Sergey Torbatov

In the northern part of the Chartalia locality, about 0.6 km east-southeast of the centre of Nikopol, a previously unknown settlement centre from the Late Iron Age was discovered. It occupies the broad and flat herbaceous terrace of a plateau-like elevation with a high and steep southwest slope and a gentle, slightly pronounced slope to the northeast. The terrace is consists of two clearly distinguishable steps, the southeast one of which is smaller in area and higher. It was fortified with a powerful earthen rampart, from which a sector with a length of about 50 m is now clearly visible on the eastern side. Its maximum preserved height is about 2 m, and the width at the base is 4-5 m. In places in front of it remains of a partly filled ditch can be seen. The rampart was probably once crowned with a stone wall. In its original form the fortification appears to have been with irregular semi-circular plan, the base of which rested on the edge of the high and steep south-west slope of the eminence, and the man-made defensive line encircled the higher step of the terrace, affording reliable protection from all accessible directions. The fortified area amounted to about 7 decares, with maximum dimensions of about 123 x 74 m. The fortification is identified with the acropolis of a settlement structure, which, in addition to it, also included an extensive unfortified quarter to the northwest, located on the lower terrace of the hill. A large amount of surface material was collected from the site – pieces of wall-plastering and ceramic fragments of household, table and storage vessels, worked by hand and on wheel, some with polished and burnished decoration. The dating of some of the specimens refers to the very end of the Late Iron Age, if not even to the initial period of Roman rule. The newly discovered settlement structure in the locality of Chartalia has all the formal features of a classical dava from the Late Iron Age, which are known in large numbers both north and south of the Danube.
According to the author, this is the Thracian settlement predecessor of the current town of Nikopol, and it is with this dava that the last stage of the functioning of the Thracian sanctuary within the limits of the Nikopol fortress is connected. That sanctuary should most likely be identified with its religious and cult centre. The two sites are in direct visual contact and are only 1 km apart of each other.

The first notices of the presence in the highest part of the Trojan Pass of an ancient fortification, located along the route and connected with the protection of the Roman road from Oescus to Philippopolis, which passed from there, date back to 1877. Its remains are still recognizable on the ground, but they are heavily defaced as a result of various construction activities in modern times. An opportunity to get an idea of the original appearance of the facility is provided by a detailed plan drawn up by Karel Škorpil at the end of the 19th century, which was recently discovered among unpublished materials in his archives. The so-called Markova or Trajan's Kapiya is a protected complex of the type of barrier walls. It was built in the locality Beklemeto, at the northern foot of the Stara Planina ridge. From a functional and terminological point of view, it is defined as clausura, but it is distinguished by a much more complicated fortification plan, compared to other similar facilities. It consists of a ditch and three parallel walls, located at some distance from each other. The defensive front faces north. A votive altar of AD 234 reused in the construction provides a terminus post quem for the dating of the structure. The initial period in the construction of similar barrier facilities in the mountain passes within the limits of the Roman Empire refers to the last quarter of the 4th and the beginning of the 5th century, and this practice became particularly widespread in the following centuries of Late Antiquity. The appearance of the defensive complex in the locality Markovi Porti should be generally attributed to the 5th-6th century AD.

The Gradishteto hill site near Riben has been known for long in the scientific literature. Until most recent time, it has been often wrongly identified with the Roman road station Ad putea. The excavations in 2013-2015 brought about much new evidence about the site but it was, unfortunately, partly misinterpreted by the investigators. The archaeological campaign in 2016 considerably enlarged the database and made it possible to thoroughly reconsider the traditional views and statements. It was proved that the habitation of the site began already in the Early Chalcolithic. In the Late Iron Age, if not even earlier, regular cult practices started to be performed here and a rock-cut sanctuary came into being. In the Roman period it developed into an important cult complex that ceased functioning in the late-3rd c. A burgus was built over its ruins in Diocletianic times, which was integrated in the mid-4th c. in the defensive system of a newly established fortified settlement. The latter suffered a destruction in the late-70s of the 4th c. but was soon rebuilt and resettled by new population, maybe of Gothic origin. By the mid-5th c. the settlement was again destroyed. The life resumed in the 10th c. and went on until the end of the 12th c. at the earliest. In the late-14th c. a coin hoard was buried in its ruins.

In May-July 2018 two consecutive rescue archaeological excavations were conducted on the southwestern outskirts of the urbanized area of St. George town quarter in Pomorie. Despite the small number of structures discovered, the results of these studies are fundamental to the final clarification of extremely important issues concerning the historical topography of Anchialus during the Roman period, the Late Antiquity and the Middle Ages, as well as to illuminating hitherto unsuspected aspects of the city's economic development in the earliest period of its existence. It was found that the studied area was inhabited or used for other purposes from the Early Roman period until the end of the 13th century, but with significant interruptions and varying intensity. The Roman city of Anchialus was established on a vast peninsula, only a small part of whose northeastern periphery had been inhabited in the previous period. The main part of it was built and the city subsequently grew not just on sterile sandy terrain, but on sand dunes. The raised eastern end of the present-day Pomorie Peninsula was still an island at that time, and it was there that the pre-Roman Anchialus, protected by "high walls", must have risen. Southwest of the ancient peninsula there was a vast and deep enough bay, which later became clogged. It is not yet known when exactly this natural ship's shelter was expanded and improved through construction of a powerful pier-breakwater from huge stone boulders, becoming a real harbour. It certainly functioned from the very beginning of the development of Roman Anchialus according to the archeological finds found in its waters. Through a series of large-scale shore protection measures caused by maritime transgression and carried out from the second half of the 6th century onwards, the harbour was improved. It is not possible to specify how long they lasted, but the accumulation processes that formed the sandy slopes and led to the formation of the current Pomorie Peninsula and the Pomorie Lagoon probably began in the second half of the 6th century and went on relatively quickly. By the end of the 11th century all that must have ended since long. The earliest among the newly discovered structures in 2018 date from the first to third quarter of the 2nd century and belong to a relatively rare archaeological monument not only in our country but also in the world-a production complex from the Roman period for processing (or possibly manufacturing?) of raw glass. Three different functional components of the installation, directly related to the technological process, were registered. In the southeastern part of the complex there is a rectangular tank furnace. Near the furnace, northeast, southeast and southwest of it, are situated three shallow basins with rectangular plan, slightly rounded corners and sloping walls. It is possible that they were used to prepare (wash? purify? mix?) the raw materials needed for the production process. An extensive auxiliary service platform with unspecified plan and dimensions was formed in the northwest, the length of which was traced in the east-west direction at more than 18 m, and in the north-south direction at about 7 m. Either from the beginning or in later time it was used to dispose of the waste remaining after the periodic breaking of the upper part of the furnace (or furnaces?) at the end of each technological cycle. A burned provincial coin of Emperor Hadrian was found on the platform, which is an approximate indicator of the time of operation of the production complex. The time of its coming into being cannot be determined in the current state of research, but it was certainly in operation until the late 70s or early

The subject of this paper is a fragment of a Roman milestone with a partially preserved Latin inscription, which is kept now in the archaeological department of the Regional Historical Museum in Pleven. The monument was found on the northern outskirts of the village of Byala Voda, Belene Municipality. This is one of the rare examples of finding a Roman milestone practically in situ.
Only the last three lines of the inscription are preserved, containing an abbreviation of the title of a provincial governor and a distance of one Roman mile:
…
leg (atum) aug (ustorum duorum)
pr (o) pr (aetore)
m (ille) p (assuum).
Based on the palaeographic features of the text and the abbreviations used in it, it seems most plausible in author’s opinion that that the milestone was placed in early-Severan times, perhaps in the years of the provincial governorship in Lower Moesia of Gaius Ovinius Tertullus, who is well known for unprecedented construction and repair activities within the province, largely focused on the road communications. The one Roman mile mentioned in the inscription seems to have been measured from the road station Securisca, whose exact location has not been established yet.
The paper also updates the database about the milestones from the Roman Danube road within the province of Lower Moesia.

The Gradishteto hill site near Riben has been known for long in the scientific literature. Until most recent time, it has been often wrongly identified with the Roman road station Ad putea. The excavations in 2013-2015 brought about much new evidence about the site but it was, unfortunately, partly misinterpreted by the investigators. The archaeological campaign in 2016 considerably enlarged the database and made it possible to thoroughly reconsider the traditional views and statements. It was proved that the habitation of the site began already in the Early Chalcolithic. In the Late Iron Age, if not even earlier, regular cult practices started to be performed here and a rock-cut sanctuary came into being. In the Roman period it developed into an important cult complex that ceased functioning in the late-3rd c. A burgus was built over its ruins in Diocletianic times, which was integrated in the mid-4th c. in the defensive system of a newly established fortified settlement. The latte...

a Секция за антична археология, Национален археологически институт с музей, Българска академия на науките, ул. A B S T R AC T In his work "De aedificiis" Procopius of Caesarea ends the description of Justinianic fortifications along the Ister River within the confines of Illyricum naming two sites located "at the extremity of the Illyrian territory"-Lapidar-ias and Lucernaria. The second one is mentioned once again in the following text of the treatise as a topographic marker, eastwards from which the territory of Thrace begins. The first part of this paper provides comments on the written evidence and presents a thorough critical analysis of its interpretations. The second part of the paper is focused on the existing localization hypotheses concerning both sites. They are decisively rejected, as being in obvious discrepancy with the ancient literary evidence, the geographic and archaeological realities. The last three parts of the paper are based on the results ...

A few years ago I received a picture of a mostly interesting altar with a Latin inscription. It had been seen and photographed somewhere in the country land between the villages of Alexandria and Abrit, Krushari Municipality in South Dobroudja. No other information about the monument is available, neither is known its present location.
The inscription is entirely preserved and reads: Dianae / Regin(а)e, / C(aius) Iul(ius) Lon/gus v(otum) s(olvit) l(ibens) a(nimo). On palaeographic grounds, it may safely be dated to the mid-2nd c. AD.
The dedications to Diana Regina were not so popular in the Roman Empire. They are characteristic almost exclusively of the Lower Danube provinces, with a single exception from Britain. The newly published monument is the next one in that series, bringing the total to 21 sure and 4 possible pieces of different kind (mostly altars but also two statues, a marble slab and a pillar).
The dedicator is a hereditory Romanized provincial, who maybe lived in Lower Moesia but whose origin was most probably from some other region of the Roman Empire. He might well be the son of the bearer of a military diploma issued in AD 112 (RMD V 344).

The “Gradishteto” hill site near Riben has been known for long in the scientific literature. Until most recent time, it has been often wrongly identified with the Roman road station Ad putea. The excavations in 2013-2015 brought about much new evidence about the site but it was, unfortunately, partly misinterpreted by the investigators. The archaeological campaign in 2016 considerably enlarged the database and made it possible to thoroughly reconsider the traditional views end statements. It was proved that the habitation of the site began already in the Early Chalcolithic. In the Late Iron Age, if not even earlier, regular cult practices started to be performed here and a rock-cut sanctuary came into being. In the Roman period it developed into an important cult complex that ceased functioning in the late-3rd c. A burgus was built over its ruins in Diocletianic times, which was integrated in the mid-4th c. in the defensive system of a newly established fortified settlement. The latter suffered a destruction in the late-70s of the 4th c. but was soon rebuilt and resettled by new population, maybe of Gothic origin. By the mid-5th c. the settlement was again destroyed. The life resumed in the 10th c. and went on until the end of the 12th c. at the earliest. In the late-14th c. a coin hoard was buried in its ruins.

The multilayered site of Gradishteto is located in the northwestern
surroundings of Riben village (Dolna Mitropolia municipality, Pleven
district). Since 2013, regular archaeological excavations are taking
place there. The research in 2016 appeared to be especially beneficial
providing significant additions and corrections with regard to the general
chronology of the site, its stratigraphy and the inner periodization of the
cultural layers registered on its territory as well as their function and
cultural and historical interpretation. A thorough scientific processing
has been accomplished of the multiple coin material from campaign
2016, found largely in a certain archaeological context. The results are
Четири колективни монетни находки от „Градището“... 143
of paramount importance for the satisfactory resolution concerning the
problems of the chronological periodization.
The subject of the work here presented are four hoards as part of the
numismatic material. Three of them are dating from the Late Roman Age,
and the fourth one – from the time of the Second Bulgarian Kingdom.
The first three hoards terminate in emissions of Valentinian I, Valens
and Gratianus. All the coins they contain have burned. The stratigraphic
observations and the construction analysis of the registered architectural
structures carried out in the course of campaign 2016 testify to the
existence of two separate phases of habitation of the fortified settlement
built there about the mid 4th c. AD. The composition and the condition of
the hoards in question coming from different places within the fortified
territory convincingly and accurately outline the time and circumstances
leading to the end of the first phase. The archaeological realities and the
numismatic data unambiguously show that the Late Antiquity settlement
near Riben not jus suffered, but even ceased to exist for a certain time
during the years of the Second Gothic war of Emperor Valens (376-378).
T he f ourth h oard c ontains t wo a sprae o f Tsar Ivan A lexander w ith
his son Michael Assen. The coins have been found beyond any adequate
archaeological context and most probably represent a small share of a hoard
its main part discovered in treasure hunters’ diggings within the site limits
not too long ago. The concealment of the hoard – among ancient ruins on
long since abandoned terrain but close to a road known for centuries and
probably still walked in those days, gives us grounds to associate it with the
tragic circumstances about the end of the Second Bulgarian Kingdom.

In his work " De aedificiis " Procopius of Caesarea ends the description of Justinianic fortifications along the Ister River within the confines of Illyricum naming two sites located " at the extremity of the Illyrian territory " – Lapidar-ias and Lucernaria. The second one is mentioned once again in the following text of the treatise as a topographic marker, eastwards from which the territory of Thrace begins. The first part of this paper provides comments on the written evidence and presents a thorough critical analysis of its interpretations. The second part of the paper is focused on the existing localization hypotheses concerning both sites. They are decisively rejected, as being in obvious discrepancy with the ancient literary evidence, the geographic and archaeological realities. The last three parts of the paper are based on the results of recent ground surveys and the use of sources of information, some of which are non-traditional for an archaeological study, together with discarding the far going misinter‑ pretation of the facts often featuring in many modern studies. Lapidarias is identified as the fortification situated in the locality Karierata/Borunya at 1 km to the northeast of Somovit. Lucernaria was placed in the locality Man‑ astirishte at 1 km to the northwest from Cherkovitsa. The Roman Danube road between Anasamus and Utus did not follow the Danube bank, but was traced across the heights between the the Osam and the Vit Rivers.

The first part of the paper discusses the written evidence about the Roman road station Anasamus, the Late Roman military fort Ansamus, and the Early Byzantine fortified settlement and later city Ἀσημοῦς/Ἀσήμος. All these toponyms refer to one and the same site, depicting its development and functional transformations through the ages. The second part presents a critical analysis of the opinions expressed so far about the exact location of the site. The conclusion is that the site had always been situated in the village land of Cherkovitsa, in the immediate vicinity of the Osam River’ mouth; on the left bank in the earlier period, and on the right bank during the Late Antiquity. The third part is a synthesis on the Early Roman Anasamus (a military camp, civil settlement and road station), based on the available archaeological and epigraphic information as well as personal ground surveys and reinterpretation of the evidence. The last part of the paper comments the remains of the Late Roman and Early Byzantine Ansamus/Ἀσημοῦς/Ἀσήμος, convincingly identified with the so-called “Osamsko kale” (Osam Fortress).

Mezdra is a small town and important railroad station in Northern Bulgaria. There is a small rocky hill on the south-western outskirts of the town, which rises on the left bank of the Iskar river. The hill is called Kaleto (‘The Fortress’) because of the impressive ruins of an old fortification, whose walls still rise to a height of some 7 m on the northern side. The walled area is about 2500 square metres. The site had been intensively excavated from 1974 until 1990. In 2003 the regular excavations resumed and are still going on.
The site proved to be a multi-layer one, with rather complicated stratigraphy. The earliest habitation layer dates back to the Late Chalcolithic. After a long hiatus, the hill was brought into use as a cult place in the Early Iron Age. Then there are no traces of human presence until the 2nd century AD.
By the middle of the 2nd century, however, a strong fort was built on the hill. It is considered to have been one of the numerous fortifications of different type constructed in the reign of Emperor Antoninus Pius to control the road traffic and internal security in the Roman province of Thrace. By size, the fort is identified as a praesidium. The control and protection of the nearby strategic ford across the Iskar river, on the Roman road from Oescus to Serdica, must have been one of the particular tasks of its garrison. The fort was in function for a comparatively short period of time and went out of use during one of the barbarian invasions between AD 170-190.
Some time after the disaster, most probably in the early 3rd century, intensive repair and restoration works began within the fort. However, they were abruptly terminated. Instead, a small temple was built in the central part of the walled area. That temple gave rise to an important sanctuary, which spread all over the territory of the former military post, using the surviving remains of the precinct wall as temenos. The earliest evidence for cult practices dates back to the time of Emperor Severus Alexander. The cult infrasructure of the sanctuary during its earlier stage includes a number of hearths and primitive clay platforms used as altars. Built altars, in the form of walled spaces of different shape and size, appeared only in the second half of 3rd century. The latest coins known so far, which are surely connected with the sanctuary, date to AD 290-292, but it probably continued to function a few decades more.
By the turn of the first quarter of 4th century at the latest the pagan sanctuary definitely ceased to exist. The remains of the former Roman fort underwent major repairs. The protected area was densely built in and settled by civil population. In the first half of 5th century the fortified settlement experienced a severe calamity, which combined the effects of a natural disaster (a powerful earthquake) and a devastating enemy attack, probably by the Huns. Restoration works were undertaken only by the end of 5th or in the first half of 6th century. The defenses were considerably modified and the walled area was totally rebuilt according to a completely new plan. Probably another assault imposed a new reconstruction in the second half of 6th century. The early Byzantine fortified settlement was destroyed during the Avar-Slav invasions in the end of 6th or in the early 7th century. A few centuries later a Bulgarian medieval settlement came into being above its ruins and survived until the Pecheneg invasions in the first half of 11th century.

The paper presents a thorough critical analysis of the available information on the ancient settlement under the modern village of Baykal, Pleven District, which was convincingly identified some 35 years ago with Palatium and Παλατίολον/Παλάστολον, known from the Late Roman and Early Byzantine written sources. The important
corrections in connection with the previously expressed views, as well as personal ground observations and reinterpretation of the evidence, enable the author to state a new and different perspective of the ancient history of Baykal.
This article is the first one of a planned series of publications designed to bring clarity and reject the existing defects in the current interpretation of various aspects related to the nature, structure and history of the Roman Danube Limes within the borders of Bulgaria.

The present day Danube border of Bulgaria constituted, at different times between the beginnings of 1st century AD to the last quarter of 7th century, an integral part of the northern border of the Roman Empire – the so-called Danube limes.
A close scrutiny of the written sources and epigraphic evidence shows that until 3rd century AD the term limes occurs rarely and when used it stands for ‘a military road’ or ‘land border’. Not a single occurrence, however, as much as even hints at limes referring to a set of defence facilities or a particular military and administrative unit. After 4th century the term came to be widely used. Nevertheless, it lacks fortification reference and was never used to mean a system of border fortifications. Rather, it stood for a particular area which comprised parts of a border province and fell within the jurisdiction of a special military commander with the title of dux (in some cases – comes). The term limes as used in the late antiquity has to be considered, therefore, to have military and administrative connotation in the sense of ‘a military zone’ which outlines the territorial scope of the competences of the specific dux (Торбатов 2002а, 435-436; Liușnea 2019, 101-114, for reference on recent research contributions to the discussion).
In any event, the term limes in the sense of a fortified defence frontier with fortification and support infrastructure constructed with care, of specific planning scheme and connected with a military road, is presently widely popular and has acquired axiomatic value in modern historiography.
Written records from the antiquity have preserved 54 toponyms, definitely connected with the infrastructure of the aforementioned section of the Danube limes: names of stations along the Danube road, legionary camps, support unit camps, fortlets, late Roman and early Byzantine fortified settlements of diverse legal status. It is assumed that some of those toponyms may refer to the same places which, for various reasons, were renamed at a later stage. It is certainly worth noting, that not even half of those toponyms have been localised unequivocally. Recent research shows that despite being widely accepted, the proposed identification stands in stark contrast to written evidence from the antiquity as well as to archaeological reality. 
Field work, on the other hand, as well as archaeological excavations, has identified the remains of over 70 fortified sites along the banks of the Danube. The overwhelming majority of those have been dated with either full certainty or a very high degree of likelihood to Roman, Late Roman or Early Byzantine eras. However, the available information on some is dated, incomplete or of questionable scientific value.  Furthermore, the dramatic hydrological and landscape changes in the Danubian riverine area over the past one hundred years, as well as the high variations and frequent changes in the local toponymical nomenclature often hamper the use of earlier data regarding specific sites.  In light of the above, the time has certainly come for the existing knowledge about the Bulgarian section of the Danube limes to be reconsidered and subjected to thorough field verification.
We believe that the productive approach to this huge challenge would be to come up with separate research and critical investigation of individual micro regions and groups of sites. Such body of research shall bring together the essence, the structure and the historical development of the Danube limes within Bulgaria. We hope this scholarly endeavour will be markedly devoid of the deficiencies in interpretation of a broad range of cross-cutting issues, typical of existing historiography.
The present work is based on two previously published studies (Торбатов 2016а; Торбатов 2016б), which in view of the time, have been revised and updated consistent with the concept of the intended scientific project.

REVIEWS Henrieta Todorova (†) (Hrsg.). durankulak, Band III. die hellenistische Befunde. verlag Marie Leidorf GmbH. rahden/Westf., 2016, 272 Seiten, 408 abbildungen, 3 Karten, 52 Tafeln The goal of this volume is to present the discoveries and finds from the Hellenistic period within the territory of the multi-layer archaeological complex Durankulak. According to the authors, they should be considered to be closely correlated with each other, testifying to the existence here between the end of the 4 th and the beginning of the 2 nd c. BC of a clearly defined territorium sacrum dedicated to the goddess Cybele. The investigations of the site began in 1975 and continued until 1996. A presumption about the existence on the Big Island of a cult complex associated with Cybele, which might have come into being as early as the 4 th c. BC, was suggested for the first time on the basis of the 1979 excavation results (Тодорова / Ботов 1980, 155). In 1983 graves were discovered on the western bank of the Durankulak Lake (Тодорова 1984, 20), and in 1988 a series of ritual pits of a similar date were also found in this area (Тодорова / Димов 1989, 15). Since 1991, the progressive increase in the amount of findings from the Hellenistic period, as well as their diversity and specifics, have made the study of Hellenistic remains in the area of the Durankulak archaeological complex a research priority. An important role for this was played by the scientific agreement with DAI – Berlin, on the basis of which Dr. J. Burow joined the research team as its representative. The successful completion of the excavations in 1996 was followed by enthusiastic work on the analysis and systematization of the finds and field observations relating to the Hellenistic period, which were gathered in the course of the investigations in Durankulak. Unfortunately, the premature death of Dr. J. Burow († 2001), who was appointed editor-in-chief of the future thorough publication, has delayed the appearance of the latter, confronting the members of the team with a number of difficult to solve problems. Prof. H. Todorova (director of the excavations in Durankulak from the very beginning), who took over the burden of editor-in-chief of the volume, spent a lot of time and did her best in seeking out, putting together and bringing in order the manuscripts by J. Burow, which turned out to be at rather different degree of readiness. Part of the chapters initially planned by him however, remained unwritten. A palliative solution to cope with the situation was found by the inclusion of brief or broader additions and comments prepared by other members of the team, as well as by invitation of external authors whose materials would contribute to clarifying and/or presenting in broader context of the discussed matters. Unfortunately, Prof. H. Todorova also did not live until the final collection and publishing of the volume († 2015). Perhaps, because of the objective circumstances or maybe due to the editorial views, Durankulak III in its present form is somewhat " non-standard " for a generic publication with a specific thematic focus: it looks much more like a collection of papers. For example, the sections devoted to the geoarchaeology, palaeo-geography and paleobotanics in the Durankulak region, placed at the