The ‘Dragon houses’; Megalithic monuments in Euboea, Greece – What we know so far

There are 23 ancient megalithic structures, called the ‘Dragon Houses‘, on the mountains of the island of Euboea, Greece, mostly in the areas of Mount Ochi and of Styra.


They are constructed of relatively thin stones, mostly square or rectangular, placed on top of one another without any binders. Huge singular stones (monoliths) are mostly used. Another impressive thing is that they don’t have foundations. Their roofs are accurately constructed in pyramid-like fashion with gigantic plates placed one over the other.

These stone-made buildings seem to be placed in strategic places capable to observe all the territory around at quite a distance. Especially the Dragon House of Ochi (or Oche) is placed at an altitude of about 1,400m.

Let’s see now what we can further learn from the published work of specialists.


Abstract In southern Euboea, Central Greece, there are several megalithic buildings known as ‘drakospita’ (or dragon houses) whose builders and purpose are unknown. On 22 March 2002 and 4 July 2004 we visited the best preserved of all drakospita on top of Mt. Oche, measured its dimensions and calculated its orientation based on the azimuth of sunset and moonrise. A Sirius-rise orientation corresponding to ca 1100 B.C., not inconsistent with previous archaeological dating based on artefacts found inside the structure, indicates a religious/astronomical purpose for the building. It could probably be argued that at least the famous drakospito at Mt. Oche was not only a place of worship but also an ancient astronomical observatory.


The first modern reference to these buildings dates to 21 October 1797, and was recorded by the British geographer and geologist John Hawkins (1758–1841). He was the first to discover the drakospito on Mt.Oche, and believed that it was an ancient temple. Other Greek and overseas researchers followed, including H.N. Ulrichs (1842), G. Welcker (1850), L.Ross (1851), M.J. Girard (1851), G. Bursian (1855) and, more recently, Th.G. Papamanoles (1954).


Three drakospita near Styra, known as Pálle-Lákka Dragò, are especially imposing, but most impressive of all is the drakospito on Mt. Oche. None of the others preserves the perfection of its construction.


In 1959 Professor Nikolaos K. Moutsopoulos studied the Mt. Oche drakospito and eleven similar buildings, and excavated the surrounding space in 1960 and 1978-1980. Inside the Mt. Oche building he discovered numerous pots, while outside the building he located an apothetes, i.e. a subterranean construction inside which some utensils and animal bones were found, as well as pottery fragments and inscriptions dating from the Preclassical Period to the Hellenistic Period; on one of the potsherds were inscriptions in an unknown kind of writing.


These relics are now housed in a small archaeological museum at Karystos (inside the Yokaleio Cultural Foundation), where one can also see a couple of finds from other drakospita near Karystos and Styra.


It should be noted that apart from southeast Euboea no other places in Greece have drakospita, if we exclude some markedly smaller similar constructions in Mane (southern Peloponnese) or, according to Carpenter and Boyd, a solitary example on Mt. Hymettos in Attica.


The southern Euboea area was known in antiquity for its quarries, as mentioned by Strabo (X 16). So, although the drakospita themselves are not made of marble, some researchers have hypothesized that they were the residences of the local quarry workers. Maybe the smaller ones could have been erected, or simply used, by such people, but this hypothesis seems improbable for the largest one of all, on Mt. Oche, because of its position on the very top of the mountain, a hard-to-reach and cold place.

The Mt. Oche drakospito lies at an altitude of 1386 m (4547 feet), on the tiny plateau formed between the twin peaks of the mountain. Access is rather difficult and requires some mountaineering ability, but not special climbing skills.


The ancient building is an approximate rectangular parallelogram made of large blocks of rock, weighing up to 10 tons each, and the way in which the blocks fit together and the overall quality of the construction is impressive. We carefully measured the dimensions of the main building. The largest of the stone blocks is 4.0 × 2.0 × 0.4 m. All of the blocks of rock seem to have been extracted from the same area, and geologically they are amphibolites, rocks composed of silicate minerals. From the inside we could testify to the excellent state of preservation. Indeed, the strength of the construction and the feeling of safety offered by this megalithic monument prompted the people to think of it as the creation of supernaturally-strong beings, dragons or Cyclops. The lowest blocks are fitted into the natural rock substrate, while—where needed—cavities were filled with smaller stones. No trace of any connecting mortar, such as mud, was detected.

The entrance of the drakospito is made of three slate blocks (a trilithon) forming a Π
shape, a common feature of all ‘dragon houses’.


The top block on the Mt Oche drakospito measures 1.2 m × 2.3 m × 0.2 m and sits at a height of 2 m. The thickness of the walls is everywhere larger than or equal to 1.40 m (for comparison, one member of the Pálle-Lákka Dragò trilithon has average wall thickness of 1.17 m, and another one 1.05 m). The interior comprises just one room, which measures 9.80 m long and 4.90 m wide, i.e. a 2:1 ratio, forming a space of about 48 m². The height of the walls is 3.45 m and that of the building approximately 4.5 m. The only wall with windows is the southern one, where two small windows exist, approximately 40 cm wide, one to each side of the door opening, allowing a small amount of light to enter the building (as is the case in most temples and churches, in order to create a proper atmosphere).

The construction method of the whole building appears to have solved serious structural problems. The construction of the roof follows the ecphoric method on all four sides, and not only on two sides, as is the case with the Mycenae megalithic monuments. In order to construct a roof with this method or system, one needs both accurate calculations and good craftsmen.


The structural study must be accurate, because if the weights of the slates is not calculated correctly, the barycenter of the whole pile will exceed the edge of the supporting wall, and the roof will collapse. The unknown constructors of the ancient building, thinking cleverly, not only made very thick walls, but also used large rocks as counterweights placed upon the first slates on the parts that were resting on the thick wall. Also, the slates are not horizontal, but were slightly inclined, for the draining of rainwater.

The lengths of the exterior walls are: 12.70 m (north), 7.70 m (east), 12.60 m (south) and 7.75 m (west). It should be stressed that the structure and texture of the walls is such that the accuracy of the measurements can be no better than approximately 5 cm. The structure should be further studied in respect to its mathematical analogies, since the ratio of length to width (1.64) is very close to the ‘golden ratio’ or ‘divine analogy’ of Φ≈1.618:1, a ratio that appears during the Classical Period mainly in the vertical plane to increase the aesthetic appeal to an external viewer.


The habit of giving an astronomical alignment to religious buildings is common in Greece, both in ancient and mediaeval times, with the sunrise and sunset at certain dates being especially favoured, as reported by Pantazis (2004). Having excluded the sunrise and sunset at solstices and equinoxes, an obvious first choice was to check for possible astronomical alignments among the brightest stars, and especially Sirius, since the orientation towards the southeast was compelling. Indeed, by using two separate astronomical planetarium programs, Redshift 5.1 and Cartes du Ciel 2.75, we discovered a rise of Sirius orientation of the southern wall for 1060 B.C. ± 30 years and of the northern wall for 1150 B.C. ± 30 years, the average for both walls being 1105 B.C. (the uncertainties correspond to the 5′ error mentioned above). The dating of the construction of the building at that time is not at odds with the archaeological evidence, as Moutsopoulos (1960) assigned this drakospito an eighth century B.C. date based upon artefacts found inside the building.


The ancient word ‘drakon’ (modern Greek: ‘drakos’, from which the modern term ‘drakospito’ was derived) can be traced back to the ancient Greek verb δέρκομαι, which means to see clearly, to watch, to observe. Indeed, the tenses of the verb are: δέρκομαι, εδρακόμην, δρέξομαι, έδρακον, δέδορκα  and εδεδόρκην. We see that the root of the past tense (drak-) gives us the word dragon (δράκων), which in Greek means “… the one who observes”! A dragon is a creature with excellent vision … Therefore, the name drakospito is a paretymological term (i.e. where the word takes on a new meaning), and a substantial use of these megalithic monuments, as suggested by the ancient Greek verb δέρκομαι, was that of an ‘observatory’: either a watch-tower (for observing the Aegean Sea) or an astronomical observatory (for observing celestial phenomena and heavenly bodies). This seems especially true for the largest and best preserved structure of this kind, the Mt. Oche drakospito.


Conclusions Whatever their actual function, the distribution and the variety of these megalithic monuments is an indication of a certain level of continuity in the construction of cyclopean buildings in Greece. The use of monoliths and the exquisite manner of fitting the stone slabs together were true architectural challenges.

The uniqueness of the drakospita provides a challenge for future researchers who now need to carefully examine the two dozen or so surviving buildings in order to ascertain whether their construction reflects some astronomical orientation or mathematical rules. From houses of dragons and giants, and palaces for the kings of the Cyclops, they were abandoned or became sheep-folds and the residences of shepherds in recent centuries. Our hypothesis that at least one member of this group of monuments was originally used for astronomical observations could give new momentum to research, quite apart from the interest drakospita present from an archaeological and architectural point of view.

(Source: “Study and orientation of the Mt. Oche ‘Dragon house’ in Euboea, Greece”, by E. Theodossiou, V.N. Manimanis, M. Katsiotis, D. Papanikolaou – 2009)


Abstract The preliminary investigation of the astronomical orientation of monuments at Styra, southern Euboea, includes the triple so called ‘dragon house’ complex at Laka Palli, one dragon house at Kapsala and one on the summit of Mount Oche, and a monumental gate in the megalithic fortification wall at Armena. Recent luminescence testing dates these remains to approximately the Classical period with apparent re‐use in Roman and later times. Thus far no definite astronomical orientation has been determined in the layout of the structures and no celestial stone markers or similar features have been found associated with the dragon houses, however, the relationship of significant stars, constellations and solar stands was well known in antiquity, and the limited results presented here suggests the possibility for further investigation. This study, which considered possible orientations related to sun rise and sun set for the summer and winter solstice, as well as, alignments towards equinoxes and major bright stars and constellations, did observe a general preference for a southern orientation at most of the sites and a possible feature for time observations in the dragon house complex at Laka Palli.


A recent dating project (Liritzis et al., 2009) of the stone foundations of the Laka Palli and Kapsala dragon houses and the Armena Gate dated these buildings to the 5th century BC. Therefore the date used in all celestial maps is ca. 400 BC.

Theodossiou et al. (2009) have suggested that the Mt. Oche dragon house is aligned to the rise of Sirius (alpha Canis Majoris). Since this alignment corresponds to the orientation of the right side of the building (i.e. the wall on the right as we face the entrance), we decided to measure the entrances and the right sides of the rest of the dragon houses as well, to compare our results.

Due to the small number of buildings investigated and their diversity in orientation, it is impossible to determine if any of the buildings had deliberate stellar alignments. There are, however, some noteworthy results to be considered. Probably the most impressive alignment is that of the Laka Palli East house to Orion’s belt. At ca. 400 BC the three stars of the belt would be positioned directly over the point on the westerm horizon toward which the house is oriented. Even if this alignment was not intended, it would hardly have been missed.

The Laka Palli South house is aligned toward the head of the constellation Draco, when it would be at its lowest point in the sky.

The declination of the Laka Palli North house and the Kapsala house does not correspond to a significant star or constellation. On the other hand, the constellations of Canis Major, Scorpius, and Sagittarius would pass a few degrees above the horizon of these buildings.

The entrance of the Mt. Oche dragon house is aligned to the constellation Crux (Southern Cross) at the time of its setting. Additionally, the two brightest stars of the constellation Centaurus would also pass over this point.


While the Mt. Oche house is aligned to the star Sirius in Canis Major, as previously proposed (Theodossiou et al., 2009), and other bright constellations, none of the other buildings follow this orientation. In the case of Laka Palli, the relevant constellations are the same to those that are visible from the entrances of the buildings, due to the layout of the complex. At Kapsala, only two constellations of low illumination were observed.

A special mention should be made about the ‘opaion’ (i.e., opening) in the ceiling of the Laka Palli East house. A view through the opening at ca. 400 BC would show many bright stars and constellations at their zenith, namely: Vega in the constellation Lyra, Deneb in
Cygnus, Capella in Auriga, Hercules, Corona Borealis, Perseus, and Bootes. Each constellation would be at the zenith on a different month, for a given time. Although it cannot be proven, the ‘opaion’ might have been used for time observations, in addition to other possible uses (religious, social). Unfortunately, the exact size (or even the existence) of the opening in ancient times is not certain.

The Armena fortress gate has a northwest‐southeast orientation, which is very close to the summer solstice sunset and the winter solstice sunrise. In fact, both solstices are within the error margins. Looking outwards (to the northwest), the gate is also aligned to the constellations Leo, Gemini, Cancer, and Andromeda. In the opposite direction, the gate is aligned to the constellations Canis Major, Scorpius, and Sagittarius, as well as the peak of Mt. Oche. However, this line of sight must have been blocked by other structures inside the fortress. Thus, only the summer solstice sunset could have been observed through the gate.


Conclusions the small number of studied buildings does not allow us to reach a safe conclusion regarding the orientation of the dragon houses. Nevertheless, there is at least an indication that the general southern orientation of the Kapsala and Oche dragon houses is significant. This research will be expanded to include the rest of the dragon houses in the Styra area, in order to create a valid statistical group that will show which (if any) of the proposed alignments are deliberate and not as arbitrary as they currently seem.

The diversity in astronomical alignments is not the only element that separates the studied buildings. The Mt. Ochi house was built at the highest point of its landscape, while the others were not. The Laka Palli houses lie sheltered at the lower part of a valley and the Kapsala house is located in an open area. The entrances of the Ochi and Laka Palli houses have the best overview of the nearby area, but the Kapsala house looks to the opposite direction (probably to avoid the northern winds). Thus, it is quite probable that these dragon houses were not built to serve the same functions (Carpenter and Boyd, 1977). Archaeological excavations are necessary to provide additional information about the function of each building.

In this preliminary study, no clear relationship between the investigated alignments and celestial elements pertaining to the gods Hera and Zeus or Hercules have been discovered that might support a potential cultic function for the dragon houses.

The possibility that the Armena gate was deliberately aligned close to the summer solstice sunset should not be hastily rejected.

(Source: “Astronomical orientations of Dragon houses (Lakka Palli, Kapsala, Oche) and Armena gate (Euboea, Greece)”, by Liritzis, I. and Artelaris, G.)


You can see the ‘Dragon House’ of Ochi in the following brief video:

NovoScriptorium: A totally different opinion is presented in the paper titled “The Dragon houses of Styra: Topography, Architecture and function“, by Reber, K., of which we present some selected parts below:

Abstract This paper deals with the so‐called Dragon houses in southern Euboea and particularly in the region of Styra. In the publications of both N.K. Moutsopoulos and T. Skouras a series of different houses are mentioned, but without precise information on their location. Thus, my first aim was to rediscover these houses and to verify wheter they correspond to the type of the well‐known houses at Palli Lakka, Kapsala and on Mount Ochi. I soon realized that some of the Dragon houses in the lists of the two authors are in fact ancient farmhouses or watch towers.

I will present the typology of the “real” Dragon houses and will also compare them with some stone houses found in other regions around the Mediterranean: in Italy, France, Switzerland, and the former republic of Yugoslavia. Despite different opinions concerning their function, I will add some arguments in favor of that proposed by Moutsopoulos. I believe that we are dealing with shepherd dwellings, used perhaps only part of the year, mostly in summer, when shepherds crossed mountains with sheep and goats or maybe even with some cows.


The buildings, briefly introduced here, have a variety of common characteristics, which may be evaluated in the definition of a Dragon house. First, all constructions are completely of stone. Second, the building material consisted of stone from the immediate vicinity. Third the roofs have the form of a corbel vault, built with large, flat stones. Fourth, the buildings stand isolated in the mountains, i.e. detached from any settlement patterns and far from commercial or traffic routes. The structures are all oriented to the south and are associated with an enclosed area or courtyard.

Stone buildings with corbel vault are known through all periods, not only on Crete, but also in many other locations around the Mediterranean. Already in 1925, F.P. Johnson had compared some ancient Carian constructions on the peninsula of Halicarnassus with the Euboean Dragon houses (Johnson 1925; cf. Radt 1970, 196; Carpenter and Boyd 1977). At Mt. Aipos on the island of Chios, Vassilis Lambrinoudakis found other ancient stone constructions with corbel vaults (Lambrinoudakis 1982), and a similar building is also known on Mt. Hymmetos in Attica (Carpenter and Boyd 1977, 189‐193 fig. 16‐20).

In addition to these ancient constructions we also know a large number of modern stone huts which continue the ancient tradition of corbel vaulting. The geographically scattering of examples from different ages show that we can hardly speak of mutual influence, but that the phenomenon of stone buildings with console or corbel vault in mountain areas can appear in different times at different places (Hamm 1974; Santillo Frizell 2001; Santillo Frizell 1989).

The isolated position of the Dragon houses in the mountainous region of southern Euboea, like the apparently timeless and geographically independent typology of the corbel construction, speaks in favor of the interpretation suggested by Ross and Moutsopoulos as lodgings for the shepherds, stables, and for cheese and dairy processing.

We may assume that the Dragon houses in the region of Styra had served exclusively as summer lodgings for the shepherds.


NovoScriptorium: We partially disagree with the above opinion; some of the smallest ‘Dragon houses’ could have been indeed linked to shepherds’ life, but, for the big ones, especially Ochi, we do not find any reason at all, no logic, why would someone put so much effort just to…house cattle.

Finally, we present selected parts of the only, to our knowledge, paper that properly deals with the dating of the structures.


Abstract The Optical Stimulated Luminescence (OSL) surface dating employing the single-aliquot regenerative (SAR) technique on quartz was applied to some small enigmatic buildings made of large marble schist slabs in a skillful corbelling technique, and a fortified megalithic gate, at Styra, Kapsala, Laka Palli and Kastro Armena in southern Euboea. The function and origins of the structures have created a puzzle that has fed the imagination and lead to various interpretations by many scholars. No archaeological excavations or methods of dating have been available for the megalithic‐like structures. The dates reported suggest the earliest construction to have taken place during the Classical period. Re‐use of these structures has occurred during Hellenistic and Roman times (the latter associated with the large scale quarrying of marbles), as well as, in Medieval times (found in agreement with the historical literature) and the contemporary period (as reported by shepherds). In all cases the datable slabs were rather reset as repairs.


The Dragon houses at Styra have variable dimensions of about 3.50 ‐ 6.00 width x 13.0 ‐ 9.00 m length, and wall thickness of ~2.00‐2.50 m.

The thickness of the megalithic slabs varies in size: 1.70x 0.84 m, 2.10 x 0.60 m, 2.00 x 0.65 m, and the slabs in the trilithon (pi like dolmen) are 2.30 x 1.60 x 0.25 m, weighing around 10 tons.

This slideshow requires JavaScript.

In this paper we approached the date of construction of the structures by applying surface luminescence dating method of monuments introduced by I. Liritzis in 1994 on the actual stone blocks (Liritzis 1994; Liritzis et al., 1997; Habermann et al., 2000; Greilich et al., 2005; Liritzis 2010a, b). The sampling was made on three sites:

a) the complex of three Dragon houses at Styra; b) the Dragon house at Kapsala, (both sites have been used by shepherds and quarry workers and are covered with thick vegetation and no archaeological excavation has taken place at either site); c) the trilithon (pi shape) Gate and fortified wall at the Armena fortress (for additional photos, see, Liritzis and Artelaris, 2010)*.


Conclusion The dating of Dragon house construction with the surface luminescence dating (OSL) of quartz grains present in the calcareous schist provided ages that define the original construction, as well as, later re‐uses. Dates of Classical, Hellenistic‐Roman, early Byzantine, Medieval and contemporary times are reported, which are in concordance to the historical and archaeological data.


*NovoScriptorium: Therefore, we still have no proper dating of the ‘Dragon house’ of Mt. Ochi, which is actually the best preserved, in the highest altitude and the most ‘complicated’ and interesting of all.

(Source: “Surface luminescence dating of ‘Dragon houses’ and Armena gate at Styra (Euboea, Greece)” by Liritzis, I., Polymeris, S.G. and Zacharias, N.)


NovoScriptorium: Megalithic structures constitute a cultural characteristic of a deeper Antiquity, other than the Classical or the Archaic Age. The closest chronically ‘megalithic building fashion’ comes from the Iron and the Bronze Age. Recently, Megalithic structures from the Neolithic Age were also found in Greece (read here:

Indeed, the Megalithic tradition appears to have had a continuum from the Final Neolithic Age until the Iron Age. But during the Classical period it was undoubtedly abandoned. Concluding, for us, the dating of the Megalithic structures in Euboea (well, not all of them) in the Classical era, is a quite unexpected and surprising result (to say the least). It implies that the Megalithic tradition was a cultural continuum expanding from the Final Neolithic Age to the Classical Age, something which is remarkable.

We are not though fully convinced about these results. If the ‘Dragon houses’ were built during the Classical era wouldn’t we have something very specific written about them, that would come from the same era, explaining their usage or how they were built? If people continued to build Megaliths during the Classical times, then why don’t we find similar Megaliths all over the Classical Greek world? Nevertheless, for the time being, and until some new credible research appears on the issue, we must stick with Science and accept that these structures belong to the Classical Age.

Research-Selection-Comments for NovoScriptorium: Philaretus Homerides


3 thoughts on “The ‘Dragon houses’; Megalithic monuments in Euboea, Greece – What we know so far

Add yours

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

Up ↑

%d bloggers like this: