THE BIG MOUND AND ANCIENT VILLA ARMIRA ARCHEOLOGICAL DATA ON EARTHQUAKES IN THE EASTERN RHODOPES

Transcription

1 Geoarchaeology and Archaeomineralogy (Eds. R. I. Kostov, B. Gaydarska, M. Gurova) Proceedings of the International Conference, October 2008 Sofia, Publishing House St. Ivan Rilski, Sofia, THE BIG MOUND AND ANCIENT VILLA ARMIRA ARCHEOLOGICAL DATA ON EARTHQUAKES IN THE EASTERN RHODOPES Gergana Kabakchieva National Archaeological Institute and Museum, Bulgarian Academy of Sciences, 1000 Sofia ABSTRACT. The report presents the observations and archaeological evidence for earthquakes in the area of the ancient villa Armira (Ivailovgrad Region, Eastern Rhodopes) in Antiquity. Data about earthquake destructions are been published about a few archaeological sites and monuments in Bulgaria (Kuzmanov, 1994, 24-35): the evidence of the ancient town near Kavarna, of Dionyssopolis (Balchik); of the Roman colony Ulpia Oescus (Gigen), of the Roman fortress Dimum and the Donube (Belene). Generally, the knowledge about archaeoseismology not only in Bulgaria, but in the whole of Europe is not large (Galadini et al., 2006, ). The ancient villa Armira was discovered more than forty years ago, 4 km South-West of the town of Ivailovgrad. A lot of research and publications have been dedicated to this ancient monument (Mladenova, 1990; Kabakchieva, 1986; 1995, ; Kabakchieva, 2005, , references). In 2001 rescue excavations were organized for one of the burial mounds from the family necropolis of the villa owners (Kabakchieva, 2002; 2003; 2005, ). Such excavations were carried out because of the destructions caused by treasure hunters in an attempt to enter the mound embankment and to discover the ancient wagon near the mound. These excavations became the reason for re-opening of archaeological field work in the area of the ancient villa Armira. The Big Mound is located in the area of Harmanite, East of the village of Svirachi and South of the villa Armira. It is located on the top of a hill which borders on the South with the Armira River and its valley. The mound was erected about 2000 years ago at the southernmost part of the above mentioned hill. Due to the massive size of the mound (60 m diameter and original height of about 20 m), its study lasted for a long time. The first archaeological campaign took place in the period , and the second in The excavations during the first campaign were funded by the Foundation Center for Supporting the Discovery and Preservation of Cultural Valuables and the Foundation Horizon, the Netherlands (the author would like to express his personal and the team s gratitude for the support). The funding of the second excavation campaign was within the project The Splendour of Armira under PHARE programme, transborder cooperation Bulgaria Greece. The studies were finalizing ones and their results are the subject of this report. The main results from the first campaign in brief are the following: the complex of the two ritually buried ancient cars next to the mound was excavated (one ritual and richly decorated and the other a common ancient wagon). The ancient cars were part of the Thracian burial ritual and were buried on purpose South of the mound. The front part of the complex was studied by archaeologists. That is where the skeletons of four horses and one dog were found in anatomic order, with rich decorations of bronze, silver and gilded applications from the yoke and horsecollar of the ritual burial wagon, as well as iron elements from the ordinary one. The applications, richly decorated with mythological scenes and characters, are preserved at the National Historical Museum in Sofia; a very massive crepis was found, which turned out to be a more complex stone step facility; the mound embankment was studied in the Western part of the mound; a rich grave of a cremated woman was discovered; three secondary graves (belonging to small children) were discovered, quite destroyed. It was established that in the period of the end of the Russian-Turkish war of 1878, the solders tried to excavate the mound. The results of the second archaeological campaign in 2007 were the following: the mound embankment was completely excavated in the Eastern part of the mound; over 30 trial trenches were made, under the level of the ancient terrain; new graves were not discovered; it was established that before the erecting of the mound, there was a spring at this spot; it was established that the spring dried up suddenly, probably due to an earthquake; 380

2 archaeological evidence was found for a natural cataclysm an earthquake from the end of the IV c. AD. This earthquake was the reason for the ruining and collapse of part of the massive stone facility in the karst caverns under the hill and under the northeastern part of the mound. On the basis of the results from the excavations at the Big Mound and the observations on the terrain of villa Armira, it was established why such a large quantity of marble and preserved floor mosaics have been discovered in villa Armira. The first issue is related to the pre-history of the Big Mound near the Svirachi village. In trench N15, in the Western half of the mound (Fig. 1; plan of the trenches), carried out under the ancient level of the terrain, the following objects were found: two gradually filled up pits from the ancient spring, which gushed from the top of the hill before the mound was erected. One pit was registered in the Eastern part of the excavation and the other in the Western. The former was smaller in size and at elevations of m. Under a thick layer of black silty clay in the two pits, there was a layer of sand, which was very easy to dig. Fragments from ceramics from the Bronze Age and up to the end of I mill. BC were found in the black clay from the filled-up pits (the former head of the spring). The vessels are thick-walled pots in shape, handmade, as well as a big semi-spherical bowl (the prehistoric ceramics was identified and dated by Borislav Borislavov, Sofia University St. Kliment Ohridski, for which the author would like to express his gratitude). A small number of animal bones were also found in this trench. Fig. 1. Trenches in the Big Mound by Svirachi Directly related to the spring are also the results from trench 14, which was planned in one axis with the centre of the trench 15, but outside the mound and immediately next to the stone crepis. There, it was found, that under the layer of rubble, spread out waste material from the treatment on site of the greenish stone for the mound, there is a thin layer of brown soil, followed by sand and gravel. The geomorphologists took samples from this trench. Even before their appearance, however, it was clear that the water sprang from here. This observation is also based on the fact that about a hundred metres down the hill, there is a spring, even today. One can suppose that it is the same one, which before the natural cataclysm the earthquake, flowed from the very top of the hill. The dating of the earthquake, which led to the interruption of the spring on top of the hill, is based on the ceramics found in the filled-up pits from the head of the spring, as well as on those, found in the mound embankment itself. Here, it must be noted that during the first archaeological campaign the following questions have been asked: where did so much black silty clay come from, once for outlining the sacred area of the mound, as well as for its accumulation. This layer, spread out at the level of the ancient terrain is 0.40 m thick and covers the whole internal area of the mound, which one would call a perfect engineering facility, and at every certain level, a layer of black silty clay was laid. It strengthened the mound embankment, but probably it also played another semantic role in the Big Mound; the second question was related to the ceramic fragments found in the embankment of the Big Mound these can be systematized in two groups: one, that were found only in the layers of black silty clay, and the others which can be related to the period of the initial plundering of the stone material from the mound. Ceramics from the first group was found only in the layers of black clay. Thus, the time when the spring existed can be determined. It is from the group of the common ancient ceramics and the group of amphorae. The vessels were made from average and well purified clay of red colour. The forms are of jugs and amphorae of a local production (Fig. 2). Very few fragments from those found are of pots (Fig. 3). The ceramics from the layers of black clay date back to the period of the second half of the I mill. BC I c. AD. The above facts allow the connection of the disappearance of the spring on top of the hill with the period of I c. BC the first half of I c. AD. As already described, there are springs existing to this day further down the Western side of the hill and on the South. The spring on the South was used for the village and fragmented clay pipes have been found. The ceramics from the second group is from the IV c. AD. The fragments are from common and varnish ceramics and the vessel forms are jugs, pitchers and a dish (Fig. 4). On the basis of the excavations in 2007 and the discovery of the location of the pits in the trench 15, it became clear that the black clay is from the long-existing spring and the buried vegetation around it. The second natural cataclysm earthquake, dates back to the time of its initial destruction in IV c. AD. In the decades after Christianity was declared as the official religion in the Roman Empire, looting of the monumental tomb monument took place gradually. First the ortostats and in the upper end of the stone facility were affected. What gave rise to the looting, however, happened at the end of IV c. AD. It was then, during a powerful earthquake with a most probable epicenter to the Southwest of the Armira River valley, that the mound was seriously shaken and when the shock wave traveled back, there was a rupture of the circle and terraced construction. It was built in a way similar to that applied by ancient architects and engineers in the construction of theatres, amphitheatres and stadiums during the Roman Age (Fig. 5). 381

3 a a b b Fig. 3. Thracian ceramics, a pots; b other shapes c Fig. 2. Pottery of the trench 15, a - pots; b plates; c other shapes Fig. 4. Ceramics from the IV c. AD Fig. 5. The stepped stone construction of the Big Mound 382

4 Archaeological evidence for the occurrence of the earthquake at the end of IV c. AD consists not only with the dating of the fragments of ceramic vessels found in the destroyed areas. During one of the first trial trench in 2001, around the Northern external face of the crepis, a coin was found, dating back to the same period (inventory number 9/2001; the coin was identified by B. Bozhkova, for which the author expresses his gratitude). So, what was actually found during the excavations at the Big Mound in 2007, in terms of archaeological evidence for an earthquake? Traces of it were most visible in the Northern and Northeast quarter of the mound. Of course, in 2001 it was already determined that in this part the mound embankment was lower and the blocks from the stone stylobate have been looted or are missing. In addition, at places, there was a visible dislocation of layers and traces of strong natural polishing between the layers, which can occur only as a consequence of a natural cataclysm. What were the results of the archaeological studies in 2007? Firstly, in the trench 26, a dislocation of the stylobate was established. Part of this stone construction was discovered at 1.20 m South of the rest of stones (the stylobate of the upper part of the stepped construction) (Fig. 6). In trenches 22, 26, 28, 33, 34 and 35, there were registered dislocations, ruptures and falling of the stone stepped construction. At other spots in these trenches, also was registered significant collapse of earth mass from the mound embankment. Thus, layers of the black and other clay were cleared and slid underground. The most probable explanation for that is that they fell in karst caverns under the hill. The inclination of fall and the slip of the layers are about 70 degrees. During the excavations, three spots of a major fall-through were determined, i.e. three big caverns under the hill in the Eastern and Northeastern part of the mound. Another indirect argument for the dating of the earthquake is the finds from the opening of the ancient villa Armira. Life there also stopped at the end of IV c. AD. Fig. 6. Trench 35, with traces of earthquake Of course, the history of the villa is a bit different. Founded and constructed as early as the third quarter of I c. AD, the villa prospered for about three centuries and was burnt and looted during the war between the Romans and the Goths in The marble palace of the owners was most probably burnt down after the battle at Hadrianopolis (present-day Edirne, Turkey) on the 9 th August 378. The ruins of the building were briefly inhabited by common people, without any sense of ancient architecture and art. They used a capitel as the leg of a table. Other architectural marble elements were used as stairs for entering the room (room 11 from the villa plan), which they inhabited. Probably the rest of the building was in such a condition that it was not possible to enter. Ten-fifteen years after the fire, there was the earthquake, which led to landslide from the hill above the villa and the burial of its ruins under a thick layer of clay brown soil. It was only after the excavations at the Big Mound in 2007, that one could also answer the question why so much marble and mosaic was found at villa Armira. The reason is that this burial and a natural cataclysm similar to Pompeii led to preservation of ancient art, which today, 2000 years later, can be seen restored by the admirers of the ancient and by tourists (under PHARE project The Splendour of Armira ). Traces of an earthquake were also registered in the next studied mound from the family necropolis of the owners of ancient villa Armira, discovered at the village of Mikridoxipara on Greek territory (Pavlides et al., 2006, 91-93). In addition, during the excavations of a public bath in Nicopolis ad Nestum (near the village of Garmen, by the town of Gotse Delchev) traces were found for its destruction by an earthquake (Kuzmanov, 1994, 24-35). The excavations and the discovered large pieces of the upper part of the building, which fell an bloc, led the author to conclude that the public bath was 383

5 destroyed by an earthquake. The newest excavation in the ancient millitary camp Dimum (Belene, Bulgaria) at the Danube display the traces of an earthquake too. A wall of a big building, probably principia fall down an bloc and this demolition can be explaned as a result of the earthquake (the excavation in ; leader of the project is G. Kabakchieva). In the center of Roman city Ulpia Oescus there are also traces of an earthquake. Several stone columns on the main street of the town were found in situ and fallen down parallel one to other (excavation of T. Ivanov). Outside the Balkan peninsula, among the Roman provinces, well studied traces of earthquakes are those from Carnuntum, the capital of the former Roman province Pannonia Superior (Austria) (Kandler, 1989, ; Decker et al., 2006, ). Similar date was established in the surroundings of the town the near-by Roman villa at Bruckneudorf (Zabehlicky, 2004). Archaeological studies do not provide a lot of examples for precisely determined situations and evidence for earthquake activity in Antiquity. This is why it can be believed that the results from the findings during the excavations of the Big Mound and ancient villa Armira in the region of the Еastеrn Rhodopes will expand the knowledge about natural factors whose impact is still underestimated in the overall study of historical monuments in Bulgaria. Further on geophysical studies may provide additional information. References Decker, K., G. Gangl, M. Kandler The earthquake of Carnuntum in the fourth century A.D. archaeological results, seismologic scenario and seismotectonil implications for the Vienna Basin fault, Austria. J. Seismology, Galadini, F., K.-G. Hinzen, S. Stiros Archaeoseismology: methodological issues and procedure. J. Seismology, 10, Kabakchieva, G Ceramics from the Ancient Villa Armira near Ivailovgrad 2 nd -4 th c. Sofia, 186 p. Kabakchieva, G On the history of ancient villa Armira and its owners. In: Heros Hephaistos, Studia in honorem Liubae Ognenova-Marinova. Veliko Turnovo, Kabakchieva, G Neue Untersuchungen an der roemischen Villa armira in der provinz Thrakia. Balacai Koezemenyek III. Vesprem, Kabakchieva, G Excavation of the Big Mound by Swirachi. Archaeological excavations and research for Sofia (in Bulgarian). Kabakchieva, G The excavation at Swirachi. Archaeological excavations and research for Sofia. Kandler, M Eine Erdbebenkatastrophe in Carnuntum? Acta Archaeologica Acad. Sci. Hung., 41, Kuzmanov, G Late Roman bath in Nicopolis ad Nestum. Archaeology, 2, Mladenova, Y Ancient Villa Armira. Bulgarian Academy of Sciences, Sofia (in Bulgarian). Pavlides, S, A. Chatzipetros, A. Zervopoulou, A. Kuercher, D. Triandafillos, D. Terzopoulou Archaeology and seismic hazard: post Roman co-seimic fault ruptures in northern Evros (Mikri Doxipara Zoni, NE Greece) case study. 11 th Intern. Symposium on Natural and Numan Induced Hazards, Petras, Zabehlicky, H Zum Abschluss der Grabungen im Hauptgebaeude der Villa von Bruckneudorf. Jahreshefte des Oesterreichischen Archaeologischen Instituts in Wien,