Roman period

7/21/2019 Roman period

1/60

VIII. THE ROMAN PERIOD


2/60


3/60

Historical outline of the Roman period| 205

HISTORICAL OUTLINE OF THE

ROMAN PERIODJen Fitz

THE NATURE OF RESEARCH

Theperiod between the 1st century and the 5th centuryA.D.brought a sharp break in the preceding, continuous develop-ment of Transdanubia. In the Roman period, Transdanubiabecame part of the vast empire ruling the Mediterranean thatabsorbed theoutstanding achievementsand cultures of antiq-uity and created a radically different cultural environmentfor the regions inhabitants within the framework of a well-organized state than ever before or for a long time after-

wards.The radical change is also reflected in the archaeological

record. Although earlier traditions survived for a long time

on the settlements of the native population, the new immi-grant population and especially the provincial governmentworked with technologies unknown in the past. The use ofcement fundamentally changed the nature and size of build-ings. Primitive sunken huts were replaced first by adobeand, later, by stone houses constructed on firm foundationsand provided with heating systems. Huge edifices wereerected, such as amphitheatres, theatres, palaces, temples,public baths, market halls, villas, military forts, bridges andaqueducts that carried water to the towns from distantsources, and many amenities of urban life were introduced:paved roads, sewers, water pipes, floor heating, etc. A net-

work of roads was constructed for the army, the postal ser-

vice and the commerce that reached even the remotest cor-ner of the vast empire. Consciously planned towns withpublic buildings, regular blocks of houses, town walls and

works of art replaced modest rural settlements.The archaeological record is not our only source of infor-

mation for the study of the Roman period. Many events, ca-tastrophes and changes in the provinces life are recordedand described in the historical sources, while the epigraphicmaterial contains information not only on the constructionand rebuilding or renovation of various structures, as well asthe name of their commissioners, but also offer an insightinto the beliefs and the life of the people, and allow the re-

construction of individual careers. Many government offi-cials are known by name, as are the important stages in theircareer: these include the successive proconsuls, the com-manders and officers of the army and their outstandingmilitary feats and the soldiers serving in the army, whosesocial and ethnic background can usually be reconstructed.

The huge number of coins that have survived is useful notonly for dating archaeological finds: when hidden as hoards,they indicate military invasions (many of which are men-tioned in the written sources, although some cannot belinked to a known event) and the size of the affected area.

The analysis of coins also reveals much about the prosperity

and decline of a particular region, while the designs on theirobverse are often an illustration of imperial policies and theprogramme of individual emperors. The changes in tradedcommodities are a good indication of the nature and thechanges in the economy, as well as of the needs, the taste andthe wealth of different social groups. Works of art, statues,reliefs, wall paintings, mosaics, the applied arts and the suc-

cessive artistic styles, revealing the aesthetic taste of the pe-riod, offer an insight into the activity of artists and work-shops, and reflect the survival of the legends and narratives ofantiquity. The richness and the diversity of the surviving evi-dence provides a wealth of detail about this period of the pastthat can hardly be compared to earlier ages or the centuriesfollowing the Roman rule in Pannonia.

PANNONIA

Together with the Vienna Basin, the Burgenland, the Dra-

vaSava interfluve in Slovenia and the northern areas ofBosnia, Transdanubia was part of an administrative unit, aprovince called Pannonia. The province was divided into twoparts (Lower and Upper Pannonia) for political and militaryreasons at the beginning of the 2nd century. Caracalla modi-fied the boundary between the two parts in 214; the province

was further subdivided into four administrative units at theturn of the 3rd and 4th centuries (Pannonia Prima and Se-cunda, Valeria and Savia). The borders of the province wereagain redrawn during the last reorganization, when Poeto-

vio/Ptuj was incorporated into Noricum Mediterraneum.With its mixed population of Celtic and Illyrian groups,Pannonia had never formed a geographical, historical or po-

litical unit in the periods before the Roman occupation. Theprovince was named after the Pannons when Illyricum, theformer large province, was carved up into smaller parts. Themajority of the Pannons lived in the southern part ofIllyricum (called Dalmatia after the division). As a matter offact, the Romans themselves did not regard Pannonia as aunit before they occupied it. This region was conquered infour phases between 35 and 4649, corresponding to themilitary policies of the Empire. The territory west of LakeBalaton was occupied at an early date, in 15, when Rome an-nexed the zone of a major trade route the so-called AmberRoad, leading from Italy to the Baltic to the Kingdom of

Noricum (lying in present-day Austria). The eastern part ofTransdanubia was occupied half a century later under Clau-dius, when the border of the Empire was extended to theDanube in the entire Danubian region.

ROMAN ADMINISTRATION IN PANNONIA

The Roman occupation followed the same successful sce-nario as the conquest of the small, central Italian city-states.

The conquest meant not only the presence of the army, butalso the irreversible occupation of the territory as well. The


4/60

206 The Roman period

native population was divided into tribal areas or districtscalledcivitates,first under the authority of the army and,later, of magistrates elected from their own ranks. A tribaldistrict usually corresponded to the already existing settle-ment territory of a conquered tribe. Larger tribes consid-ered to be unreliable were re-settled in different districts.

Military and political considerations often led to the major

disruption of earlier conditions. In the Roman period, theknown settlements and find assemblages of the native popu-lation show a rather dense distribution in the frontier zonealong the Danube and a looser one in the inland areas of theprovince, as for example around Lake Balaton. This nodoubt reflects a conscious settlement policy designed tostrengthen the frontier zone and reserve the more pro-tected inland areas for the large estates.

The creation of tribal districts for the native populationalso meant that the greater part of the new provinces terri-tory came into the ownership of the conquerors. The fortsand watchtowers of the army and the settlements beside

these forts were established in these territories. Pasturelandfor the mounted troops and other military areas, road sta-tions and custom houses were carved out from these areas.

The greater part of this area was populated according to aconscious settlement policy. New population groups ar-rived in the wake of the Roman conquest, a process that hadalready begun before the Roman occupation proper withthe settlement of Italian tradesmen and the representativesof Italian firms along the busy Amber Road. The number ofnew inhabitants who swelled the ranks of Pannonias pop-ulation as a result of this conscious settlement policy wasmuch higher than the number of individual immigrants. Inaddition to the immigrant groups who arrived from north-

ern Italy, the veteran soldiers discharged from the army sta-tioned in the province made up the largest part of this newpopulation. The veterans received generous grants of landon their retirement. They were settled on land allocated tothem near a military fort or in the area of major crossroads.

These settlements were granted the rank of a town as thenumber of their inhabitants grew and urbanization pro-gressed. The state also founded new towns, called coloniae(such as Savaria) for the retired legionary soldiers. The firstimmigrants were mostly Italians or arrivals from neigh-bouring provinces. After gaining Roman citizenship, thelatter adopted the Roman lifestyle and spoke Latin. Lured

to the Danube region by the prosperity at the beginning ofthe 3rd century, the number of Orientals Greeks, Syrians,Jews and Egyptians also rose, while the towns attractedthe native population with their promise of a better liveli-hood. The native population came to be the dominant ele-ment among the town-dwellers and the ordoof the townsfounded at later date, a tendency that became more striking

when a settlement was granted the rank ofmunicipium. Thepromotion to urban rank usually meant the dissolution ofthe earlier tribal districts, with the villages of the nativepopulation placed under the authority of the municipalcouncil.

The overall number of native population groups was lowin Inner Pannonia, as was the number of settlements thateventually grew into towns. Quite a few villas, residences oflarge estate owners decorated with frescos and mosaicfloors and surrounded by economic buildings, are knownfrom the area around Lake Fert, in the area north of LakeBalaton and in the Mecsek Hills (e.g. at Balca). The names

in the inscriptions suggest that their owners were families ofItalian origin.The lower echelons of society, the slaves and the freed-

men, are known from inscribed monuments and tomb-stones. Some of them were soldiers who had been takenprisoner in various wars, but most were slaves brought toPannonia from the East and from Africa. They were put to

work on the large estates, in commerce and in privatehouseholds. Bright and gifted slaves were liberated aftersome time or were able to buy their freedom. These freed-men often rose to a prominent position in local commerce,industry or in certain professional bodies. Many of these

former slaves were employed by the state in various eco-nomic offices and at the customs.The administration of a province was a relatively simple

affair. Power was wielded by the governor, the legate repre-senting the emperor who, in the first three centuries, wasalso commander of the provincial army. In the 4th century,the civilian and military administration was separated, theformer placed under apraeses,the latter under adux. Sepa-rate organizations headed byprocuratores,whose authorityusually extended over several provinces, were responsiblefor economic matters (the Danubian provinces formed asingle customs territory).

THE CHANGE IN LIFEWAYS

Following the Roman occupation, Pannonia was integratedinto a world empire. A modest craft industry catering to lo-cal demand (pottery, iron and bronze metallurgy) contin-ued its activity for a fairly long time in the regions inhabitedby the native population. The immigrants and the army,however, were supplied by freshly settled new craftsmenand the empires commercial network. The appearance ofan Italian lifestyle also brought a regular flow of wares andcommodities that had earlier arrived but sporadically to the

Danube region. Oil and wine imported from Italy and His-pania were part of the provisioning of towns and militaryforts. High quality pottery, the most common types of

which were the redsigillatawares, were first imported fromthe major Italian workshops and, later, from southern andcentral Gaul, as well as from Africa in the imperial period(terra sigillata chiara). Bronze vessels were imported fromItalian, Gaulish, Balkanic, Anatolian and various eastern

workshops. Local workshops later began the production ofwares in the Roman taste: for example,sigillatawares werealso produced in local workshops, such as the ones in

Aquincum and Gorsium. Household pottery and simpler


5/60

Historical outline of the Roman period| 207

articles were mass-produced on the industrial settlementsthat grew up beside towns and military forts. The spread ofthe Roman lifestyle also brought a demand for artwork.

The interior furnishing of public buildings, the palaces ofthe proconsuls, the villas in the centres of large estates, thetemples and the public baths conformed to the general stan-dard throughout the empire. The decoration of these build-

ings, the colourful wall paintings, the mosaic floors, thestatues in the sanctuaries and home shrines were rarelymade in local workshops. Stone carving was the single ex-ception: these were produced in local workshops. Of thetombstones erected over graves a custom that becamecommon in the Roman period only a few can be linked tostone carvers or workshops in Italy, Noricum or the East.

PANNONIAS ROLE IN THE ROMAN EMPIRE

The significance of Pannonia for Rome was primarily deter-

mined by Italys defence. While the Alps provided a naturalprotection for the central areas of the Roman Empire in thewest and the north, it could be easily approached by a hostilearmy marching along the Amber Road. This threat was noidle speculation: from Augustus time, the Romans regardedthe Marcomanni and the Quads, Germanic peoples who hadfor long centuries settled in the Bohemian Basin and UpperHungary, their most dangerous Barbarian enemies. Theforts of thelimes,the defence line built along the Danube,

were from the early 2nd century garrisoned by the troops ofone of the Empires strongest armies that repeatedly played adecisive role not only in the wars fought against the peoplesbeyond the borders, but as a large army stationed close to

Rome also in the civil wars of 69, 193197 and the decadesafter 249. Aside from the few civil wars, the centuries of Ro-man rule brought a long period of peace and prosperity last-ing for many centuries for the inhabitants of the inland prov-inces of the empire. This was not the case in the border prov-inces, such as Pannonia. The Germanic and Sarmatianpeoples on the left side of the Danube lived under consider-ably worse circumstances than their contemporaries on theother side of the river, and the prosperity of the province wasa constant temptation to stage raids and looting expeditionseven in times of peace. The army stationed in the provincefought the first war with the neighbouring peoples and the

Dacian state between 86 and 106. Trajans victory on bothfronts brought a period of peace to the Carpathian Basin.The next, even more taxing war, the so-called MarcomannicWar, in which practically all the peoples of the Danubian re-gion joined forces against the Empire, was fought by Romeunder Marcus Aurelius reign between 167 and 180. One in-dication of severity of this war was that the Sarmatians wereforced to set free a hundred thousand prisoners of war whenpeace was concluded. The next ordeal in the life of the Ro-man army stationed in the Danube region began in the mid-dle decades of the 3rd century. Under the pressure of easternpopulation groups, the Goths who had migrated to the

Ukraine were pushed towards the Danube delta in the 230sand thus set in motion the successive population movementsof peoples dislodged from their homeland. Fought with

varying success until the Goths were finally exhausted, thiswar eventually forced Rome to surrender Dacia in 270. Al-though these wars rarely affected Pannonia until Dacia stillexisted, the province suffered the greatest catastrophe in its

life during this period, when the pretender Regalianus armycontrolling the Danube region was dealt a crushing defeat in260 by the Sarmatian Roxolani who had earlier settled in theDanubeTisza interfluve. The numerous coin hoards fromthese years and the extensive destruction layers observed onmost settlements suggest that the greater part of Pannonia

was plundered and that many towns and military forts werecompletely destroyed.

PANNONIA IN ANTIQUITY

Pannonia never enjoyed a peaceful and continuous develop-ment similar to the flourishing provinces and towns of theempires inland provinces owing to its military role, the great

wars and the devastation brought by recurring raids.Pannonia differed from the more fortunate provinces of theempire not only because of the repeated need to start lifeafresh. The consolidation of the province, a process lastingfor some eighty years, came at a time when Italys impor-tance began to wane within the empire. Even though theItalian element was strong in the western part of the prov-ince, their presence, culture and taste nevergained a footholdin those areas of the province that were conquered later. Adegree of Romanization, disseminated by the army, can be

demonstrated in these areas also. At the same time, theRomanized population included increasingly less landown-ers, traders and craftsmen of Italian origin. The majority ofthe Romanized population was made up of soldiers who hadsettled in the towns after their retirement from the army.

Most of these soldiers had been recruited from the nativepopulation living in the shelter of the chain of forts and theyonly acquired a superficial Romanization during theirtwenty-four years of military service. The young men re-cruited from the peoples living in other parts of the empire

were no different from the average native of Pannonia whousually married a girl from a neighbouring village. A uniform

peasant-military society proud of its military prowess, butcharacterized by a superficial Romanization, emerged in thefrontier zone. Except for the belt of the Amber Road andcertain areas in the DravaSava Interfluve, the towns lyingfarther from the border remained insignificant in the lack offresh Italian immigrants. Even setting up a municipal councilin these towns often proved quite difficult. The historian

Ammianus Marcellinus noted that Valentinian, his contem-porary, who had been born in Pannonia, did not speak Greek considered the sign of a cultivated mind and that he evenhad difficulties with Latin. It is therefore hardly surprisingthat not one single renowned artist or scholar of antiquity


6/60


came from Pannonia; in contrast, the inscriptions and biog-raphies have preserved the names of countless excellent sol-diers and outstanding military feats.

The moderate level of Romanization in Pannonia can inpart be associated with the economy of the province. It wasnot a particularly rich province, lacking major industrial cen-tres and large enterprises whose products were exported to

distant territories. This was also reflected in the provincessocial make-up. There were fewer representatives of the twoleading social classes of the empire, the senators and theEquestrian Order (who played a prominent role in the econ-omy) in the entire province than in a larger town of Gaul,Hispania or Africa. The first Pannonian individual to join theranks of the senators was a certain M. Valerius Maximianus,an excellent soldier and military leader in the Marcomannic

wars, who killed the king of the Naristi with his own hands.

CHANGES IN THE 4TH CENTURY

The great wars of the 3rd century that brought destructionto the Danubian region and the front on the Rhine, as wellas the Tigris and the Euphrates region, coincided with amajor crisis of the empire. The precarious economic bal-ance of the 2nd century was shattered by the Marcomannic

wars, by Commodus reckless overspending, the five yearsof civil war between 193 and 197 and the significant raise ofthe soldiers pay under Caracalla. A rapid inflation can benoted in the 3rd century, reflected by the disappearance ofsilver coins and the debasement of bronze coins. The un-successful military campaigns led to successive civil warsthat disrupted the administration of the empire for decades.

The government of the Augustan period that had in essenceadjusted the administration of the Republic to the needs ofa world empire while preserving the local governments ofthe towns, had became ineffective and could no longer bemaintained. Diocletian introduced a series of radical mea-sures that changed the administrative system and the divi-sion of power, leading to the abolishment of all forms of in-dependence and the creation of a rigid central administra-tion. Another major change in this period was the victory ofChristianity that is also reflected in the archaeological re-cord, for example in the temples, the burials, wall paintings,mosaic floors, symbols, etc. Lying between the western and

eastern half of the empire, Pannonia lost the former mili-tary and political significance it had enjoyed in the 3rd cen-tury and in consequence of the weakening of the westernpart of the empire, the province proved unable to check thesuccessive waves of population movements and migrations.

QUESTIONS OF SURVIVAL

Roman rule in the Carpathian Basin ceased in the early de-cades of the 5th century. The border forts were unable tohold back raiding groups and the mobile field army was

transferred to Dalmatia. In accordance with the terms of anagreement, in 430 Rome ceded the eastern part of Pannonia(Valeria province) to the Huns. The written sources con-taining a wealth of detail on the Roman period in this re-gion again fall silent. The archaeological finds and theproducts of local workshops rarely enable an ethnic attribu-tion. Even assuming that the greater part of the population

left the province for westerly and southwesterly regions, wecannot speak of the absolute cessation of Roman life. A sur-vival was possible in the forts enclosed by walls and in thetowns of Valeria (the most likely candidates being Sopia-nae/Pcs and Herculia/Szabadbattyn). West of Lake Ba-laton down to Slovenia, where the toponyms of the Romanperiod survived until the Hungarian Conquest period (e.g.Sala/Zala and Arraba/Rba), the flourishing of the so-calledKeszthely culture indicates the survival of the population ofthe Roman period, while the name Valeria Media suggeststhe unbroken existence of the population and its assimila-tion into the freshly arrived Slavs. Judging from the council

held here during the Carolinigan period, we may also as-sume the survival of Christianity.

THE BORDER DEFENCE

OF PANNONIAZsolt Visy

The study of the border defence of Pannonia and of its ar-chaeological remains and history has been one of the mainconcerns of Hungarian archaeology for generations. Theinitial enthusiasm sparked by investigations in other areas

of the one-time Roman Empire conducted in the early 20thcentury was followed by debates, rather than actual researchprojects. The situation did not change later and there werealways scholars who regarded the advances made in thestudy of the limesthe yardstick of the archaeological re-search of the Roman period in Hungary. The initial dreamof a research project comparable to thelimesstudies in Ger-many or Austria and the regular publication of findings inthis field of research have still not been realized. The regu-larlimescongresses, especially the one held in Hungary in1976, gave a fresh impetus to this research, as did the majorconstruction projects in the cities that now overlie the set-

tlements along the one-time Pannonian border and theplans of the Danube dam that called for large-scale rescueexcavations before the construction was begun. These exca-

vations yielded a wealth of new finds and information; theirpublication will no doubt modify our current knowledge.Even so, there still remain a number of areas, where re-search has barely started: these include the Pannonianlimessection south of Paks and other areas, where our knowledgeof the network of the military stations along the limes,theripa Pannonicais rather incomplete.

Two legionary forts and many auxiliarycastellalie in theHungarian territory of Pannonia(Fig. 1). It is impossible to


7/60

The border defence of Pannonia| 209

precisely determine the number ofcastellasince many of theones lying in the inland areas of the province or along theripahave not been identified yet. The discovery of new auxil-iary camps can be expected at TokodErzsbet-akna, Blcs-keSzentandrs-puszta and in the Szekszrd area. Moreover,some of the known forts have not been excavated yet; in

other cases, the findings of a few already excavated forts arerather controversial and it is impossible to determine whichof these forts had been used continuously, which had beenfounded at a later date and which had been abandoned ear-lier, even if only temporarily. A number of new forts wereconstructed in the late Roman period, some of which fol-lowed the traditional groundplan, while others, built onmountains peaks and atop steep hills, were adjusted to the lo-cal terrain. The latter form a rather dense chain compared tothe preceding period, especially in the Danube Bend. As a re-sult of more recent investigations, the number of known

watchtowers has nearlydoubled and the limesroad can some-

times be followed along 3040 km long stretches.

FORTIFICATIONS

The fortifications system created along the ripain the 1stand 2nd centuries was the result of a more or less linear oc-cupation. The early camps of the legions and the auxiliarytroops that were later rebuilt into proper forts all lie alongthe Danube, the frontier river of the province and, also, ofthe Roman Empire. This linear border defence system,conforming to the general defence policy of the Roman

Empire, was obviously adapted to local conditions and thestrategic necessitates. The river border, theripa,provided arationale for the genuine military occupation of the fron-tier. This linear defence system remained unchanged even

when Pannonia was carved up, first into two and, later,three frontier provinces.

The local terrain was taken into consideration both inthe organization of the defence system on a provincial leveland in the siting of individual forts. The diagonal roadsleading to the Danube rarely stopped at the river, but con-tinued into the Barbaricum on the other side. The riversflowing into the Danube were natural and excellent trans-port facilities on both sides of the river. It has since longbeen noted that the army troops were at first concentratedin these locations along the Danube and that the earlycamps were established by the major crossing places.

Even though there is little archaeological evidence in thisrespect, we know that in Augustus and Tiberius time troops

were stationed in the region of Carnuntum and near themouth of the Sava by the Danube. It is unclear whether thesewere permanent camps or temporary deployments linked toa particular campaign or diplomatic manoeuvring (69 and17). The first legionary camp of Carnuntum was built inClaudiuss time and the earliest auxiliary camps at Arrabona,Brigetio, BudapestVizivros, Lussonium and Lugio toodate from the same period. It is also quite certain that severalauxiliary troops were stationed near the mouth of the Dravaand the Sava in the region of Mursa and Sirmium during thisperiod, even though there are no archaeological finds to con-firms this. Moreover, in the case of Vetus Salina (whose

Fig. 1. Legionary camps. 1. Aquincum (buda), 2. Brigetio (Szny). There is a late Roman fort on the eastern side of the Aquincum camp and asimilar fort can be assumed on the northern side of the Brigetio camp


8/60


foundation date is uncertain), we can hardly speak of theDanubian terminal of an important road.

Most of the auxiliary troops were transferred to theDanubianlimesunder Vespasian and Domitians reign. Theearliest legionary camp of Aquincum and a number ofcamps for auxiliary troops were also built at this time(Fig.2). However, in many cases very little is known about these

forts owing to gaps in the archaeological record and becausethe finds from the forts that have already been investigatedare unsuitable for resolving problems of chronology.

Troops were installed in Solva, Cirpi, Aquincum and prob-ably in Intercisa along the Hungarian border section under

Vespasian; the first auxiliary camps at Albertfalva andAdony were constructed at the same time, if not earlier.The camp at Campona may have been founded in Do-mitians time, but if so, the later stone fort was not builtover the earlier palisade fort since the stockade uncoveredunder the former can be dated to Trajans reign. The

Aquincum legionary camp was established in Domitians

time, in 89. A number of problems concerning the campsestablished under Domitian still need to be clarified: one ofthese is that while the epigraphic evidence indicates a signif-icant rise in the number of auxiliary troops during the 80s,this is not reflected in the foundation of new camps.

The full and permanent occupation of the frontier zonewas completed under Trajan. The legionary camp at Brige-

Fig. 2. Aquincum, the southern gate of the legionary fort,from the west

Fig. 3. Auxiliary camps: 1. Quadrata (Bartfldpuszta), 2. Ad Statuas (csVaspuszta), 3. Azaum (Almsfzit), 5. Ulcisia Castra(Szentendre), 6. Matrica (Szzhalombatta); late Roman forts: 4. Tokod, 8. Lussonium (Dunakmld)


9/60


tio was founded during his reign, sometime around 100.

The other known auxiliarycastellatoo were built during thisperiod, suggesting that the border defence system of thetwo Pannonias that remained virtually unchanged for twocenturies had been established by the beginning of Had-rians reign at the latest. The new auxiliary forts dating to

Trajans time were Ad Flexum, Quadrata, Ad Statuas, Aza-um, Ulcisia Castra, Campona if there was no earlier pali-sade fort here and Matrica(Fig. 3).

In some cases the fort was relocated. It is also quite likelythat a number of early palisade forts that were abandonedand replaced by another fort in its vicinity or somewhat far-ther will be discovered in the future. In the case of Matrica

we cannot speak about the relocation of the fort, as was as-

sumed earlier, while this seems to have been the case atIntercisa. A ditch with a V shaped section, suggesting theexistence of an early fort, was found south of thecastellum. Asimilar relocation of the camp also seems likely at Lusso-nium.

The recently discovered auxiliary fort near Srszentgotawas also a palisade fort that, judging from its size, was builtby acohors. Although there is nothing either in the archaeo-logical record, or in the epigraphic material to confirm this,its location and distance from the Danube suggest that itcan be dated to the mid-1st century, i.e. the Flavian period.

The same uncertainty surrounds the fort at Tokod

Erzsbet-akna. Although there is no archaeological prooffor an early Roman camp, Italian sigillata fragments andother finds characteristic of early Roman military camps

were found together with a rich assemblage of wares pro-duced by the native population, a stamped brick of theala IBritannicaand the military diploma of a soldier serving intheala Frontonianaissued in 110. These finds indicate thepresence of an early fort in the area.

The auxiliary fort near Bem Square in Budapest isslightly different. Recent excavations have uncovered Ro-man buildings with several occupation levels; the early

wooden constructions can be assigned to the mid-1st cen-

tury on the basis of Italian sigillata finds. It is unclear

whether the later rebuilding of these early structures, sug-gesting the presence of a military fort, served military orcivil purposes.

The military forts discussed above were, with few excep-tions, all palisade forts. The structure of the palisade couldbe reconstructed in some cases. In cases when there was noevidence for posts arranged into one or two rows, theearthen bank was most likely buttressed with mud bricks.Palisade forts were built even after Trajans reign.

Several new forts have been identified in the Brigetioarea, many of them from the aerial photos made by OttoBraasch. Some of these lie in Pannonia and some in theBarbaricum on the other side of the Danube(Fig. 4). A total

of eighteen temporary earthen forts (marching camps andperhaps practice camps) have been identified to date andthe line of a ditch enclosing a Roman camp was discoverednear Horvtkimle as well. Although the date of these campsremains uncertain until the evaluation of the finds collectedduring field surveys and excavations, the findings ofSlovakian excavations suggest that the majority of the fortsat Brigetio were founded at the time of the Marcomannic

wars.While the legionary forts built along the Pannonianlimes

section lying in Hungary were stone constructions by theend of the 1st century, the auxiliary forts were not. It is pos-

sible that the rebuilding of the auxiliary camps of Solva,Ulcisia Castra, Albertfalva and Vetus Salina in stone can bedated to Trajans reign. In many cases, this rebuilding wasonly begun under Hadrian; many of these reconstructionscame to a standstill during the Marcomannic wars and wereonly continued or finished under Commodus.

The castellum of Intercisa was rebuilt in stone inCommodus time. The southern gate was protected by asimple wooden tower and the walls of the fort extendedonly to the gate itself. The two gate towers were built laterand a section of the wall was pulled down to make place forthe towers. The bedding trenches survived inside the

Fig. 4. Brigetio, marching camp Fig. 5. Intercisa, the southern gate of the auxiliary camp, during theexcavation in 1975


10/60


towers. The towers were rebuilt in stone at a later date,probably in Caracallas time, but most certainly before Gor-

dianus IIIs reign(Fig. 5). The same could be observed atthe fort of Quadrata.

The fort system set up by the middle of the Roman pe-riod existed for about two hundred years. The auxiliaryforts were distributed more or less evenly and they weresuitable for defending the river border of the two Pan-nonias. The average distance between two forts was aboutten miles or, occasionally, fifteen miles when the Danubeforked into several branches or was flanked by marshland,as for example along the Moson branch of the river and inthe Srrt region. The largest distance, 32 miles, was mea-sured between Solva and Cirpi. This distance can be attrib-

uted to the mountainous terrain.The late Roman fort system can only be understood inknowledge of the late Roman military reform. Begun underDiocletian and completed under Constantine, the reformultimately meant the division of the army into central mili-tary troops and the provincial armies assigned to defend thefrontier and, also, a reform of thelimitaneusarmy by raisingnew troops. The reforms also affected the already existinglegions and auxiliary troops: these could keep their names,but their organization changed and their number decreasedsignificantly. These changes influenced the fort system as

well. The genuine threat of enemy incursions called for the

reorganization of the linear defence system into an in-depthone, in which the fortified towns and their militia could bemobilized in times of danger, played an increasingly impor-tant role beside the military bases. The forts along the bor-der were reconstructed as part of the reform and this, inturn, had two consequences. One was the adjustment ofthese forts to a defensive role and to making them capable

of withstanding sieges, calling for the reinforcement of thefort walls and, even more important, the construction ofprojecting angle and interval towers that also meant that theearlier ditch near the wall was filled up and another one wasdug farther away. The other was the rebuilding of the fortand its barracks to accommodate a garrison of fewer troopstogether with the civilian population. The most salient re-flection of these changes was the demolition of theaggeronthe inner sides of the defence walls to make place for newbuildings built against the inner face of the wall.

The chronology of these late Roman constructions is un-certain in many cases. The existing forts were renovated

and rebuilt to some extent under Diocletian, Constantine Iand Constantine II, and the determination of their exactdate and architectural-typological features calls for furtherstudies. The last major military constructions can be datedto Valentinians reign. The replacement of the horseshoeand fan shaped towers of the inner Pannonian forts withlarge round bastions begun shortly before his death has notbeen observed in the Pannonian provinces, suggesting aradical drop in the funds spent on the forts of theripa Pan-nonicaand, consequently, thelimitaneusunits.

No late Roman constructions and reconstructions havebeen observed in the two legionary forts of the province.

The rebuilding of theporta praetoriaduring the Tetrarchy

was the last construction project in the Aquincum fort. Thefort was soon abandoned and a new fort was built on theeastern side in Constantines time. Since no fan shaped tow-ers were built in the legionary fort of Brigetio either, it isquite possible that the same happened at this fort also.

Almost all of the earlier existing forts that have been ex-cavated yielded evidence for late Roman construction work.

The only exception is the Albertfalva fort that was evacu-ated during the 3rd century crisis or shortly afterwards. Thelarge rebuildings in the excavated forts have been generallydated to the earlier 4th century, to Constantines reign; thisdating seems acceptable, even though conclusive proof for

it is often lacking. At Intercisa, for example, the rebuildingwas carried out under Constantine II or Valentinian I, al-though the latter date is hypothetical. The coins found inthe fan shaped towers at Quadrata and Ad Statuas suggestthat they were built before the 350s.

New forts were also built in the late Roman period(Fig.6). Some of these were standard rectangular constructionsconforming to the traditional layout, such as the ones at

Mrcius 15 Square in Budapest and at Tokod built underDiocletian or Constantine, while others were built on hill-tops and the adjustment to the local terrain meant that theirgroundplan diverged from the standard one. These forts

Fig. 6. Late Roman small forts: 1. Lussonium (Dunakmld),2. Cirpi (Dunabogdny), 3. Azaum (Almsfzit), 4. VisegrdGizella telep, 5. Solva 19 (PilismartMalompatak)


11/60


indicate a consistent adherence to the imperial defensivepolicy and the practice of adjusting the overall defence-

works to the terrain in order to repel hostile attacks. Theprecise date of these forts is in many cases unclear. Asidefrom Pilismart that can be identified with Castra adHerculem and dated to the Tetrarchy, these forts includ-ing the one at Lussonium can be broadly dated to the

reign of Constantine, Constantius II and Valentinian.The last construction period of theripa Pannonicasawthe erection of 1030 m large, tower-like forts, either sep-arately or on the territory of already exiting forts. This

would suggest that at the time of the construction works,the garrisons stationed in these forts were unable to de-fend the forts enclosed by several hundred metres long

walls. The 40 m long and 3.2 m wide foundation wall un-covered at Arrabona was probably the almost completelongitudinal wall of a small late Roman fort. While thepresence of a fort of this type can only be assumed at

Arrabona, its existence has been proven at Lussonium. A

10 m by 9 m large small fort with 2.3 m thick walls wasfound in the interior of the late Roman fort, a few metresfrom the southern gate.

According to Sndor Soproni, these forts were built afterthe Battle of Hadrianopolis, in a rather critical situation

when the mobile field army, after suffering grave losses, wasreplenished fromlimitaneusunits; since the troops stationedalong the borders could not be brought up to strength, theforts were adjusted to the considerably smaller garrisontroops. The excavations at Intercisa and Lussonium indi-cate that the small forts were not built immediately after

Valens death, but at the turn of the 5th century.

WATCHTOWERS

The research of watchtowers is perhaps the best indicationof the state of Pannonianlimesstudies. The identification oflarge military forts is considerably easier than the detectionof these rather small constructions, of which little survives ifthey were built from wood. The excavation of watchtowersthus often lags far behind the investigation of larger fortifi-cations. Even in areas where the research ofcastellais wellunderway, virtually nothing is known about the network of

watchtowers. We know that the efficient defence of the em-

pires borders was a major priority of the Roman militarypolicy from the very start. In his description of Lentuluscampaign on the Lower Danube, the historian Florus men-tionscitra praesidia constituta;however, it is uncertain whattype of military outpost he meant. Obviously, the control oftheripa,the frontier, was a major concern from the very be-ginning of the military occupation under Tiberius and,later, under Claudius reign, even in times when relations

with the peoples on the other bank were fairly good. This isprimarily reflected in the construction of auxiliary forts,rather than watchtowers. The defence of the river borders

was probably organized around regular patrols at that time,

rather than sentries stationed in watchtowers. Trajans col-umn, however, suggests that the chain of wooden watch-towers along the borders became common by the end of the1st century.

The study of older and more recent aerial photos, com-bined with field surveys and excavations, brought major

advances in the research of watchtowers. The number ofknown watchtowers has almost doubled in the past twentyyears and new ones have been identified in areas whereonly one or two such military installations were knownearlier. These areas include thelimessection at Arrabonaand the territory south of Annamatia. In contrast, few new

watchtowers have been discovered between Szekszrd andHungarys current border. Still, it is quite certain that in-tensive research in this field will bring the discovery ofnew sites.

Finds dating to the 1st century have so far only been re-covered from the ditch of the Solva 11a watchtower. The

Fig. 7. Watchtowers: 1. Intercisa 6 (Kisapostag), 2. Arrabona 9(Gny), 3. Gerulata 4 (Bezenye), 4. Azaum 3 (Neszmly),5. Ulcisia Castra 1 (Szentendre), 6. Ulcisia Castra 2 (Budakalsz),7. Lugio 1 (Dunafalva), 8. Intercisa 10 (Kisapostag)


12/60


tower itself perished completely, and thus its size and struc-ture could not be reconstructed. Only its ditch (having a di-ameter of 30 m) could be unearthed. Judging from astamped brick of theala I Britannicafrom the 1st century,the Azaum 1aburguscan perhaps be dated to the same pe-riod, although the site has not been excavated yet.

The number of watchtowers from the 2nd century is not

much higher. The Crumerum 1 burguscan be mentionedfrom among the earlier ones since its finds suggest that itwas used from this time, as was theburgusidentified in thePilismart area, indicated by the remains of wooden towersand a ditch. Some of theseburgihad a stone tower whose

walls measured up to 16 m (Solva 20), although this was notthe standard length in the light of more recently unearthedor discovered watchtowers from the 2nd century.

Theburgifrom the Commodus period mentioned in thenearly identically worded building inscriptions are regularlyquoted in Roman period studies. It must be noted, however,that these inscriptions mention not onlyburgiproper, but

alsopraesidia,suggesting that they were two different typesof structures. Since these were mentioned in the context ofthe secret activities of the latrunculi, it seems likely thatthese buildings were erected in places that were deemedsuitable for their arrest. The towers built on the elevation atIntercisa (Intercisa 11, 12, 17 and 13) may have been besuchburgi.

Only the late Roman towers or, to be more precise, theones from the Valentinian period are suitable for a more de-tailed analysis. The watchtowers in the Danube Bend wereall built of stone, while the ones along the southern limessection in the Intercisa and Annamatia area were woodenstructures. Their walls were 10 m long and they were usu-

ally enclosed by a quadrangular ditch, although double rect-angular ditches have also been found in the Intercisa areaand elsewhere, as have circular ditches on some sites.

Palisades have only been found in association with thewatchtowers of the Valentinian period. These were usu-ally constructed on the inner side of the ditch and theyalso functioned as a buttress for theagger. They occasion-ally ran along the outer side of the ditch, as at theCrumerum 2burgus. Depending of the size of theburgus,the roof of the upper floor was supported by posts. Onepost was uncovered in the small Solva 19 fort, while fourposts were used in the larger burgiof Solva 23a, Cirpi 2

and Ulcisia Castra 2(Fig. 7).The identical size of the watchtowers from the Valen-tinian period is rather conspicuous, suggesting that they

were built according to a standard design. Since this uni-formity can also be observed in the case of the ditches andthe palisades, it is therefore instructive to include them inthe discussion of theburgi. The axial length of the enclo-sure ditches was most often 2526 m or its double, about52 m, corresponding to 100 and 200 feet (whose mostcommon unit was roughly 27 cm). The same can be said ofthe wall thickness that measured 45 feet or, occasionally,6 feet. The ditch of the Cirpi 2 burgusenclosed a 32.5 m

wide area, corresponding to 120 feet. The same unifor-mity in size can also be noted among the towers enclosedby double ditches. The inner ditches also had an axiallength of 2528 m.

The towers enclosed by lozenge shaped ditches form aseparate group that can most likely be dated to the Tet-rarchy, at least on the evidence from the Intercisa 10 tower.

Another group is made up of towers with double rectangu-lar trenches, probably from the Valentinian period, that canmostly be found along thelimessection between Intercisaand Lussonium, although similar towers can also be as-sumed at a few other sites. This tower type is conspicuouslyabsent from thelimessection between Solva and Aquincum,

while the tower type characterizing this section is probablyabsent farther south. It seems likely that there were differ-ent military districts in these two areas of Valeria.

A distinction must be drawn between watchtowers andsignal towers. They cannot be distinguished from eachother in areas where thelimesroad runs directly along the

Danube as, for example, in the Danube Bend. Elsewhere,however, the watchtowers erected on the Danube bank,on the edge of higher plateaus and by gullies that func-tioned as part of the Danubian defence system can beclearly distinguished from the signal posts built on the in-ner side of thelimesroad running at some distance fromthe river. In many cases, there are no or only minimal ty-pological differences between the two, as at Pilismart,

where the line of thelimesroad coincides with that of themodern road, and in the Ercsi, Rcalms, Kisapostag andBta area, where thelimesroad ran farther from the Dan-ube. In these areas a second chain of watchtowers wasbuilt near the riverbank.

The chain of bridgeheads along the Danube is a charac-teristic feature of the late Roman border defence system.Comparable structures are known from the Rhine region as

well. A total of fourteen bridgeheads are known or assumedalong the limes section in Pannonia Prima and Valeria,

while only the bridgehead at Bcs, lying slightly fartherfrom the Danube, is known from the border section inPannonia Secunda. Sndor Soproni noted that bridgeheads

were built on both sides of the Danube and that they usuallyoccured in pairs. He believed that their construction anduse fell into the same period (between 324 and 378) as theDevils Dyke, the large rampart system in the Great Hun-

garian Plain. The Romans regarded the several kilometreswide zone on the other side of frontier rivers, such as theDanube, as part of the empires territory. The bridgeheadsthat functioned both as military bases and supply depots

were built in accordance with the terms of the treatiessigned with the neighbouring peoples and in keeping withthe general military reform in the 4th century, and theycannot therefore be directly associated with the DevilsDyke in the Great Hungarian Plain. It is nonetheless con-spicuous that the highest number of these bridgeheads canbe found along Valerias border, in the northern part of thearea enclosed by the rampart system.


13/60


14/60


15/60

Late Roman forts in Pannonia| 217

Hill and the coins found in two features suggest that theconstruction of the Sgvr fort was begun under Constan-tine the Great (312337), perhaps at the end of his reign. Inphase I, the fort walls had U shaped and fan shaped towers,the wall thickness conformed to the standard 150 cm. AtSgvr, the towers had rectangular foundations, althoughsimilarly to the towers of the Tc fort, the corners of the

vertical walls were probably rounded. The rebuilding(phase II) also affected the defence works. The walls of thefort were rebuilt with a width of 200 cm, 250 cm and270 cm, and round bastions with diameters of 1415 m anda wall thickness of 200270 cm were added at this time.

This rebuilding can be dated to the end of Valentinian Isreign (364375) or slightly later, as suggested by the chro-nology of the Fenkpuszta and Alshetny forts(Fig. 10).

The Tc fort was not rebuilt. It is unclear whether thisshould be taken to indicate that its military use came to anend. The Fenkpuszta fort with its round bastions was builtat a later date than the other ones.

The excavations also yielded a wealth of information onthe diverse range of economic activities pursued in theseforts. Bread was baked in sunken ovens from the cerealsstored in thehorrea(Fenkpuszta, Sgvr and Alshetny).Iron smithies and the production of agricultural tools areindicated by the pig iron and tools found in the forts(Fenkpuszta and Alshetny). Evidence for animal slaugh-tering and butchering is provided by refuse pits filled withanimal bones unearthed at Fenkpuszta and the meat smok-ing places discovered south of the Alshetny fort. The in-land forts were supply bases, established as part of the lateRoman military reforms. They provisioned and quarteredthe troops stationed along the limesand, if necessary, the

mobile field army (thecomitatensestroops) when it marchedthrough the province. Agricultural tools and implementsfor the population in the area were also produced in theseforts. The 300400 strong infantry units stationed in theforts also functioned as a frontier garrisons.

Vegetius, a military writer of the late Roman period, cau-

tioned that camps should not be set up near higher hillssince the enemy might capture them (EpitomaI. 22). TheBarbarian armies invading the province during the Sarma-tianQuadic incursion of 374 occupied eastern Transdanu-bia for months, indicating the need for strengthening theseforts. Besides thickening the walls of the defence works, a

watchtower was built on Tmlc Hill overlooking the Sg-vr fort(Fig. 11). The tower had a unique structure, un-known elsewhere in Pannonia. It measured 12 m by 12 m,similarly to the watchtowers along thelimes. The tower hada foundation of solid stone and a wall thickness of 5 m onthe ground floor. The Fenkpuszta fort was practically in-accessible because of Lake Balaton and the surrounding

marshland, reflecting the importance of defensibility in itssiting. It would appear that the forts were rebuilt after thebitter experiences of the SarmatianQuadic incursion of374 to provide an even more efficient defence.

About six hundred tombstone fragments, most of themcarved from marble, an altar fragment and the fragment of alarger than life marble emperor statue(Fig. 12)were recov-ered from the foundation walls of the Alshetny fort. Theuniform style of the tombstones can be associated with thestonemasons workshops active in western Transdanubiaand southeastern Pannonia. Their dates range from theclose of the 1st century to the end of the Severan period.

Since no settlement granted municipal rank is known fromsouthern Transdanubia, it is possible that these carvingscame from the cemetery of amunicipium,whose location re-mains unknown for the time being. This town was probablyabandoned by the 4th century and the tombstones from itscemetery were used as building material.

The inland forts reflect an ambitious, carefully plannedand rational construction project in Valeria and PannoniaPrima the 4th century. More building material was used forthe construction of the bastions of the Alshetny fort thanfor the replacement of the earlier wooden watchtowers withstone ones along the entire Transdanubian limes in the

Fig. 11. Reconstruction of the watchtower at Sgvr

Fig. 10. The round bastions of the Alshetny fort


16/60


Valentinian period. The large military fortifications of the

province were identical with the 4th century military fortsin other parts of the Roman Empire and were built accord-ing to the same design. The overall concept and plan ofthese construction projects and, more importantly, their ex-ecution called for significant public work, directed and su-pervised by the military. The forts functioned as supplybases. Thecohortesgarrisoning these forts were unsuited topartaking in the provinces defence. These inland forts canthus be regarded as military installations of the late imperialperiod controlled by the army.

These impressive constructions withstood the ravages oftime for long centuries. In the Middle Ages, they were usedas convenient stone quarries, the only exception being the

Fenkpuszta fort that was spared owing to its favourable lo-cation. The medieval chronicler, Simon de Kza was famil-iar with the ruins of several Roman buildings in northeast-ern Transdanubia: he associated these buildings with theHuns and various events of the Hungarian Conquest. Ac-cording to his chronicle, Svatopluk was defeated in an an-cient town beside Bnhida, whose remains could still beseen the town in question was the Krnye fort.

VINCEN-TIA

Krnye

QUADRI-BURGIUM

Sgvr

IOVIAAlshetny

?Fenkpuszta

Size of the fort 16 ha 7.3 ha 21 ha 7.8 haThickness of thefort wall

230300 cm 280 cm 150160 cm 230250 cm

Diameter of theinterval towers

1580 cm 1300 cm 1350 cm 1400 cm

Wall thickness ofthe interval towers

220230 cm 200 cm 250260 cm 240270 cm

Diameter of thecorner towers

ca. 1600 cm 1700 cm 1600 cm

Wall thickness ofthe angle towers

250 cm 275 cm 260 cm

Diameter of thegate towers

1370 cm 1350 cm 1430 cm1380 cm

ROMAN ROADS IN TRANSDANUBIAEndre Tth

A road-map (the Tabula Peutingeriana) and two itineraries(theItinerarium Antoniniand theItinerarium Burdigalense)list the sturdy roads of Pannonia. The road system was in

essence based on two major roads. One of them, the contin-uation of the ItalianVia Posthumia,led eastwards from theEmona and Poetovio area in southwest Pannonia to Sir-mium and the Danubian border between the Drava and theSava, ensuring communication between Italy and the west-ern part of the empire with Asia Minor and the easternprovinces.

The other road led northwards from Aquilea, the borderof Italy, to the Danubianlimesof the empire through EmonaandPoetovio.ThisroadisknownastheAmberRoad,amod-ern name given to this route after the trade in amber con-ducted along this road. A milestone found near the southern

gate of Savaria records quite exceptionally the distancefrom Rome: the 675 miles correspond to the 1000 km be-tween Savaria and Rome on Roman roads(Fig. 13).

Following the occupation of the Danubian border at theend of Claudius reign and the construction of the chain offorts and watchtowers along the river in the Flavian period,a network of roads was also built to ensure communicationbetween them. The significance of the Danubian road in-creased owing to the single-line defence system.

In theItinerarium Antonini,the main roads of the prov-ince, starting from Poetovio, are listed in a sequence to pro-

vide a clear overview of the military and civilian administra-tive centres. After describing the main roads leading

through the province (the military road along the Danubeand the SirmiumSopianaeSavariaAugusta Treverorum/

Trier road starting from Byzantium), it goes on to list thelocal branches from Savaria and Sopianae/Pcs, the twoproconsuls seats during the Tetrarchy. It describes theroads leading from Savaria to the legionary fort of Vindo-bona/Vienna and Carnuntum, and thence to the legiosta-tioned in Brigetio/KomromSzny through Arrabona/Gyr, and finally to Aquincum. Another road led fromSopianae, the seat of the civilian governor of Valeria, to

Aquincum, Brigetio and Carnuntum. In its description ofthe main stretches of these roads, the itinerary always gave

the civilian and military governors seat and legionary fortsas the end-stations.The identification and mapping of the major Transdanu-

bian roads of the Roman period was begun in 1980. It soonbecame clear that of the roads described in the itineraries,only the remains of the roads that had been continuouslymaintained and resurfaced could be identified during fieldsurveys. The course of the western Transdanubian roadscan be identified along almost their entire length, whileonly shorter sections of the roads in eastern Transdanubiacan be determined. The information contained in theseitineraries is incomplete since many passages were repeat-

Fig. 12. Fragment of an emperor statue from the Alshetny fort


17/60

Roman roads in Transdanubia| 219

edly copied and some stations, together with the distancesbetween them, were left out of later copies. The analysis oftoponyms and old maps is also invaluable for identifyingthese roads. Land surveyors interested in artificial topo-graphical features often sketched Roman road sections inthe early 19th century.

The course of the known roads, such as the AmberRoad and the SavariaArrabona road, indicates that mostof them ran along a straight line whenever the terrainmade this possible. In more hilly regions, the road-build-ers usually skirted the slopes and tried to keep to the samealtitude (this could be best observed in the case of the road

that cut off the Danube Bend and ran along the southwest-ern edge of the Pilis).

The entire Pannonian section of the Amber Road isknown. The road crossed the Drava south of Lenti, continu-ing to themunicipiumof Salla/Zalalv on a course that ismore or less parallel to the one-time VarasdBratislava road.From here it continued to the Ndasd area, where it branched

into two before crossing the Rba. The western branch lednorthward to Krmend and Savaria. The eastern branchcrossed the Rba north of Katafa; its course can be traced toSavaria.A small fortenclosed bya ditch controlledthe roadonthe southern bank of the Rba. The bridge over the Rba isshown on a drawing from the 19th century. There was a roadstation on the western side of the road at Sorokpolny untilthe mid-3rd century(Fig. 14). The road functioned as thenorthsouthmain road of Savaria; leaving thetown, it entered

Austria at Olmd; its course is known up to Sopron, whence itcontinued to the Danube and Carnuntum.

The course of the SavariaBrigetio road can be well

traced from Savaria to the Rba. It branched into two northof Srvr. The southeastern branch continued towardsAquincum. Dnes Gabler unearthed a road station on thenortheastern branch. The section between Ostffyasszonyfaand Ppc is known only from a 19th century manuscriptmap since it was destroyed by ploughing. This branchturned eastward at Ppc, crossed the Rba and continuednorth-northeast toward Egyed, passing a road station. Amilestone was ploughed up beside the Roman road north ofRbaszentandrs; its inscription only indicated the dis-tance: XLVII MP, the distance between its findspot andSavaria. The road crossed the Rba and passed through themunicipiumof Mursella, joining thelimesroad to Arrabona

near Mnfcsanak, from where it continued to Brigetio.The posts of the bridge across the Rba on the road from

Savaria to Aquincum have been preserved in the river bed.The road can only be traced along short sections fromSrvr to Celldmlk. It reached the Bakony Mountainseast of Soml Hill and ran towards Aquincum near present-day Road 8 on a course that is more-or-less known.

Even though only short sections of the SavariaSopianaeroad are known, the terrain more or less outlines its course.

The road branched off from the Amber Road north ofSorokpolny. Few traces of its gravel surfacing survived upto the Rba, where the remains of a road station mark the

crossing place. From here the road turned southeast; prac-tically nothing has survived of its gravel surfacing. Writingin the 19th century, Flris Rmer noted that the Romanroad was still in use in the Keld area. The road ran south-

wards through Balatonhdvg or Fenkpuszta, then turnedsouthwards toward Szigetvr through Somogyzsitva. A doc-ument from 1217 mentions a large road called Via Impera-torispassing through the village of Basal lying north of thetown that can perhaps be identified with the Roman road.Crusader armies, such as the one led by the Emperor Fred-erick Barbarossa, marched down this road on their way toConstantinople, explaining its medieval name.

Fig. 13. The milestone fragment from Savaria and its inscription


18/60


Even less has survived of the roads leading northwardfrom Sopianae to Arrabona, Brigetio and Aquincum. Thethree roads left Sopianae on two courses. Judging from theterrain, the eastern road to Aquincum probably passedthrough the Mecsek Mountains. along the HosszhetnyZobkpusztaMagyaregregy line, although it is quite possi-ble that it more or less coincided with the line of present-day Road 6 up to the Bonyhd area, where it turned north.It ran somewhat east of Vajta and Cece toward Tc, whereit either passed through Gorsium or nearby Herculia (Sza-badbattyn?), continuing first to Trnok and then to

Aquincum. The other two roads probably followed thesame course until the Dombvr area. The western roadreached Lake Balaton through Alshetny and Sgvr. Itskirted the eastern banks of the lake and arrived to

Arrabona through Litr, Zirc, Veszprmvarsny andPannonhalma. The course of the eastern road is not known.It most likely led through the Szabadbattyn area and pass-ing between the Vrtes and the Bakony Mountains alongthe MrOroszlnyKocsMocsa line, it eventuallyreached the legionary camp of Brigetio.

The road that cut off the Danube Bend followed a coursethat can be clearly traced partly from the terrain and partly

from the ploughed-up gravel surfacing along the Pilis-borosjenPiliscsvKesztlcDorog line, joining thelimesroad at Tokod. The foundation walls of stone watchtowersfrom the close of the 4th century ensuring the safety of theroad were unearthed beside the road at Pilissznt andPiliscsv.

The course of the well maintainedlimesroad runningalong the Danubian border of Pannonia and of the em-pire is known from aerial photos analyzed by Zsolt Visy.

This road connected the legionary and auxiliary forts andthe watchtowers.

The roads usually ran on a 5070 cm high embankment.

The embankment supporting the Amber Road has survivedin a fairly good state of preservation at Ndasd, while that ofthe SavariaArrabona road could be observed at Kemenes-szentpter. The road was 58 m wide. The embankment ofthe western Transdanubian roads was constructed of gravelrammed down hard to a thickness of 6080 cm, usually

without a foundation of larger stones. In the upper layers of

the roads, the gravel was held together with a binding agentmixed with mortar. The binding material adhering to thepebbles has survived in some spots. A pavement of stoneslabs was only made for the stretches passing throughtowns. The roads were paved with stone slabs in Savaria,Scarbantia and Aquincum(Fig. 15). Since the roads restedon an embankment, there was no need for a ditch alongtheir sides. Although later agricultural cultivation has lev-elled and destroyed most of these embankments, theploughed-up gravel layer very often outlines the originalcourse of these roads. Since gravel was more scarce in east-ern Transdanubia, the embankment was made of crushed

stone, of which little survives. Longer road stretches andchanges in the course of the roads can only be documentedfor thelimesroad.

Little is known about the inns, relay stations and bridgeson these roads. We know that these were spaced 15 kmapart on the Amber Road, depending on the terrain and,more importantly, on river crossings(Fig. 16). The mainfording places were controlled by fortified sentry towers,such as the one at Katafa. A road station has been foundsouth of Sorokpolny. The foundations of the pillar sup-porting the bridge of the road leading westwards fromSavaria across the Perint stream have been identified. Thisbridge is also mentioned in a historical source that recounts

how Quirinus, Bishop of Siscia, was thrown into the streamfrom this bridge. The road leading from Savaria to

Aquincum crossed the Rba river northeast of Srvr. Thebridge was supported by wooden posts set in the river bed;the spaces between the posts were filled in with basalt

Fig. 14. Road station on the Amber Road at Sorokpolny

Fig. 15. Detail of a northsouth street in Brigetio


19/60

The towns of Pannonia| 221

stones. The pillars of the bridge have been preserved in the

bed of the Rba river. The dendrochronological analysis ofthe timber remains dated the felling of the trees to the 180s.Similar posts supported the wooden bridge across the Dan-ube near the legionary fort of Aquincum.

The western Transdanubian Roman roads were still usedin the Carolingian period. Various documents from the

rpdian Age mention wide bands of gravel, calledtte-vny,casting, in Hungarian. The word also survived as atoponym. In some spots, the mortar binding the gravelcould still be observed on the surfaces of the roads; these

were calledopus cementariumin medieval documents, whilethe expressionvia lapidosa, paved road was used for theroad along the Danube.

THE TOWNS OF PANNONIA

THE EARLY DEVELOPMENTOF PANNONIAN TOWNS

Mihly Nagy

The Roman period heralded a new era in the history of Hun-gary: this period saw the emergence of urbanization in the

modern sense of the word. New construction techniquesmade their appearance, for example in the installation ofpublic utilities (aqueducts, sewers and drainage networks)and in the heating system of private and public buildings andbaths concealed under the floor and in the walls.

The appearance of towns in Pannonia actually meant theemergence of communities to whom certain rights of localgovernment and autonomy were granted. In contrast to themodern meaning of the term, the basic criterion of urbanstatus was the legal status of the community that controlleda particular area, rather than the size of the settlement orthe degree of its urbanization. An autonomous community

was in effect an organization of the native population occu-

pying a specific area that was granted the right of self-gov-ernment recognized by the state administration. An auton-omous community of this type could be a colonia, amunicipiumor acivitas peregrina,depending on its earlier de-

velopment. The army, and the civilian groups arriving in itswake, played a key role in the early development and evolu-tion of the Pannonian urban communities.

The Romans organized the native populations that hadsurrendered or had been subdued by the army into admin-istrative units calledcivitates peregrinae. At first, these cameunder the authority of apraefectus civitatis(an officer of theRoman army troops stationed in the area) who was laterreplaced by apraeposituschosen from the native popula-

tion; when this happened, thecivitasitself was granted agreater degree of autonomy. According to Andrs Mcsy,the organization of thecivitates peregrinaein Pannonia wasbegun after the suppression of the first phase of thePannon-Sarmatian, rebellion in 8 A.D. at the latest. Thelist of these earlycivitatesis known from the works of Plinythe Elder. Except for thecivitas Eraviscorum,the inscribedstone monuments of the Roman period rarely mentioncivitates (Fig. 17).

Certain areas of the province remained part of theterritoriumof thecivitatesafter the Roman conquest, whileanother part was probably expropriated and declaredager

publicus,state land. Theterritoriumof a particular civitaswas probably already separate from the territories con-trolled exclusively by the military administration, whosefunction was to provide for the needs of the military forts.

The first town, Emona, was founded in an area under mil-itary administration after the legion stationed in the forthad left and veterans were settled in the abandoned fort in15 A.D. The next town, Savaria, was probably alsofounded in the place of a military fort during EmperorClaudius reign by settling legionary veterans. Two colo-niae,Siscia and Sirmium, were created by settling veteranson the strategically important road leading eastwards in

Fig. 16. Depiction of a wagonand rider on a tombstone

from Intercisa


20/60


the Sava valley under the Flavians, probably in 71. Twomunicipia, Neviodunum (municipium Latobicorum) and

Andautonia, were also founded during the Flavian periodon the same road. The often large territory of the muni-cipiawas divided into smaller units called pagiand vici.Other communities with a certain degree of autonomy

were also established in territories under military control

(such as thecanabaebeside the legionary camps or thevicibeside auxiliary forts). Scarbantia was also founded duringthe Flavian period. The settling of the veterans in coloniaecontinued under Trajan. Poetovio was founded in theplace of a former legionary camp, most likely before thestart of Trajans Dacian wars. Most of the Pannoniantowns were founded under Hadrian. The emperorfounded the coloniaof Mursa at a major junction of theroad leading from Poetovio to Sirmium along the Drava.

Another first during his reign was the granting ofmuni-cipiumrank to Carnuntum and Aquincum, two civiliancommunities(canabae)near legionary forts that had not

been founded by the military. These towns, the capitals ofUpper and Lower Pannonia, were the largest towns of theprovince. The civilian communities that developed besidethe former military fort at the crossing place of the Zalaat Zalalv (municipium Sallensium), Mogentiana and

Mursella were promoted to the rank ofmunicipiaduringHadrians reign. In the lack of conclusive evidence, nei-

ther the exact location, nor the date of the foundation ofVolgensium, an inner Pannonian municipium is known.Inscriptions from the early 3rd century mention its townmagistrates. Themunicipium Iasorum, lying between theDrava and the Sava, is also considered to have beenfounded under Hadrian. The form of the name of thetown suggests that, similarly to themunicipium Latobico-rum, the original civitas peregrina of a population grouphad been reorganized into a municipium. It seems likelythat Cibalae and Bassianae were both granted the statusofmunicipiumby Hadrian.

In the lack of conclusive evidence, it is impossible to

Fig. 17. The tribes living onthe territory of Pannonia andthe early towns, up to the mid-1st century A.D.


21/60


determine the exact date when Iovia, lying on the roadalong the Drava in Upper Pannonia, was granted the statusofmunicipium,an event recorded on an altar from the 3rdcentury. The date when this rank was conferred on themunicipium Faustianensium, lying between the Drava andthe Sava, is similarly unknown.

Theviciof Brigetio and Vindobona were destroyed dur-

ing the Marcomannic wars (169172, 166). The honorifictitle(municipium Aurelium)of the civilian community of theBrigetiovicussuggests that it had received this title duringthe reign of one of the emperors bearing this name at thelatest. The lengthy wars also arrested the development ofseveral other northern Pannonian towns. Rebuilding pro-

jects continued with a greater zeal during the Severan pe-riod, at the beginning of the 3rd century.

The available evidence would suggest that no other townswere founded in Pannonia from the Severan period. Septi-mus Severus conferred the title of Septimia on twocoloniae,Siscia and Sirmium, and promoted threemunicipia,including

the two provincial capitals Carnuntum and Aquincum, andprobably also Cibalae, to the rank ofcoloniain 194. EmperorCaracalla granted the status ofcoloniato the Bassianamuni-cipium. Themunicipiumof Brigetio was also given the statusofcoloniasometime around the mid-3rd century.

The prosperity of the Severan period was shattered bythe Barbarian attacks that devastated Pannonia in 258260,reflected also in the countless coin hoards. The rebuildingof the damaged or destroyed buildings only began yearslater. Public and private constructions were underway evenas late as the 4th century. The decline of urban life is re-flected in the fact that the inscriptions erected by town offi-cials gradually disappeared, indicating that office-holding

in the town administration had lost its attraction. Contem-porary sources suggest that the Pannonian towns were in arather bad state by the later 4th century. Archaeological in-

vestigations have shown that certain town quarters wereabandoned and subsequently used for burial. The continu-ity of Pannonian towns was interrupted at the end of theRoman period. New towns emerged in their place duringthe Middle Ages Szombathely over Savaria, Sopron overScarbantia, Buda over Aquincum and Pcs over Sopianae since they were usually located at important crossingplaces.

RELIGIOUS LIFE IN PANNONIAN TOWNSIstvn Tth

About half of the towns of the one-time Pannonia provincelie in Hungary. Only five of them yielded assemblages thatprovided information on religious life. Two of these fivetowns, Aquincum and Sopianae, lay in Pannonia Inferior,another two, Savaria and Scarbantia, in Pannonia Superior,

while Brigetio was annexed from Pannonia Superior toPannonia Inferior in 214.

The proximity to Italy, the early foundation date and the

settling of legionary veterans in these towns during the 1stcentury essentially determined the religious life of Savariaand Scarbantia, the two western Pannonian towns. Theseare the only towns in the province from where a templededicated to the Capitoline Triad and monumental statuesof the deities are known(Fig. 18). These statues were madefrom Greek or Italian marble. Priests from other towns of

the province regularly dedicated altars in Savaria, the centreof the provincialara Augusti. Associated with the events or-ganized by the concilium provinciae, the cult of Nemesisplayed an important role in the town, just as in Scarbantia,

where a Nemesion with a rich find assemblage was uncov-ered in the amphitheatre.

The deities of the native population were also revered inboth towns. Local elements can be detected in the cult ofSylvanus, Diana, Liber Pater, Ceres, Hercules and Mercu-rius. In Savaria, this was complemented by a number of im-personal deities, such as the Numina, Fatae, Sphinces, Dii

Augurales, Dii Itinerarii and Semitatrices, indicating the

survival of strong pre-Roman traditions (similarly to the DiiMagni in Gorsium).

Fig. 18. White marble statues of Jupiter, Juno and Minervafrom the Capitoline temple of Scarbantia. Sopron,mid-2nd century A.D.


22/60


Egyptian cults arriving from Italy in the 1st century too

played an important part in the religious life of both towns.In Savaria this is reflected in the creation of a sacred pre-cinct, and in Scarbantia in the erection of a private sanctu-ary. A direct oriental cultural influence can be noted besidethe Italian one, leading to the emergence of a hereditaryEgyptian priesthood in Savaria and the appearance of anIsis-Bubastis dedication in Scarbantia, a unique phenome-non in the West(Fig. 19). Other oriental cults were repre-sented by a Iuppiter Dolichenus sanctuary in Savaria and a

Mithraeum in Scarbantia.Christianity also left its mark on Savaria. Beside a large

early Christian cemetery and numerous Christian represen-tations, this town was the birthplace of Saint Martin of

Tours, one of the renowned saints of the 4th century.Clashes between Christians and pagans in the late 4th cen-tury can be observed in both towns. The last pagan shrines

were destroyed and the statues of the Capitoline Triad werebroken. Although there is no evidence for the existence of abishopric in either town during the Roman period, thisnonetheless seems quite likely.

The cultural impact of thelegio IandII adiutrixon reli-gious life on quite obvious in Aquincum and Brigetio. Thecult of the state and military deities, Jupiter and Juno, Mars,Fortuna, Victoria, Venus and Dea Roma dominated in bothtowns. The highest number of altars was erected to Jupiter,

followed by dedications to Sylvanus and Mithras. Manyarmy officers and proconsuls were followers of the cult ofthese two deities. The Mithras sanctuary unearthed in thecentre of the Aquincum fort was lavishly furnished by thetribunus laticlaviiof the legion.

A deity blending a number of native elements was re-vered by many army troops. Called Hercules Illyricus, thisdeity can in a sense be regarded as a self-portrait of the 3rdcentury army transferred to the religious sphere. The cultof the emperor too played an important role in the military,especially in the 3rd century.

Many of the altars erected by the proconsuls in Aquin-

cum were dedicated to the imperial deities: Sylvanus andMithras. The cult of Mercurius and Nemesis has been doc-umented in the proconsuls palace.

Religious life in the civilian towns lying beside the le-gionary forts did not differ significantly from that in thefort towns. The dominant cults were those of Jupiter and

Juno, Silvanus and Mithras. The cults of the healing gods,

Aesculapius, Hygieia, Apollo and of Sirona, a Celtic god-dess, were prominent in both towns. In Aquincum, thepresence of Telesphoros, depicted asgenius cucullatusandinvoked in inscriptions, was a unique phenomenon.

The religious centre of the Eravisci lay on top of GellrtHill, south of Aquincum. Many altars dedicated to Jupiter

Teutanus have been found here, most of them dating to the2nd and 3rd centuries. They were mostly erected by theauguresof the community or the magistrates of Aquincumfor the welfare of thecivitas Eraviscorum.

The oriental deities popular in the 3rd century includeJupiter Dolichenus, Heliopolitanus and Adonis. The syna-

gogue of the Jewish god, called Deus Aeternus, is knownfrom Brigetio; an inscription dedicated by anarchisynagoguswas found inside it.

Relatively few Christian monuments have survived in thetwo towns. The augurs staff(lituus)with gold inlay recov-ered from a 4th century burial at Brigetio is a remarkablefind, reflecting the ultimate decline of paganism, indicatedby the burial of the high priestssignum.

Few finds that can be associated with religious life beforethe advent of Christianity have been uncovered in Sopianae.

There is evidence for the cult of Jupiter, Juno and Sylvanus,as well as of Liber pater and Terra mater. The altars werededicated by both the civilian population and thebeneficiarii

consulari.Lavish early Christian assemblages from the 4th5th

centuries are known from this town. Burials, mausoleumsand mortuary chapels, frescos and small finds preserve thememory of a flourishing Christian community and of con-tacts with Mursa and Sirmium to the south.

SAVARIAEndre Tth

In the mid-1st century, the Emperor Claudius founded a

town called Colonia Claudia Savaria on the Amber Roadthat led to the Danubian border of the Roman Empire,where he settled the veteran soldiers of the legio XVApollinaris. After the division of Pannonia in 107, thistown became the centre of the emperors cult and the seatof the provincial assembly of Upper Pannonia. The altarof the imperial cult, the sacred precinct and the buildingsof the provincial assembly were built at the western end ofthe town(Fig. 20). Following Diocletians administrativereform, Savaria became the seat of the civilian administra-tion of Pannonia Prima. The proconsul moved his resi-dence to the town and troops were stationed here to

Fig. 19. Bronze statue of the Apis bull from Gorsium


23/60


ensure its defence. Savaria was occasionally visited by theemperors (Constantius II, Valentinian I). Some of townscitizens and the inhabitants in the surrounding area fledto the south in the early 5th century. The abandonedbuildings were destroyed by an earthquake on August 7,

456. The name Savaria, however, survived until the Mid-dle Ages.

The research of Hungarys oldest town began sometwo hundred years ago. In 1791, Istvn Schnvisner, pro-fessor at the university of Pest, wrote a history of thecolonia, based on the epigraphic evidence, the survivingrelics and various other sources. His study gave a major

impetus to the collection of Roman finds. The first sys-tematic excavations were conducted in the late 1930s. Aceremonial hall with a mosaic floor measuring 17 m by46 m, part of a group of buildings from the late imperialperiod, was no doubt used for imperial representation(Fig. 21). The sacred precinct with the Iseum was uncov-ered in the mid-1950s. The rescue excavations precedingurban reconstruction projects vastly enriched our knowl-edge of the topography of this Roman colonia, enablingthe reconstruction of its layout and street system, as wellas of the various periods in its life. The town had agridded street system with the Amber Road, traversing

the town in a north to south direction, in its axis. The 6 mwide streets were paved with flat basalt slabs. The excava-tions conducted more or less continuously since the 1990shave brought to light a number of stone houses from the2nd century that overlay the earlier timber framed build-ings of the 1st century and kept the original orientation of

Fig. 20. The southwestern town quarter of Savaria with thesanctuary of Isis

Fig. 21. The Roman palace building from the 4th century A.D. and the Carolingian round fort in Savaria


24/60


these buildings. The roads were paved with basalt slabs atroughly the same time. A few residential blocks with ar-caded street fronts have also been unearthed. A vaultedunderground sewage system constructed of stone carriedthe waste water from the houses. The successive periodsof the towns development correspond to the increasingadministrative significance of the settlement.

SCARBANTIAJnos Gmri

The ruins of Roman Scarbantia lie four and a half metresunder the centre of Sopron. The southern part of theforumcan be seen in the cellar of the modern building on the cor-ner of j Street and Szent Gyrgy Street(Fig. 22). To its

west stood thebasilica,the house of legislation with its rowof slender columns. In the north, the imposing temple oftheCapitoliumtowered over the rectangular square measur-

ing 45 m by 46 m.The settlement of wooden buildings and adobe huts wasknown asoppidum Scarbantia Iuliain Tiberius time (1437).

Major constructions were begun in the earlier 2nd century,a few decades after the town had been promoted to the rankofmunicipium Flavium. The stone amphitheatre buttressed

with earth was also built in the 2nd century, approximately

Fig. 22. Excavation plan of the forum of Scarbantia

Fig. 23. The town and the walls of Scarbantia from the north


25/60


at the same time as theforum. The excavations conducted in1925, 1933 and 199192 revealed that the arenameasured42 m by 63 m, while the width of thecaveawas 21 m. Theruins of the town bath built in the 2nd century and the re-mains of thehypocaustumsystem that heated the hot-waterbasin from below were uncovered in the cellar of what is to-day the Szent Orsolya School in 1950.

The inscribed stones found in Scarbantia and its envi-ronment indicate that the town had held numerous sanctu-aries and temples; the ruins of some have been uncovered(Nemeseum, the temple of the Capitoline Triad, the

Mithraeum in nearby Fertrkos). The merchants erectedan altarstone in the sanctuary of Mercurius standing besidetheforum. The followers of Liber Pater, another name forBacchus, also presented their sacrifices in a neighbouringtemple.

The pottery kilns were found on the site of the Ikvashopping centre and between Szchenyi and Dek Squares.

Travellers arriving to Scarbantia from the south passed the

cemetery flanking the main road, containing the tomb-stones of prosperous families, often with painted scenes orrelief decoration. The inscriptions on the tombstones re-

veal that some of the people who settled in the Scarbantiaarea came from the trading houses of Aquileia and neigh-bouring Tergeste/Trieste.

At first, Scarbantia was an open, undefended settlement.The town walls were built at the beginning of the 4th cen-tury. The town centre was enclosed within a 3.5 m thickand 8.5 m high stone wall with 35 horseshoe shaped bas-tions. The north-south axis of the oval fort was a 400 mlong stretch of the Amber Road. A 100125 m wide zone

was fenced off on either side of the 6 m wide road paved

with gneiss slabs leading to the fort. The town had two gatetowers. The northern gate, opening towards Carnuntumand Vindobona, is still visible under the present-day belfry.

The southern gate toward Savaria probably lies under thetown wall section between 18 Szchenyi Square and theOrsolya School. A pedestrian gate was cut into the easternand the western side of the wall. The groundplan of the fortis clearly outlined by the town wall along the VrkerletSzinhz StreetSzchenyi Square line that was rebuilt dur-ing the Middle Ages. The ashlar faade of the Roman town

wall and its bastions can still be seen in several places(Fig.23). When the Avars invaded the Carpathian Basin in 568,

the towns inhabitants evacuated Scarbantia, and, togetherwith their bishop Vigilius, they joined the Lombards andmigrated to northern Italy.

AQUINCUM: THE CIVILIAN TOWNPaula Zsidi

Aquincum developed at one of the most important crossingplaces of the Danubianlimes,at the junction of the east

west road arriving from the Solymr valley and the roadrunning along the Danube. There was a settlement on the

site of the town before the Roman period. In view of thestrategic importance of the area, the earliest Roman archi-tectural features were military in nature. They probablydate from Domitians reign, when thelegio II adiutrixwastransferred to buda.

Being a settlement belonging to the capital of Pannonia

Inferior province, thevicuswas granted the rank ofmuni-cipiumaround 120

Documents

Roman period