logoAA Conference logo
foto dr Berislav Horvatić, fotodr Borut Juvanec

Water-permeable (!) Corbelling - the Shepherds’ Cisterns (bunari) of the Village of Draga Bašćanska on the Island of Krk in the Northern Adriatic

For at least six thousand years corbelling has been the ingenious way of constructing a top covering of a dry stone building that is in dry stone itself, i.e., using only stones and no binding material, no mortar or concrete, as well as no truss to support it. The problem was brilliantly solved in the late Stone Age or early Bronze Age, yielding the so-called false dome. If the stones are laid properly, the dome comes out rainproof, as it should be.

A corbelled dry stone roof deliberately constructed to be water-permeable does not make much sense, unless it covers a cistern for collecting and storing rainwater.

A fine and rare example of the latter unusual building technique are the shepherds’ cisterns of the village of Draga Bašćanska (locally: Baška Draga or just Draga) on the island of Krk in the Northern Adriatic. A dozen or so of them are scattered over the unenclosed, communal grazing ground (= common grazings = rough pasture, komunada) on the barren northeastern plateau (Gornji/Gorinji Vrh, ‘The Upper Plateau’) of the southern part of the island. They have been built, maintained and used by the local shepherds who graze their sheep in the corresponding ‘patches’ (pajiz) of the common grazings during the summer half of the year. The names of the localities are the following: (1) Za Vodinami, (2) Za Kijcen, (3) Na Ohodu, (4) Semjuni, (5) Pod Maričini, (6) Pod Franićevi, (7) Pod Ostri, (8) Na Trnovici, (9) Navrh Vala, (10) Podprodanji, (11) Pod Zminu, (12) Na Guvninah, (13) Va Graci.

The cisterns are named bunar (pl. bunari) by the villagers of Draga Bašćanska, while the term zdenac (pl. zdenci) is in much wider use by other inhabitants of the island of Krk. Both dialectal terms literally mean ‘well’, but locally they refer to cisterns rather than wells in the usual sense of the word. Also, both denote only the cisterns in the field, while another word (gušterna) has been reserved for a classical, fully closed and mortared cistern one has at home as the source of water for the household.

The cisterns built by the shepherds of Draga Bašćanska are exceptional for having an above-ground dry stone shelter with a water-permeable top covering, usually corbelled or partly corbelled, built above the underground storage tank. The shelter serves several purposes: (i) The outside surface of the covering, flattened with packed small stone (gruh), is the (only) rainwater catchment area. (ii) Unlike an ordinary house roof, the dry stone covering of the shelter has no gutter to collect and carry away rainwater, as well as no downspout to convey roof runoff to the storage tank. Both of these functions are fulfilled by the permeable false dome which allows the harvested rainwater to permeat through the interstices among the stones directly down into the reservoir. (iii) The shelter protects drinking water from pollution. (iv) Last but not least, it ensures natural cooling of water due to constant streaming of air through the dry walling, driven by the difference in temperature between the sunlit side of the building and the shady one. It is the ancient solar refrigerator, with the cooling proportional to the intensity of sunshine, being most efficient on a hot day with a baking sun - just when most needed! A thirsty person who comes across a cistern like that will certainly appreciate the refreshing coolness of both the shelter and the drinking water inside.

Since the cistern water is only for people, the building usually has a narrow entrance impassable by sheep, a sqeeze stile or through stile (škalica), consisting of two wall heads built close together, forming a narrow vertical slit to edge one’s legs through. An adult sheep is too corpulent to squeeze through, but a question naturally arises whether a little lamb could make it. Presumably it could, but shepherds maintain that it would not venture through without its mother, leaving her out of sight on the other side. Until proven or disproven experimentally, this thesis should be taken at its face value.

Preventing sheep from entering a cistern is not just a minor matter of hygiene. The problem is that sheep are “unable” to walk backwards - in order to move back, they have to turn round first. So a sheep that somehow manages to enter a cistern and gets stuck in the narrow inner space simply does not know how to get out since it has no space to turn round. It will sooner drown and contaminate the cistern than pull out going backwards. That is why the entrance has to be carefully constructed to be either 100% “sheepproof”, which is the prevailing solution, or, quite the opposite, wide enough for a sheep to get both in and out (e.g. Na Ohodu). May this be understood as a tiny contribution of ethology (not ethnology!) to architecture, aiming at a possible influence of the behavior patterns of animals on the structure of a building.

The squeeze stile is followed by a small uncovered vestibule, from which a stone staircase leads down and inside the shelter to the water. The staircase is often curved, i.e., descending sideways, since the entrance to the roofed part of the cistern is preferably made at an angle with the main axis of the squeeze stile and vestibule, which renders it better sheltered from the unfavourable outside influences (e.g., wind bringing litter, direct sunshine to warm the water.)

The dry stone shelter itself is usually oval-shaped in the ground plan, with perimeter walls almost vertical on the outside, and relatively low. The entrance of the shelter is constructed in the usual manner, with a massive stone slab as a lintel.

The whole building is in uncoursed, rough dry stone masonry (= random rubble) except for the underground storage tank (pit), which is partly hewed in stone and partly dug in the ground and constructed with half-dressed stones, plastered over with concrete (nowadays). In the “old days” when there was no concrete, cisterns were built only in places with impermeable, loamy soil. The pit was constructed in dry stone that was not plastered over, and it was the surrounding loamy soil that held the water to some extent. Cisterns like that are termed na zemlju, na blato, na teren (‘on earth, on mud, on terrain’), and are less reliable than those na beton (‘on concrete’), since their pits are not absolutely watertight. Of the extant larger ones, only the one Na Trnovici is still ‘on earth’.

Apart from these general, systematic features, each of the cisterns has a peculiarity of its own, reflecting the individuality and creativity of a particular builder or builders. E. g., the cistern Semjuni has two chambers, the one Za Kijcen has a kind of gutter on the rim of its roof, … Not two of them are completely alike.

Geomorphologically the southern part of the island comprises two barren and dry plateaus rising to some 500 metres above sea level, made up of Upper Cretaceous limestones, and a flysh valley between them, carved by a small stream with an imposing name of Vela Rika (‘Big River’). Of all the islands of the Adriatic Sea the island of Krk is the only one that has running water, and two of them at that: Veli potok (‘Big Brook’) in the central part and Vela Rika in the south. Unfortunately for the villagers of Draga Bašćanska they settled in the valley of Vela Rika “too late”, presumably in the 16th century, when most of the arable ground in the valley had already been taken by the earlier settlers. Therefore they were forced to take to the stony and dry plateaus, away from the village and running water, not only for their sheep rearing, but for their tillage as well. Most of their farming was restricted to ‘newtakes’, small areas of the communal grazing ground enclosed for private use as fields (with the permission of local authorities, occasionally even without one.) This unfortunate circumstances seem to have greatly fostered their diligence and creativity in dry stone architecture and earned them (deservedly!) the local fame of extraordinary builders on the island, as regards both quantity and quality. One gets mixed feelings while inspecting their exceptional accomplishments enforced by sheer necessity.

Although traditional small-scale farming on the plateaus was mostly abandoned soon after the 2nd World War, traditional sheep rearing remained and even expanded in quantity. The villagers of Draga Bašćanska alone graze some two thousand plus sheep, which is by no means exceptional either on the island of Krk or on the remaining four big Northern Adriatic islands of Cres, Lošinj, Rab and Pag.

The problem of drinking water on the dry plateaus is solved by harvesting rainwater. Water holes (lokva, pl. lokve) are for sheep, dry stone cisterns for people. Both are a necessity for traditional sheep rearing in dry, stony areas without running or underground water.

Water holes are small ponds formed naturally in shallow dips in the surface of the ground with impermeable, loamy soil, in which rainwater collects and remains for a longer time. They range from a few metres to some fifteen metres in diameter and up to a few metres in depth. What nature conceives is then supplemented and maintained by human hand. Water holes are deepened and cleaned from time to time by local shepherds.

As communal, public sources of drinking water, the shepherds’ cisterns of Draga Bašćanska are located in the open, on the unenclosed, communal grazing ground. The only exception is the one Na Ohodu, which is situated within a dry stone paddock (graja, pl. graje), but is also easily accessible through a gap in the dry stone wall bordering the paddock, closed only by an inserted bunch of thorny branches.

What about other, more standard corbelled dry stone buildings, like dry stone shelters (for people) or sheepcotes (for livestock), constructed by the same builders? There are almost none! The strong pressure of the local building tradition of the village has had it otherwise: covering with stone, corbelled or not, was for cisterns only, while all other dry stone buildings had to have thatched roofs. A typical dry stone shelter for people (hramac, pl. hramci) is rectangular in the ground plan, both inside and outside, and covered with a thatched roof (slon) supported on a wooden frame, mostly a gabled double-slope one. Sheepcotes (mošuna, pl. mošune) are of two types: either rectangular, being just the enlarged versions of shelters for people, or oval, with an oblong opening (zjalo) along the middle of the roof.

Some of the shelters and sheepcotes are real masterpieces of dry stone masonry, built in dressed or half-dressed stone with much care and skill, but gaping desperately roofless nowadays. A thatched roof requires regular maintenance, at least once in five years or so. As the vast majority of these buildings were abandoned a few decades ago, their thatched roofs have gone with the wind (literally!)

Not that builders from Draga Bašćanska were ignorant of the much more durable possibility. Their neighbours from Punat have built a multitude of corbelled dry stone shelters, and indeed, the only two ones built by the shepherds from Draga Bašćanska stand near the “border” with the common grazings of Punat. Also, they did use a specific kind of corbelling to cover their cisterns. The point is that the dictates of a local tradition often prove much stronger than any pragmatic reasons, turning sometimes into a tyranny of local tradition. The “refusal” to corbel a dry stone shelter is just a tiny example.

Back to the concept of water-permeable corbelling. Historically the problem was posed in just the opposite way: how to construct a rainproof covering in dry stone. Anyone could make a dry stone roof that leaks, but it takes a lot of creative effort to construct a rainproof one! It probably takes no less creativity to turn the problem the other way round and undo the established “normal” concept, constructing a corbelled dome that is as permeable as possible.

When I first entered one of these cisterns, and the finest one at that (Navrh Vala), I was surprised and rather disappointed by the “poor” job I encountered inside. The stones of the false dome seemed to had been laid rather “sloppily” and in the “wrong” way - decent corbelling just isn’t done like that! It took some time for me to realize, and get confirmation from local shepherds, that what seemed like “sloppiness” is indeed intentional - the builders did not want the corbelled dome to be rainproof, on the contrary, and they did their best.

Za povečavo kliknite na sliko
Click on picture to enlarge