottene the scallops consistently present over all the passage surface, again indicates complete water filling and a steady current, at the time of their formation. Interpretation of the direction of flow is simple in cases where the asymmetry is marked, but where it has been obscured by later subaerial erosion or corrosion interpretation tends to become subjective. This was shown in the entrance to Pikhaugrotte Nr. 2 where different directions were deduced from the scalloping independently by two parties. However, in every definite case; the general movement of water was seen to be down the outcrop from N.E. to S.W. involving occasional instances of uphill flow.

The general conclusions to be inferred from the passages and the scalloping are that, at the time of their development, the passages were completely filled with fast moving water, under hydrostatic pressure, and that these conditions were maintained uniform over a long period of time. These cannot be termed true vadose conditions, as there was no free air surface, nor can they be called phreatic since their position relative to the regional water table cannot be proved. The tern pseudo, or paraphreatic, as suggested by Professor Tratman, is therefore used to cover these conditions.

The abrasive material necessary for the corrasion, is now represented by the deposits of fine light sand. They were deposited during a phase in which the energy of the water became less than that required to carry its load, probably due to decrease in pressure gradient and flow of water. A succeeding stage of increased activity is shown in some passages in Petty Cury, where a channel has been cut into a bank of semilithified sediment (fig. 8).

Superimposed upon the older paraphreatic system are the local effects of vadose modification. Small water courses have developed connecting the surface to the lowest levels, generally down the North West limestone schist interface. At the lowest level, light, clean-washed passages are formed, the walls being finely scalloped and white. These show several criteria listed by Bretz for vadose formation - vadose trenches incised into the floors of passages of types (a) and (b), small potholes etc. The active drainage is inaccessible in a tight network underlying the main series and conforming to the conditions required by vadose flow - under gravity, as compared to hydrostatic pressure. The lowest accessible levels, such as Petty Cury and the mud chokes, show signs of periodic activity. The latter have a thin coating of moist dark mud up to a certain level on their walls. This activity would occur during periods of heavy rainfall and thaw; and would account for the recent gravel deposits and animal remains. On the whole, however, the present drainage is on a very small scale and incongruous with the extensive system it has adopted.

(b) Surface criteria.

The relation of the cave system to the surface topography also has an important bearing on its origin. As has been mentioned before, the top levels of the system appear to have been truncated, leaving in places isolated remnants and false entrances. This indicates that erosion, most likely glacial, occurred after the main development of the system, ie. that the latter preceded the last phase of glaciation in Pikhaugene. This conclusion was also reached by Horn (5 p.35).

This last glacial phase also left stranded blocks of granite porphyry. These are now perched delicately upon 10 - 15cm pedestals of limestone which must have been protected from the subaerial erosion which lowered the general level of the surrounding limestone. Similar examples are known from elsewhere in the Svartisen district, and also in Britain, as at Austwick. If the rate at which the limestone surface has been lowered could be determined independently, it would be possible to calculate the date of the last glacial phase in Pikhaugene. Corbel (2), by correlation with the readvance of Alpine glaciers assumes that the last glacial readvance was between the 14th and 16th century. This would involve an average rate of lowering of 0.15mm/year, which he considers reasonable under very humid periglacial conditions. Gunner Horn (5) by contrast, thinks that the rate of lowering was far less, of the order of 0.02mm/year. This would correspond to glaciation during the late glacial periods some 6-8,000 years ago. As recent detailed work on the weathering of limestone seems to be lacking, the date of the last glacial phase in Pikhaugene cannot be deduced from these limestone pedestals to the required degree of accuracy. The possible range of conditions which have elapsed since the limestone was glaciated are also too varied, in the local detail considered here, to be accounted for. Also, correlation with Alpine readvances, occurring within the Ice cap distinctly separate from the Scandinavian ice cap seems unjustified, especially when applied to such local glacial movement.

At the present day, the Flatisen glacier, 4.5km N. of Pikhauggrottene and some 150m lower, is receding at a considerable rate. To reach Pikhaugene, this glacier would have to have been about 100m higher in the main Glaamaaga valley, when it would also have coalesced with a small hanging glacier, at present perched up above the head of Pikhaugene. The joint glacier, or the deflected smaller glacier would then have overflowed into the niche on the west shoulder of Kamplitind, that forms Pikhaugene. Perhaps this did occur only 500 years ago, as Corbel suggests, but 6000 years would be a more probable date. The main development of the cave system therefore most likely preceded the end of the Pleistocene.

(c) Conditions and Age of formation.

So far formation can be said to have occurred over a period or periods during or preceding the Pleistocene. During this period or periods, the passages were completely filled with water moving at moderately high velocities, these conditions being maintained uniform. Two possible situations are suggested under which such conditions could be obtained; either the cave system drained a preglacial lake, or else it was actually a part of the subglacial drainage of a glacier.

The first case would occur if the outflow from a glacier exceeded the rate at which the system could discharge the water. Water would then accumulate on the surface behind some topographical feature to form a proglacial lake. Equilibrium would be achieved as the increasing pressure head increased the rate of flow through the cave. This would provide the necessary "paraphreatic" conditions (fig. 16).

[fig. 16]

The deposits, under such conditions would show a rough annual variation since the subsurface drainage would disturb the quiet settling conditions necessary for the deposition of varves, and the bulk of water within the cave would be too small to allow such extreme differentiation of grades. The banks of sediment in Petty Cury, near the mud chokes, do show a rough repeated graded bedding.

Such a lake, blocked at the N.E. end by a glacier, would contract to Pikhaugvand on the withdrawal of the glacier from the valley. Proglacial drainage would be temporary only, governed by the movement of the glacier; though could be extended over long periods of time.

The second case, subglacial drainage, has been shown to be possible by the researches of Sverdrup and Werenskiold (13) in Spitzbergen. They found the ground beneath the glacier they studied to be unfrozen. Again, assuming the supply of subglacial water exceeded the rate of normal discharge under gravity through the cave system, a pressure gradient would be established under the load of the glacier, until equilibrium conditions of discharge equalling supply were achieved. A relatively high velocity would therefore be maintained whilst a valley glacier occupied the whole of Pikhauggene (fig. 17).

[fig. 17]

The present long profile of the Pikhaugene limestone band shows that Pikhauggrottene lie within the highest part, that is above the lakes to the N.E. and S.W. This explains the present inactive state of the system which only takes the immediate drainage from the valley sides. Glacial erosion must have modified this profile to a certain extent, but it is difficult to say exactly in what way, even qualitatively - except that over Pikhauggrottene the surface was lowered sufficiently to intersect the top passages. Presuming, for the sake of argument, that the profile had the same general shape as today, both the conditions suggested can be represented as figs 16 and 17.

It is probable that both these situations were established several times during the glacial fluctuations in the ice age, and that these phases account for the main development of the system. Interglacial periods would possibly see conditions comparable to those of today, prevailing, with inactivity or slight vadose modification of the paraphreatic system in existence. The results of such periods, trenches, sediments, etc., would be obliterated by the following glacial period - and no evidence in Pikhauggrottene of interglacial periods is seen. It is unlikely that an interstadial period would see the deglaciation of the valley, and more probably the proglacial phase of development might occur at this time. Horn (5) suggested generally for the caves of South Svartisen that an incipient small scale system was probably already in existence in the early Tertiary Period. This would correspond to the initial the phreatic stage of the two cycle theory of formation, envisaged by Davis (3).

Both Horn (5) and Holmson (4) discuss the regional uplift and silting that has affected North Norway and whose magnitude can be inferred from the river terraces, and raised beaches along the coast line. Pikhaugene, however, lies above the extrapolated sea level, and hence its history is not complicated by this factor.

(d) Conclusions:

Pikhauggrottene developed in the following stages:


The Fosshullet-Glomvand system carries the drainage from the un-named lake in Pikhaugene underground to the Glomvand shore. The stream draining the south end of the lake continues down valley a short way and then swings round into the limestone band. Here it disappears underground, dropping about 10m to form an impressive entrance waterfall, as it does so. The stream can be followed underground in Fosshullet for a short way, exploration being halted, on the upstream side of a sump of unknown extent. It is then not seen again for about 2.5 - 3km, when it is met in Storbekkengrotte, 300m lower down in the section of the limestone outcrop overlooking Glomvand. Storbekkengrotte is essentially an effluence cave although the actual effluence is impenetrable and entrance is made through a karst window upstream. The stream flows out from it, across a levelled stretch of schist to disappear into another band of limestone and resurge roughly 100m beyond as a powerful spring on the shore of Glomvand.

As already stated, the Fosshullet sink was dyed with fluorescein and the resurgence at Glomvand turned green ten hours later. This indicates a velocity of the water averaging approximately 1m/sec. This is a relatively high value for limestone drainage when compared to British systems which have been proved by fluorescein - such as Mossdale caverns, where the slow rate of movement had been accounted for by large stretches of phreatic passages. This condition does not appear to arise in the Glomvand-Fosshullet system which is seen to be vadose at both ends, and is presumably so in the intermediate section where the 300m drop in level is involved. It is unfortunate that no entrance to this section was found although the surface outcrop of limestone was briefly explored. The remaining local drainage in Pikhaugene remained superficial, and one small stream actually crossed the limestone band. On the shoulder of Glomvasstind, drainage concentrated into a small stream did sink into the limestone, but this system was very tight and, in effect, a direct tributary to Storbekkengrotte. Other small caves were either dry joint systems showing no indication of connecting with the main drainage, or, those that looked more promising silted up completely after a few metres, It is quite possible that an open way into system that was missed, does exist.


The character of Fosshullet is in direct contrast to Pikhauggrottene. Instead of being essentially a deserted dry system, Fosshullet is active and takes a fair sized stream from the lake North East of it. The drainage forms the core of the system which is therefore relatively simple when compared to the network of passage perforating the limestone band in Pikhauggrottene. Exploration in Fosshullet was limited, but altogether, about 45 hours were spent in it, surveying and photographing.


The passages can be related to some of the types recognised in Pikhauggrotten although development by the main stream has also resulted in large chambers. The waterfall chamber, nearly 20m high, is angular in outline and presumably involved a certain amount of "cavern collapse" in the later stages of its formation. This process is not obvious in caves in this area though, since bedding planes are not operative, and the jointing is widely spaced. The places where passages intersect one of the schist containing walls shows the effects better, since the schist is mechanically weak and collapses easily, large slabs flaking off, especially from the hanging schist wall. Otherwise the passages can be related to the "bedding plane" (type a) or Trykkledning (type b).

The "bedding plane" type of passages are governed usually by one or two sets of oblique joint planes. Their control of the formation of the cave is marked and they can also be observed on the surface of the limestone where they are widely spaced and regularly developed (fig. 18).

[fig. 18]

This sometimes gives a passage of similar cross-section to fig. 1 although development along the dip of the joint plane is often more extensive as in the section connecting the main stream passage to "King's Parade" (fig. 19) and also in the subway series.

The "Trykkledning" (pressure tubes) are perhaps even better developed than in Pikhauggrottene, "King's Parade" being the true example. This has a well developed elliptical cross-section, modified by the bedding as shown in fig. 2. This type of cross-section is also present in the main stream passage and the smaller tubes of the subway series. Another possible example is that of the dry North East series which has an arcuate roof but is almost completely silted up (fig. 20). The inclined rift at the end of this series is reminiscent of the type 2 of Pikhauggrottene (the steep angle rift - fig. 4).

[figs 19-21]

With the exception of the North East series, and the higher levels, most of the passages show signs of periodic activity. The walls are generally white, clean washed, and show small scale scalloping of the order of 2 - 8cm across. One passage in the subway series showed a diminution in scallop size from ceiling (5cm) to floor (2cm), as has been described in Pikhauggrottene. An interesting feature seen in the connection between the subway series and the main stream passage (shown dotted on the survey) was what appeared to be a phreatic network with half tubes and some poor sponge-work. This was the only place in which this was seen, though, and elsewhere vadose phenomena - trenches, small potholes etc. were the obvious features of the active passages.

The North East series is of a different nature. The passages are dry with a fine sandy floor, and have brown stained walls which show large scalloping ( - 20cm across). The series is similar to the Pikhauggrottene type of passage, and formed under similar conditions.


The sediments of the North East series resemble those of Pikhauggrottene. The light fine sands have the same heavy mineral assemblage as those of the Aven series. The deposits of the rest of Fosshullet are coarser in grade. A coarse dark sand is associated with the schist exposures, deriving from the disintegration of the schist. Pounded pebbles and stones (5 - 25cm) are common in the active passages - for example the main stream passage, the subway series etc. Within the latter a bank of the gravel has been semi-lithified, and subsequently eroded to form a scarp (fig. 21).

Living fauna were recorded in Fosshullet in the form of insects. Brown and black beetles were found and also a type of mosquito, all as yet unidentified. In the dry sands of the North East series, a rodent's skull was found, the genus being Lemmus.

No formations of any type were seen in Fosshullet, the walls being completely bare of stalactites and concretions.


On the surface, there were indications of a superficial drainage channel crossing the limestone just south of the "whale back" within which the Fosshullet entrance lies. This may still be periodically active under flood conditions, but probably predates the Fosshullet system. A few small streams also sink into the limestone above Fosshullet and enter the system by small inlets down the schist limestone interface.

The main stream in Fosshullet, after descending about 25m, starts off up the outcrop towards the N.E. It is then deflected back through 180° and is lost at the unpassed sump in the main stream passage. From here it cannot be followed directly, but is seen again in various minor sumps in the subway series before disappearing at the final sump at the end of "King's Parade". The survey showed this series of sumps to lie on a distinct gradient - dropping about 1m in 75m. Fluorescein introduced in the main stream passage, appeared in most of the minor sumps in the subway series, and in the final sump after 35 minutes, so proving them all to be interconnected. Attempts were made to investigate the final sump and main stream sump by floating in and kicking, though no attempts were made to dive them. The passages appear to continue with some dimensions at some depth below the water, and would be ideal for cave diving, but ended normal exploration of the system.

The problem of the cause of these perched sumps is interesting. Such high level sumps are often met with in cave systems, as at Fountains Fell in Yorkshire, and they probably result from a variety of causes, although these are rather obscure. Several mechanisms have been considered, which could account for the sumps in Fosshullet. Tectonics resulting in a silicified fault zone, or a simple extensive quartz vein could give a perched water table. There is, however, no surface evidence of any tectonic activity, and although quartz lenticles do occur along the original bedding or in the oblique joints, the largest seen was 3m x 15cm. This would not be sufficient to dam back a passage of the dimensions of King's Parade.

Another possibility is that the fluctuations in width of the limestone band, observed in the surface exposure, are accentuated at depth. If the limestone pinched out and the position of the feather edge varied vertically, a perched water table could easily arise, as shown in fig. 22. There is, however, no evidence of such extreme variation in thickness, and no schist is intersected in the subway series which is quite extensive across the width of the limestone. This explanation is therefore unlikely in the case of Fosshullet.

[figs 22, 23]

A more likely possibility is based on observations on passages where they come against the "hanging" schist wall. Here, passages are often seen to move diagonally up the schist face for a short distance, before swinging back down into the limestone band again. This could cause a siphon to form locally in the passage concerned, which may well be the case in King's Parade. It is represented diagrammatically in fig. 23.


From the nature of the cave system, it appears the main development of the large chamber and related series had been by the outflow from the, un-named lake, entering at the same point as it does today. The dry North East series, however, appears to antedate this stage and it resembles the passages in Pikhauggrottene closely. It is therefore likely that a system formed over the same periods and under the same conditions: as Pikhauggrottene to the N.E. and was later intercepted by the drainage from the lake and subsequently adapted and modified to its present condition. The date at which this modification became dominant would be when the earlier system had become inactive - probably in the late glacial period, although earlier interglacial phases are possible. Modification would consist of the formation of passages leading down to "King's Parade" and the enlargement of the latter. The conditions would be nearer to vadose than those of the preceding paraphreatic phase, as then if the system did become locally completely water-filled the hydrostatic pressure built up under gravity flow would be less than that under subglacial or proglacial conditions. Any earlier large scale scalloping such as has been preserved in the North East series, would therefore have been replaced by the smaller "King's Parade". From the ubiquity of this scalloping, it can be inferred that the modification occurred when far more water was entering the system than today. This would have occurred when the level of the lake was much higher, being supplied directly with local glacial drainage in the early post-glacial times.

ie. probable stages of development:



Storbekkengrotten unfortunately, was incompletely explored and only partially surveyed (fig. 24). Its character is different from both Pikhauggrottene and Fosshullet, although certain features are held in common. This in places it shows a tendency towards passages of the types (a) and (b) whilst it is also essentially an active stream system, clean washed with small scalloping on its walls.

Two features, however, are particularly striking, the marked vertical development and the remarkable phreatic forms present. Storbekkengrotten consist mainly of a large rift passage running N. 28° E. To the N. the rift narrows down to a deep canal carrying the fast moving river until a sump is reached. To the South, the rift opens out to form a large chamber some 10m high. Beyond this chamber, the stream branches off to the right into a passage which has an opening at the top, narrowing down to an inclined rift containing the stream below (fig. 25). A dry series continues on from the large rift chamber, still marked by a large vertical extent, although the cross section is different (fig. 26).

[fig 26]

Tributary to the main passages at various levels are small meander-passages, of cross section shown in fig. 27. They show signs of phreatic development - spongework, half tubes, etc. The spongework, which is also present all over the upper parts of the rift walls is extremely well developed (fig. 27). Large complex pendants protrude up to 50cm, irregularly though smoothly shaped and unrelated to the banded limestone structure which can be seen in them. They are inactive, and stained brown.

In contrast to these phreatic forms are the vadose features formed by stream erosion, which also very well developed. In the lower stream section remarkable potholes 75cm in diameter go straight down some two metres. The stream has cut neat trenches into the limestone which it occupies to form powerful water chutes. Large swirly embayments have also been cut into the limestone where the passage turns through 90°.


Storbekkengrotte shows very clearly two distinct phases of formation. In the earlier one, the passage networks formed with water moving very slowly under hydrostatic head. Slow solution of the limestone occurred by almost static water to form the sponge-work - (there is no obvious directional element in them, except the vertical). Conditions were truly phreatic and this part of the system must then have been below the regional water-table. The subsequent lowering of the water-table resulted in the system being adapted for gravity flow, with concentration of the drainage and development of one series of passages.

The movement of the water-table is interesting. Attention has been paid by Norwegian geologists working in this area (Horn (5) - Holmson (4)) to the regional uplift and tilting of Scandinavia that followed the removal of the ice cap. From various raised beaches, river terraces etc., the position of the sea level on the depressed land mass can be reconstructed. Extrapolating this work into the Glomvand area, (unpublished work by P.J.Kirkland), the original sea level was found to be just higher than Storbekkengrotte. It is therefore likely that the early phreatic phase was one of submarine development and that the succeeding vadose phase is therefore comparatively recent, and certainly post-glacial.

If correlation with the other end of the system is justifiable, the early phreatic phase in Storbekkengrotte would correspond to the paraphreatic phase in Fosshullet and hence to the main development of Pikhauggrottene. This would place the latter as late Pleistocene or earlier. This framework of correlation is, however, rather tenuous and would require further work to justify it. It is unfortunate that so little time was left to investigate Storbekkengrotte, particularly as a study of sediments belonging to the earlier phase may well have helped to establish that marine/brackish water conditions did exist in the system at that time. Also a comparison with the Pikhauggrottene heavy mineral assemblages might possibly have helped with correlation of the various stages of development.

B I B L I O G R A P H Y.

(1) Bretz J.H. 1952 -
Vadose and phreatic features of Limestone Caverns. Journ. of Geology. L pp 675-811.
(2) Corbel. 1957 -
Le Karst du Nord Ouest de L'Europe et de quelques regions de comparaison.
(3) Davis, W.M. 1930 -
Origin of Limestone Caves. Bull. of the Geol. Soc. of America. Vol.41. pp 475-628
(4) Holmsen G. 1932 -
Rana. Beskrivelse til det Geologiske Generalkart - Norges Geologiske Undersokelse. Nr. 136. Oslo.
(5) Horn G. 1947 -
Karsthuler i Nordland - Norges Geologiske Undersokelse Nr 165. Oslo.
(6) Lugeon M. 1914 -
Annales de Geographic. Vol.23-24.
(7) Malott C.A. and Schrock R.R. 1933 -
Mud stalagmites - American Journ. Science 25 pp. 55-60.
(8) Marstrander R. 1911. -
Svartisen, dens Geolgi - Norges Geologiske Undersikelse Nr. 59. pp 14-17.
(9) Maxson and Campbell. 1935 -
Stream fluting and stream erosion. Journ. of Geology. Vol. XLIII. pp 729-744.
(10) Maxson. 1940 -
Fluting and facetting of rock fragments. Journal of Geology. Vol. XLVIII. p 717.
(11) Oxaal J. 1914 -
Kalkstenshuler i Ranen - Norges Geologiske Undersokelse. Nr 69. Oslo.
(12) Railton L. 1954 -
Caving in Norway. Trans. of the Cave Research Group of G.B. Vol.3 No.1 p 17.
(13) Sverdrup and Werenskjold. 1922 -
Frozen earth in Spitzbergene Geofysiske publikasjoner. Vol.11 No.10.
(14) Wells, O.C. 1957 -
Some notes on Larshullet and Lapphullet Caves, Northern Norway. Trans. Cave Research Group, G.B. Vol.5 No.1 pp 23-35

The area is covered by the map:- Topografisk Kart over Norges - Svartisen Gradteig J.15. (Scale 1 : 100,000).


We would like to express our gratitude to the many people whose help made the Expedition possible, in particular, Rector Arne Grønlie for looking after us in Mo-i-Rana, and our friends Bjørm Granlund, Bernhardt Ravna and Karl Westermark, whose help and company we appreciated on many oocasions in Svartisen.

A further list of our benefactors follows:-

> Index to published Norway expo articles
> Back to CUCC top page