Cave Research Group of G.B. Newsletter
No. 87, Mar. 1963, pp 1-5


The summer of 1962 saw a party of Cambridge cavers in the Pindus Mountains of north-west Greece. The group was first attracted to the area by the accounts of limestone massifs and gorges given by a recent botanical expedition to the region.

The Polje and Caves near Ioannina:

We chose the town of Ioannina as the centre for our investigations. This town, the economic and administrative hub of Epirus, is as notable for its picturesque Turkish influence as it is for its spectacular setting on the shores of a lake beneath the foothills of the Pindus Mountains. The lake, which is about 1600 feet above sea-level, lies in a long tectonic trough of limestone which has subsequently been modified by solution, and it is drained by "katavothrai", or subsurface outlets. This trench, or polje, forms a convenient boundary between the predominantly impermeable rocks of the coastal range, and the Pindus Mountains, formed of a succession of Cretaceous and Jurassic limestones, to the north and east. The Pindus form the main watershed between the rivers flowing to the Adriatic and those flowing to the Aegean. They are highly dissected on the western edge and they have been subjected to considerable faulting. Rising to an average height of 6000 feet, but occasionally exceeding 8000 feet they are barren and have maquis as the dominant vegetation covering. The mountains support a sparse population, consisting mainly of Vlachs - nomadic peoples similar to the Lapps who live off herds of goats and sheep. Many of the villages scattered over the region are built around old religious institutions, but past glories are largely forgotten in the face of the difficult environment and the toll extracted by emigration. Roads in the area are generally bad and the mule or donkey is the main transport medium. The motorbus is however bringing some measure of unity to what was originally an area as dissected socially as it is physically.

To the north-west of the Ioannina lake, and in the centre of the polje is a small hum (isolated erosional remnant) and in this hillock is the Perama show cave. The hill is covered with weathered limestone scree and the entrance to the cave was not discovered until 1940. The cavern is a very ancient landscape feature - there is no stream in it, and its main interest lies in the magnificent limestone formations which are gradually filling it up. The route taken through the cave is a purely arbitrary one, weaving in and out of the stalactite formations. I feel sure that the cave could be extended, but this would involve breaking down formations, and is unlikely to yield more than extra shapes with which the guides could play their word-games.

A similar hillock exists today as a small island in the middle of a lake. On investigation, this was found to be pitted with small collapsed sink-holes. A resurgence cave with a small stream was discovered down near the lake. This was explored upstream for about 300 feet. The passage rose very steeply and the end, which was blocked with boulders, was close to the surface. The cliffs around the island were interesting, being covered with an impressive array of calcite curtains and stalactites, some of them quite large.

Limestone Features in the Mitsikoli Range:

To the north-east of the Ioannina trough rises the Mitsikeli Range. This foothill range of the Pindus is a limestone area consisting of an irregular succession of variously-sized solutional depressions and rounded hills. The range is an excellent example of a maturely dissected landform assemblage, and as such it is totally lacking in accessible caves and potholes. The whole range is thickly mantled in scree. The hills have a rounded form and all cave entrances are masked. The transportation of sediment towards the lake (presumably in a former wetter phase) has resulted in the formation of large alluvial fans which form the main village and cultivation sites in the area. The cone-shaped hills are an interesting feature since it has been suggested that a tropical climate may have obtained in Greece at one time. The cone-shaped hill is one of the characteristic features of limestone scenery in the tropics today, notably in Cuba and Malaya. Deposits suggestive of laterites were noted, and deep rock-rotting, sometimes to a depth of 10 or 12 feet could frequently be seen in road-cuttings. Clints on the higher parts of the foothills have been weathered to such an extent that they were leaning crazily against one another and could be moved around with the foot. Obviously without much more investigation the hypothesis of tropical weathering is no more than tentative, but writers from other parts of the Greek mainland and from Yugoslavia have also seen suggestions of it.

Moraines in the Pindus Mountains lead inevitably to considerations of glaciation and periglaciation, the effects of which are found all over the region. For the time being it is best to think of the deep rock-rotting and scree formation as being due to a much higher rainfall and more intense frost action under periglacial conditions rather than to tropical climate weathering. Similarly, the vegetation cover was much thicker in historical times and would have provided the necessary mechanism for the greater acidulation of the precipitation and the speeding up of all solutional processes. The main result of these processes is that a thick layer of scree mantles the hills and blocks many of the joints, thus precluding cave development.

Konitsa and the Vikos River Gorges:

Having seen the features of the Mitsikeli Range - a complete lack of drainage and a mature landscape rotted to a considerable depth, we decided to investigate the country northward, towards Konitsa and the Albanian border. At first the road runs along the Ioannina trench and then strikes off into the hills and poljes of the Pindus foothills. At the summit of this range it descends in a series of Z-bends to the course of the Vikos river, where it debouches from a gorge on to a wide flood-plain. This is now a military zone, and passes have to be Obtained from the army authorities in Ioannina. These are however easy to obtain. Our first camp was at the end of the gorge, by the side of a dam which has been built to provide irrigation facilities for the region. The gorge itself is very spectacular. The massive limestone walls are some 800 to 900 feet high in this part. Higher up, they attain heights of anything up to 1500 to 2000 feet.

A preliminary investigation of the gorge seemed quite promising. The limestone was solid to the touch and obvious-looking cave entrances could be seen in the sides. When we came to study these entrances more closely however, we found they never went in more than about ten feet. They are scour hollows formed as the river cut down through successive stages. They are found at widely differing heights, both near the top of the gorge and down at river-level, where they may be seen in the process of formation today. Caves such as are found in the Tarn Gorge do not exist - the water merely percolates down through the fissures in the limestone and appears in the form of tiny springs at river level. It is probable that the down-cutting of the main river has been so rapid that the water in the interstices of the limestone is continually adapting itself to a changing base-level. This precludes all possibility of great solutional enlargement of the joints. The plateau surfaces above the gorge scarcely warrant such a name since they are a mass of dissected pinnacles and spreads of felsenmeer. There is no indication of water-courses, and no possibility of water collecting in to one central funnel, as there is for instance, on the French Causses. The great gullies near the cliff-tops do however give the impression of swallow-holes sawn in half and viewed in section. At some time in the past it appears that streams have collected, and disappeared down deep shafts, only to be abondoned by the lowering of the base-level, and then almost completely weathered away. Here, as in other parts of the Greek mainland, ancient karst forms lie almost alongside extremely immature forms.

Provatina and other potholes on the Astraka Plateau:

A short investigation north of Konitsa showed that the rock here was of flysch rather than limestone, so we left this area and went further up the Vikos Gorge, to the village of Papingen. This village is situated on a small plateau above the gorge, and behind it the limestone massif of Astraka rises to a height of nearly 8000 feet. It was on the summit plateau-surface of this impressive mountain that we found "Provatina". The English-speaking Greek in the village told us about the hole, which he had seen whilst hunting wild goats, and so, on the following day we set out to look for it. The gradient on the lower slopes of the mountain is not steep, and here again, scree is widespread. Once above the tree-line we soon reached the base of the cliffs, and the only way to get to the summit was to climb one of the gullies which score the mountain's north face. From the top (after about an hour's climb), the hole was seen on a grassy plateau to the left of the dried-up valley. It lay at the base of a small cliff and had no water-course leading into it. In surface dimensions it was similar to Alum Pot, but the depth is far greater - we reckoned the total depth of the first pitch to be of the order of 600 to 800 feet. A 250 foot rope lowered with a rock on the end failed to touch a ledge of any sort, and a stone thrown down could be heard rattling for thirty seconds - there being a second pitch as deep as the first one! The vertical sides of the hole were well fluted, though it is difficult to say whether water action has been a modifying or a formation agent in this case. A thin capping of flysch does exist on the top of Astraka, and this may indeed have provided the necessary conditions for vertical development (as on Ingleborough). However, against this must be set the fact that Provatina is the only deep hole in the area, and it would seem more reasonable to invoke faulting as a formation mechanism, as in the case of the Yorkshire pot-holes of Long Kin West and Rift Pot. Other pots do exist on Astraka, but these are generally of the order of 100 feet deep, choked with rubble or snow at the bottom. The possibility of extending these does not seem to be very great. Provatina thus seems to be something of an anachronism. Certainly its position away from the present valley, and the fact that it has no water course leading into it are rather unpromising features, making it appear that faulting is the main mode of formation. If this is the case it is unlikely that the hole can be extended deeper than its estimated depth of about 1200 feet. On the other hand, the hole may be a very ancient feature, the old valley leading into it having been eroded away. Such a hypothesis does not fit in with the poor development of other karstic landforms in the area.

Several smaller caves in the sides of valleys tributary to the Vikos River were investigated. These were in a band of Reef limestone. They contained no streams, but were liberally decorated with formations. The passages were of the rift type and closely controlled by the jointing. Developmemt was almost certainly in accord with the changes in base-level: after this dropped the small cave passages were left dry in the valley sides.

The Karst of the Monodhendri Region:

From Papingon our final expedition was further still up the Vikos River to the village of Monodhendri. Again this village was perched high above the gorge and received its water supply from perched water tables on flysch or flint within the limestone. The geological succession in this region consisted mainly of a great thickness of very thinly bedded limestones. These have been eroded into fantastic shapes where they stand out above the surrounding doline-country. The finest succession of dolines we encountered was in the Monodhendri region where they varied from a few yards to half a mile in width. Many of the depressions had pot-holes associated with them. These were frequently at the lowest point of the doline, and drainage must have gone down them in the past. The anomalous position of some holes in the district leads again to the probability of faulting as the dominant mechanism. All the pots are now dry and all are choked at the bottom with scree. We descended all the holes shown us by a local guide. The deepest was 300 feet and went down in one big pitch. The height of these pot-holes is about the same as the summit surface of Astraka, and the general accordance of the highest plateau in the Pindus of this area suggests a high-level erosion surface with rather better karstic development than we encountered at lower levels near Konitsa.


Apart from the finding of Provatina, which we could not attempt with the tackle we had at our disposal, our expedition was rather disappointing. We had expected a terrain similar to that of Yugoslavia, but did not find it. Surface drainage of any type is almost lacking in the Pindus Mountains and a thick layer of scree seals off most joints in the limestone, preventing any concentrated percolation. These rock accumulations are probably the result of very intense frost action in periglacial times, on the edge of small local ice caps. Although overworked, the term 'maturely dissected' ideally describes the rounded hills and ancient landform assemblages to be found in the Pindus. Only on the highest erosion surface did we find any significant vertical development. I would advise any future expedition to the area to concentrate on the summit plateau at about 7000 feet.

Cambridge, October 31st, 1962.
D.C. Mercer.

> A slightly fuller version of this article was published in CUCC's own Journal in 1965
> Index to Cambridge Underground
> Back to CUCC top page
> You can read an account of an early (1973) SRT descent of Provetina on the NPC Website.