Riverside Maria and the Fall of Vindobona

MAGYARUL

There are "speaking" geographical names along the Danube, sometimes in the most unexpected places, often far from the river, such as one of Vienna's oldest churches. The German name of the Gothic church of Maria am Gestade clearly indicates that it stood on the riverbank when it was built, but today it is several streets and 320 metres away from the Donaukanal. The evocative name of the former church of the Danubian boatmen takes us back to the ancient history of Vienna's urban waterways and explains the peculiar shape of Vindobona's late Roman legionary camp.

The idealized image of the castra Vindobona before the III. century (source)

42 steps. That's how many steps you have to climb to get from Salzgries Street to the Roman Catholic Church of St Mary of the Riverside, on Passauer Platz, named after the former owner of Bishopric of Passau. But there are just as many steps for those arriving from the Tiefer Graben (Deep Trench - another "speaking" geographical name!), but probably even more for the medieval boatmen who climbed up from the Danube harbour below the bank to the church on the high bank, since the landfills that has since been done has reduced the difference in level considerably over the centuries. This high bank, marked by steps, can be followed eastwards to Schwedenplatz, and the site of Vienna's oldest church, the Ruprechtskirche, legend has it that it was founded in 740. The riverside Church of St. Mary and the medieval city wall immediately adjacent to it have dominated the Danube landscape of Vienna for centuries, initially the wooden chapel, then the stone church rebuilt in Gothic style at the turn of the 14th and 15th centuries, and its 56-meter-high tower, which was already raised by another 15-meter of the higher ground. 

42 steps of the Marienstiege towards the Passauer Platz.

The right bank of the Danube in Vienna can generally be compared to a flatter Greek theatre, where the surface slopes gradually and arena-like towards the Danube. At the end of the Last Glacial Maximum, this slope was even less uniform, with flatter surfaces alternating with steeper slopes, as if they were terraces. This stepped, terraced pattern was formed by the changing water flow of the Danube during the Ice Age, mainly for climatic and tectonic reasons, and the gravel material of the higher terraces was also covered by loess layers during the late Ice Age. However, since their formation, these terrace generations have been gradually exposed to the very strong lateral erosion of the Danube, caused by the local braided, anastomosing bedforms of the Danube. Vienna's urban core was also created on one such flood-free Danube terrace, the youngest, namely on the so-called Prater Terrace. The Prater terrace was formed during the Würm glaciation, and its gravel was deposited in Vienna at an altitude of 145-153 metres above sea level.

Danubian Terraces of Vienna. The Prater Terrace is nr. 6 with lilac color. (source)

The Roman military engineers had a fairly good sense for choosing camp sites, typically settling along river crossings, on flood-free ground, at road crossings, in well-protected areas and near fresh water sources. In Vienna, too, they built as close to the river as possible, where, after the conquest of the province of Pannonia, the Roman Empire built an earth-wood military camp for its auxiliaries on the flood-free Prater Terrace. In the case of Vindobona, crossing the Danube was somewhat problematic, as the floodplain reached a width of 6-10 kilometers and dotted with islands and gravel bars, and often had no specific main branch. The Romans chose the Danube estuary of the Ottakringer stream, nowadays known mainly for the yellow-labeled beer, which cut deep into the loess surface, as the military base, following the erosion valley of the Tiefer Graben, providing a natural moat for the military camp of Vindobona. There was also a good reason why the fortress was not built on the banks of the Wienfluss: the marly rock and soil conditions of the Wiener Wald mean that rainfall can only drain away to a limited extent, so at major rainfalls discharge often increased to devastating floods towards the Danube, and the cone of alluvium deposited in the riverbed not only impeded navigation but also had a profound effect on the development of the bends in the Danube branches on the right bank.

Cross section of the Danubian Terraces. Lilac: landfill, gray: loess, green: terrace gravel (source)

At the turn of the 1st and 2nd centuries the standard and regular Roman legionary camp was built, with a standard layout, familiar from many places along the Danube, but it is important to note that the orientation of the camps was not always aligned with the cardinal points, at Vindobona there is a 45 degree angle of deviation, i.e. the sides of the camp do not give the cardinal points, but the corners. Within the walls, the command buildings, barracks, baths and stores were arranged in a predetermined order in a rectangular area of about 450*500 metres (18.5 hectares) with three gates. The castrum's 3 metre wide stone walls were divided by towers, which were often rebuilt in differing shape from the ruins after the major destruction. A regular network of streets was laid out in the inner area of the legion camp, and this network of streets, together with the city walls and moats, still defines the street network and layout of the city of Vienna today. However, the archaeological excavations have not been able to reconstruct this regular layout in the Austrian capital, no finds have been recovered from the northern part of the castrum and no building foundations have been identified. The excavated Roman city wall in this section followed a natural break between the Maria am Gestade and the Ruprechtskirche. Since it is unlikely that the Romans in Vindobona deliberately deviated from the usual patterns, it can be assumed that the fortress wall may have suffered some irreparable damage that prevented it from being restored to its usual form. No barbarian opponent of the time was capable of such a feat, only the Danube. 

The line of effect of the Salzgries-arm within the legionary camp. (source)

There are two plausible scenarios for the destruction; a gradual erosion, which has been taking its toll over decades, gradually eroding the high bank and the fortress wall, and a single, extraordinary event. The latter is more likely. Sometime during the 3rd century AD, an extremely high flood may have flowed down the then westernmost branch of the Danube in Vienna, which science has named the Gonzagasse-Danube from the street that runs along its present-day course. This was the branch of the Danube on the banks of which Vindobona was originally built. The Roman legionary camp, however, was situated on the outer curve of the bend, where the proximity of the current line meant that bank erosion could occur even at mid-water. This was probably not simple riverbank erosion, but the displacement of a branch of the Danube, which also had a tectonic prediction in the fault lines of the Pannonian clay layers forming the terrace material, i.e. slides could have exacerbated the situation. 

The Danube has carried away more than 100 metres of the river terrace, shortening the lower reaches of the Ottakringer stream, removing up to several million cubic meters of sediment material from the Prater terrace, and at the same time from the western suburbs of Vindobona and the northern corner of the castrum of Vindobona. It is even possible that the suburban amphitheatre stood on the same site and also collapsed into the Danube. Similar changes in the riverbed were not at all uncommon in the pre-regulation period in the vicinity of Vienna, with sediment or other riverbed material often blocking off branches and causing islands, gravel bars and even settlements and bridges to be demolished in search of new routes. Severin Hohensinner's images illustrate this process over the last two millennia. 

42 steps to Maria am Gestade church

It is thought that material from the collapsed wall of the camp was used in Roman times to stabilise the collapsed bank and rebuild the northern wall of the camp along the newly formed river branch, which science calls the Salzgries branch from another street. This street name is also an evocative geographical name, a reminder of the salt trading ships that harboured here. Vindobona was by then past its heyday, having experienced a decline in socio-economic terms broadly similar to that of Aquincum. At one point in its history, a dwindling and impoverished population moved within the fortress walls until, sometime in the early 5th century, Roman administration ceased and the camp and its suburbs fell into ruin. If there was a continuity in the population, archaeology has found no evidence of it, the latest coin dates from 408 and after the middle of the century a layer of no archaeological remains was formed over the Roman ruins. 

Steps to the Ruprechtskirche, Vienna's oldest church

But the Salzgries branch of the Danube remained where it was, in the northern foothills of the ruined Roman fortress walls, and a thousand years later the church of Maria am Gestade was named after it. The name was a timely one, for the river, which was once quite navigable in size, gradually became silted up and narrowed during the 12th century, and the Danube bed moved away from the high bank to the north-east to form the so-called Porzellangassen branch, but still remained within the present-day Danube Canal to the west. 

Riverbed changes during the antiquity and in the Medieval Era in Vienna's 1st district.
1. Gonzagagasse-branch 2. Salzgries-branch 3. Porzellangassen-branch 4. Ottakringer stream
10. Wienfluss 11. Trench system of Vindobona (source: Wasser | Stadt | Wien)

The shifting of the Danube riverbed in the 3rd century left such an indelible mark on Vienna's urban structure that it affected not only the differences in geomorphology but also the city's road network. Neither the thousands of years of human landforming nor the great Viennese river regulation of 1875 were able to erase the work of the Danube that had been preserved precisely because of the building development. And if we listen carefully to the geographical names that speak, they tell us about the historical and geographical changes that have taken place over the millennia. 

Literature :

• https://tobias-lib.ub.uni-tuebingen.de/xmlui/bitstream/handle/10900/61026/CD71_Gietl_et_al_CAA_2003.pdf?sequence=2&isAllowed=y
• https://www.wgm.wien.at/hydrogeologische-forschung/news/flut-von-vindobona
• https://www.geschichtewiki.wien.gv.at/Salzgries
• https://www.wien.gv.at/kulturportal/public/grafik.aspx?bookmark=0nZLRk2zFUYHn7dFw-aI3RRwpAtZGVBFvuBteonQ1N1C4dSRsFu7fFg-b-b
• https://de.wikipedia.org/wiki/Vindobona
• https://www.wgm.wien.at/fileadmin/docs/hydrogeo-forschung/2022/Vortr%C3%A4ge/Pr%C3%A4sentation_WGM2022.pdf

Translated with DeepL.com (free version)

Blasting the Schwalleck

MAGYARUL

Despite the fact that sirens throughout the city had given half an hour's notice of the historic event, the explosion of the Schwalleck rock in Grein, which overlooks the Danube, caused serious collateral damage. Collapsed walls of apartment blocks, roads buried by debris and damaged railway embankments signalled the beginning of the process of making the "Austrian Iron Gates", the infamous Strudel Strait, navigable. 

The Schwalleck peninsula. A detail from the map of Strudengau. (Leopold Franz von Rosenfelt, 1721.)

The term "Austrian Iron Gates" is no coincidence, the Strudengau in Austria was as much a dreaded passage for sailors as the Iron Gate, Islás or Tachtalia on the Lower Danube. On this 25-kilometer stretch, the Danube cuts its way through the 300-million-year-old granite cliffs of the Bohemian Massif between Ardagger/Dornach and Ybbs/Persenbeug, where the rocky reefs and the whirpools they create have caused the destruction of many ships, for example between Wörth and Werfenstein. Just as the (second) regulation of the Iron Gates began in 1890 with the blowing up of a symbolic rock, the Grében, so the regulation of the Strudengau began with explosions, the first being the castle of Donaudorf on the right bank near Ybbs on 20 December 1955, and the second being the blowing up of the left bank of the Schwalleck promontory near the town of Grein in Upper Austria.

Outlook from the Schwalleck towards the town and castle of Grein.

Until Friday, 13 June 1958, at 12.30 pm, the Schwalleck Cliff reached far into the Danube, and in a very bad place, for navigation, as the drift line of the river led ships travelling downstream straight into it, while above it, at the mouth of the Greinerbach, a shoal was formed where flating ice were regularly stuck, and the cross-section of the river at the cliffs was also narrowing, causing the river to speed up, which made life difficult for sailors heading in the opposite direction, towards Linz. Geographically, the 250-260 m high Schwalleck was a counterpart to the cliff on which Grein Castle was built at the end of the 15th century, and these two heights defined the local skyline, the town of Grein being either depicted from here or depicted as being on the picture. It was about 40 to 50 meters above the Danube's zero level.

The doomed Schwalleck, and the evacuated buildings at its foot. (source)

It was originally topped by a cross called Halterkreuz, the origin of which is told in a local legend: a shepherd from Grein used to graze his cattle near the cliff during high water. While grazing, he was trying to fetch scrap wood out of the flood, which he wanted to store for firewood during the winter. When he tried to drag a large branch ashore, he slipped and fell into the water. Unable to swim, he desperately squeezed the drifting branch and, fearing for his life, vowed that if he managed to escape, he would place a cross on the shore to the glory of God. The river eventually swept the branch ashore, where he managed to cling to a tree hanging in the water and was lucky to escape. He kept his vow and the so-called Shepherd's Cross stood at Schwalleck Cliff until the explosion. [1] There was another cross here, the Schwalleck Kreuz, which was also destroyed by the blast, but was later reerected on the side of the cliff face opposite the Halterkreuz [2]. 

The moment of the explosion (source)

Austria's first hydroelectric power plant on the Danube was built jointly with Germany between 1952 and 1956, on the border between the two countries, just next to the unique Jochenstein cliff, which fortunately was not blown up along with the statue of St John Nepomuk on it. Austria then began to barrage the Danube, and construct the next power station which was located at the lower end of the Strudengau, between Ybbs and Persenbeug, precisely because of the navigability of the Strudengau. 

On 11 June 1958, the mayor of Grein sent out a leaflet informing the population of the impending explosion. The precaution was certainly justified, as there were already inhabited houses less than 100 metres from the Schwalleck cliff, which the town did not wish to demolish. Some nearby structures had to be sacrificed, such as those built at the base of the cliff. Doors and windows had to be left open, parked cars were removed, and valuables and furniture were moved to the far corners of the rooms for safekeeping. However, due to the proximity of the site, the blast caused significant damage to the city [3]. 

Landscape after the blast (source)

On that day, 5,800 kilograms of explosives blew up the cliff, from which more than three hundred thousand cubic meters of stone were then extracted in several stages, ensuring the unobstructed flow of ice and the necessary width of the shipping lane. Interestingly, a small cliff was left as a memorial between the Danube and the new main road, and the Shepherd's Cross was put back on it when the works were completed. Since then, the vegetation has conquered the cliff, the cars the new route and the boats the less dangerous bend in the river. The Ybbs-Persenbeug hydroelectric power station has raised the water level by 11 metres at the power station, about 5 metres at Wört and slightly less at Grein, and has removed the rapids, reefs and cliffs of the Strudengau. There is now no threat to navigation on the uniform river.  

The boulder of Halterkreuz left as a memento (forrás)

Everyone seems to have well served. If you're passing by, you don't miss anything, as if everything was already like this. What we don't know doesn't hurt. To those whose hearts we might have hurt by this old story, we apologise. 

But we will continue.  

Translated with DeepL.com (free version)

Danke an die Autoren des Österreichischen Donaubuchs für die Idee!

[1] https://www.grein.at/Tourismus/Sehenswuerdigkeiten/Halterkreuz

[2] https://www.grein.at/Tourismus/Sehenswuerdigkeiten/Schwalleck_Kreuz

[3] https://www.im-fundus.at/das-greiner-schwalleck-gefuerchtetes-schifffahrtshindernis/

Where the Amber Road Crossed the Danube

MAGYARUL

Stopfenreuth is probably not one of the best-known settlement names along the Danube, even in Austrian terms, and the riverine forest on the outskirts of the village, which barely belongs to Engelhartstetten, might even seem like the end of the world from Vienna, even though in ancient times one of the arteries of European trade, the Amber Road, ran crossed the Danube here. There is no visible evidence of the site's importance in the muddy riverine forest today, but when archaeologists scratch the sediment layers of the Danube, they come across ancient Roman walls.  

The tower of the military camp, unlike what we are used to, was located inside the wall.  

On 20 February 2025, the website of the Donauauen National Park published an article that archaeologists from the Carnuntum Museum and the Austrian Archaeological Institute had discovered a Roman fortress in the Stopfenreuther Au area of the Austrian national park. To be more precise, only what was left of it by the Danube; a short section of wall in the northern corner of the military camp, together with the base of an inner tower. The island, called Stopfenreuther Au, is located on the left bank of the Danube, in fact in the Barbaricum, opposite today's Bad Deutsch-Altenburg, but slightly to the north-east of the ancient legionary camp of Carnuntum. Excavations began towards the end of the vegetation season in order to minimise disturbance to wildlife, as the ruins lie on the floodplain, under 0.5-0.8 m of sediment, which is now completely covered by a vast riverine forest.

The fine-grained Danube sediment, which covered the ruins in thick layers, is clearly visible.

The Boris flood on the Danube in September delayed and hampered the work, but the excavation was completed in November, the trenches were reburied and, despite the fact that the Roman fortress (unlike Hungary) has been a World Heritage Site in Austria since 2021, the national park has requested that the vegetation be allowed to reoccupy the strictly protected area. This will be followed by an evaluation of the finds, which will hopefully shed light on the circumstances of the fort's construction and destruction, and will allow conclusions to be drawn about the hydrological conditions of the period, since archaeology is currently unable to say for certain on which bank of the river this Roman military camp, which was largely destroyed by the Danube, was originally built.  

The Stopfenreuth excavation in 2024 as seen from above.  

Erőd-ellenerőd példákat a Duna számos pontjáról ismerünk a római kori Pannoniából, Brigetioval szemben ott állt az izsai Leányvár, Pestről ismert a Március 15. téri erőd, a kisebb kikötőerődökről nem is beszélve (Dunakeszi, Nógrádverőce, Dunaszekcső, stb.) Stopfenreuth-nál alig maradt valami a római erődből, de a meglévő falszakasz alapján egyértelműen látszik, hogy kiserődről van szó, nem pedig kikötőerődről, erre utal a fal íve és az ív belső oldalán épült trapéz alaprajzú torony, sőt a toronyalap helyzete alapján azt is feltételezik, hogy ez nem egy későrómai építmény, amikor a patkó alakú tornyokat az erőd falsíkjához képest kívülre építették. 

Examples of fortress and counterfortress pairs are known from many places on the Danube from Roman Pannonia; the fortress of Leányvár in Izsa opposite Brigetiovo, the fortress of Március 15. square in Budapest, not to mention the smaller harbour fortresses (Dunakeszi, Nógrádverőce, Dunaszekcső, etc. ) At Stopfenreuth, hardly anything remains of the Roman fortress, but the existing section of wall clearly indicates that it was a small fortress, not a harbour fortress, as is indicated by the trapezoidal tower built on the inside of the arch and the position of the base of the tower, which suggests that it was not a late Roman structure, when the horseshoe-shaped towers were built outside the wall plane of the fortress.  

Location of the Ödes Schloss fortress and the ferry crossing on the III. military survey (1870s)

Although all the remains of the fort are covered in floodplain sediment, the ruins were not unknown to archaeologists, with substantial wall stumps still standing on the bank in the 1860s. Where the Danube had just cut through the walls of the fortress, a cross-section of the stone wall was cropped out perpendicular to the bank, and the places where it was built were referred to by several names, one of the most common being 'Ödes Schloss', or Desolate Castle, a name that also often appears on maps, However, there were also expressive hydrological names, such as 'Hungerstein' or 'Durstkugel', which refers to lean years with significant draughts, as their appearance was probably related to record low water levels. 

The archaeological interest in the nearby legionary camp of Carnuntum (Bad Deutsch-Altenburg) and its civilian settlement (Petronell), despite the three kilometre distance, was the main reason why excavations were carried out here relatively early, at the end of the 19th century (1896, 1898, and 1900), when researchers already assumed that it was an important element of the fortified Roman Danubian border (ripa) connected to Carnuntum. The figure below shows the same section of wall as the one excavated in the autumn of 2024. On this site, archaeologists fount the brick stamp of the leg. XV Apollinaris, a legion known to have been stationed in Carnuntum between 9-61 AD and 73-117 AD, i.e. either the camp was built at this time or the building material was reused at a later date.

Site plan of the 1898 excavation.

Between Klosterneuburg and Bratislava, the Danube meandering through the Marchfeld was once a major obstacle to north-south traffic. This 60-kilometer stretch of the Danube cannot be characterised by a single main riverbed, which widens to 4.5 kilometres in places, where the constantly changing pattern of tributaries, backwaters, gravel bars and islands made it virtually impossible to create permanent crossings until the river was regulated. At Carnuntum, however, this Danube floodplain narrowed slightly, with the Kirchenberg limestone block, which emerged as a peninsula, forming a flood-free surface directly on the bank of the main branch, while at Stopfenreuth the floodplain narrowed to 1.7 km. This is still a relatively large distance, but the generally lower Danube water levels of Roman times allowed the establishment of a river crossing, which was an important cross-continent economic artery of antiquity, leading to the Baltic amber deposits from Aquileia. 

The stump of a section of wall destroyed by the Danube.

At the time of the first archaeological excavations between 1896 and 1900, the section of the fortress wall formed a spur in the Rosskopf branch of the Danube, and behind it the backwash formed a small bay in the inner part of the fortress. At the end of the 18th century, the Rosskopf branch was even wider than the branch that washed the high bank of the Carnuntum, now the main branch. Generally speaking, the river's course was constantly changing due to its course, and there were no two maps alike in that the islands, gravel bars or the banks are in the same position, which is why the cartographic representation of this stretch is much less accurate than that of other, more stable stretches of the Danube. The confluence of the two branches was generally at the foot of the Kirchenberg, where, according to contemporary maps, a ferry crossing operated. The southern crossing point of the ferry occupied a fixed position in the northern part of Deutsch-Altenburg, but on the northern bank the ferry port was constantly in flux, where it landed on the island surrounded by the Rosskopf branch, but mostly landed at the Ödes Schloss, suggesting that the crossing continued to operate after the loss of the importance of the Amber Road, albeit with less traffic.

The route of the Amber Road between Aquileia and the Danube border.

Geographical considerations were the main reason for the route of the Amber Road between Aquileia and Carnuntum, which avoided mountainous areas impassable in winter and was as close as possible to them, it had to cross as narrow a watercourse as possible in the province of Pannonia, the Sava at Emona (Ljubljana), the Drava at Poetiovo (Ptuj), the Mura and then the Zala at Sala (Zalalövő), the Raba south of Savaria (Szombathely) before reaching the Danube east of Carnuntum (Deutsch-Altenburg). At this point, it is assumed that there was a temporary pontoon bridge, which was replaced by a ferry crossing in the Middle Ages and modern times. After 1951, this was replaced by a cable ferry, which also transported cars. The latter was a relatively brief moment in the history of the crossing, and in January 1973 a new Danube bridge was inaugurated on the site of the crossing, which since 2012 has borne the name of the Lower Austrian Landeshauptmann Andreas Maurer, the only bridge between Vienna and Bratislava to date. The fact that the bridge now crosses the Rosskopf branch where the Roman army built the military camp almost two millennia ago says a lot about the Roman engineers' choice of location. 

Translated with DeepL.com (free version)

Source of news, pictures: 

• https://www.donauauen.at/aktuelles/news/das-roemische-brueckenkopfkastell-in-der-stopfenreuther-au
• https://de.wikipedia.org/wiki/Kleinkastell_Stopfenreuth
• https://www.lobaumuseum.wien/cms/erinnerung-an-die-rollfaehre-bad-deutsch-altenburg/

Early Holocene Island below Budapest's III. District

MAGYARUL

Thanks to the joint research of prehistoric archaeologists and geographers, more and more details are becoming known about an ancient Danubian island and its first inhabitants, located under the 3rd district of Budapest. This big island once stretched from Csillaghegy (north) to Hajógyári Island (south), but its side arm was blocked off from the main Danube some 6000 years ago for climatic and tectonic reasons. In the Early Holocene, this Western Danube arm was gradually occupied by the surrounding watercourses, but traces of the riverbed can still be seen in the street network of the Mocsárosdűlő and Csillaghegy. 

Early Holocene hidrography of the Óbuda plains.  (I. Viczián) [2]

The proceedings of the XI. MΩMOΣ Research of Prehistory Conference, 10-12 April 2019, focused on the relationship between the environment and humans. The Óbuda area has been a priority area of research in landscape and environmental reconstruction, especially in relation to the Roman past of Aquincum. In the publication published in 2023 as Volume III of the Prehistoric Studies, a total of three papers discuss the environmental reconstruction of the prehistoric Danubian floodplain of Óbuda in three different excavation areas. At the site of Királyok útja 291-295, the researchers found the Danube mouth of the Csillaghegy Ditch, i.e. the northern tip of the prehistoric island [1]. Shallow drilling was carried out at the swampy area of Mocsárosdűlő in connection with the prehistoric sites along Pusztakúti street [2], and the Neolithic settlement of the island was studied at the site of Nánási street 75-77 [3]. Sediment samples from the excavations in the Csillaghegy Ditch and the Mocsárosdűlő site yielded almost identical results regarding the date of the Danube branch's bedding, despite the fact that the subsequent filling of the former Danube bed created different hydrographic characters at different sections. 

A summary of the end-of-quaternary climate history, with the river meandering activity.

In the Early Holocene period, the Óbuda floodplain between Békásmegyer and Újlak was a very different landscape compared to today. The Danube riverbed was not nearly as well canalised, with its tributaries freely flowing through the floodplain on both sides. During the Early Holocene, the dominant main branch of the Danube had already developed in the Óbuda area, but at that time, even larger side branches were still surrounding the relatively numerous and large islands (e.g. the Óceán Ditch on the east). The Early Holocene Óbuda-Danube branch can be relatively easily identified on a map, as the street network largely preserves the contours of the old riverbeds. The street network of Csillaghegy still shows the Csillaghegy Ditch parallel to the Árpád Street, which turned southwards near the Árpád Bath of Csillaghegy. Here it could form additional islands in the widening channel in the Mocsárosdűlő area. The Early Holocene Óbuda Island was over 5 kilometres long and covered an area of about 600 hectares.

Previously, two ideas competed over where the Óbuda Danube could return to the main branch. According to Ferenc Schweitzer [4], the estuary was originally located at the northern tip of Margaret Island, north of Újlak. This supposed Danube branch has been identified by archaeologists in the form of a drainage ditch at the foot of the Kiscell hilltop, but the archaeological fact that the legionary camp of Aquincum, i.e. the built-up area, extended from the Danube to the foot of the hilltop, rules out a Roman Danube bed in this area. If there were smaller creeks below the Kiscelli Hill in prehistoric times, they were displaced from the area by the Roman period at the latest and diverted north of the legionary camp into a new, shorter and straightened channel. 

The other possibility is that Óbuda was last an island in the Pleistocene. The Óbuda Danube originally flowed along the lower reaches of the old channel of the Aranyhegyi creek in the Early Holocene, and its estuary was north of Hajógyári Island. This idea is supported by the age of the Danubian terraces excavated in the Óbuda area and their height relative to the Danube level. 

The highest elevations of the Early Holocene Óbuda Island are marked by the level of terrace II/a, which was raised during the whole Holocene by sand layers blown out of the barren western mountains, the Transdanubian terrace surface and the riverbed [4]. The terrace no. I alias the higher floodplain level had already been formed (at the Pleistocene-Holocene boundary - see the Gábris-Nádor Figure 2007), but the lower floodplain levels were only carved out by the Danube later, during the Subboreal. This means that each terrace level was 'one step lower' in the numbering: the present Terrace I (high floodplain) existed then at a low floodplain level, while the now flood-free Terrace II/a may have been a Terrace I (high floodplain) in the Early Holocene. 

Archaeology and earth sciences study the same layers in excavations, but their methods, approaches and basic objectives are different. Sometimes, however, the common sets create fruitful collaborations between representatives of each discipline, giving a more complex and comprehensive picture of the interaction between past people and their former environment. Archaeologists categorise buried layers of soil primarily on the basis of human factors such as artefacts, while geoscientists draw conclusions from the geochemistry, grain size, colour and mineral composition of the layer. The working procedures, the methods of dating and the instruments used deserve a separate entry, but fortunately the studies in this volume discuss this in a paragraph or two. In short, each archaeogeographical environment is characterised by different sedimentation. If a poorly sorted, organic matter-poor, gravelly, gravelly sandy layer is found, it may be inferred that the sediment was deposited in the bed of an active river. The total thickness of this late Quaternary river sediment in the Óbuda catchment is 10-15 m. The Danube beds were cut into this terrace gravel. The sediments are rich in organic matter, typically fine-grained and dark in colour, and typically settle in a marshy environment. Thus, based on the type, succession and thickness of sediments, it is possible to basically define a sedimentary environment, including the life cycles of a buried riverbed.  

The following sequence of layers was reconstructed at the northern section of the Csillaghegy Ditch (Királyok útja 291-295) on the basis of the Optically Stimulated Luminescence (OSL) dating:

6500 BC: Danube sediments, gravel, pebbles (active riverbed)

6000-5500 BC: shallow water, floodplain sediments of the Danube (sedimentation, decreasing water discharge)

5000-3000 BC: sediment deficit. Presumably, the reverse flow of a small creek in the Danube valley in the Danube lowlands carried sediments away. 

AD 1-1000: deposition of marshy sediments due to decreasing water flow, possibly caused by Roman drainage, which drained the waters of the Mocsáros marsh and Rómaifürdő springs mainly to the south.

1500 AD: accelerating sedimentation, marshy sediments.

Riverbed reconstruction based on sediment samples taken at the former mouth of the Csillaghegy Ditch  (Gy. Sipos) [1]

On the basis of the discovered Neolithic settlements, the researchers assume that in the Csillaghegy area the Western Óbudai Danube has already been blocked off from the main Danube by the Middle Neolithic. The water network was radically altered by the deglaciation, as the former Danube arm flowing in a southerly direction was replaced by the Csillaghegy Ditch, which transported waters of the mountainous area and springs into the Danube, but in the opposite direction, northwards. The Óbuda-Danube branch was influenced by the precipitation of the Atlantic climate phase. The sediments suggest that the Danube bed migrated slightly eastwards in parallel with the drift and incision, while the mouth of the stream in the Csillaghegy Ditch was dragged southwards. The latter is a phenomenon commonly observed in lowland streams flowing into the Danube. 

The cross sections of two different floodplain levels excavated during the excavation, "Section B is the higher floodplain, Section A the lower floodplain. (Viczián I.)

At the same time, the excavation at 295 Királyok útja revealed different processes at different levels of the floodplain in prehistory. At the higher levels, suitable for human settlement, only 30-60 cm of sediment including artefacts was deposited from the Middle Neolithic to the Late Iron Age, while the lower floodplain sediment deposited during the same period was 250 cm thick. This means that sediment accumulation on the higher ground was already taking place in a typically terrestrial environment at this time. This includes the human cultural layer, airborne dust, organic matter from vegetation and silt deposited by the major Danube floods. Meanwhile, flooding was much more frequent at the lower level, i.e. the differences between the two floodplain levels were gradually eroded by the Danube-derived sediment, making the initially lower floodplain level increasingly suitable for human settlement. In this respect, it is no coincidence that the ruins of the town of Aquincum, lying on a nearly flood-free surface, are covered in places by less than 25 cm of sediment.

Cross-section of the Mocsárosdűlő explored by shallow geological drilling
and the stratigraphic boundaries identified on the basis of macroscopic features (Sipos Gy.) [2]

In the Mocsárosdűlő area, the OSL dating reconstructed the following sequence of strata:

5930-4910 BC Sand and gravel sediment, active fluvial formation, island-structured, multi-branched river network. This layer was overlain by a fine-grained clayey silt layer, which was already a precursor of sedimentation. 

5410-4610 BC Another layer of gravel was deposited on top of the silty layer, indicating renewed river activity. This was the last active period of the Western Óbudai Danube. 

4610-2800 BC After this period, the presence of river sediment in the marshland area ceases. The decrease in the sedimentation rate indicates that, over time, floods reached the area less and less frequently. The Mocsáros, which at that time still had an open, lake-like water surface, was supplied by groundwater and streams from the surrounding valleys, but at the same time the alluvial cone of the Aranyhegyi stream was building up and expanding at the expense of the open water surface.

2800 BC - AD 14th c. Shallow, eutrophic, marshy environment, increasingly dark, organic-rich sediments. The carving out of the low floodplain, followed by Roman drainage, further depleted the groundwater in the area, and thus the open waters of the Mocsáros marsh. From the Middle Ages to the present day, sedimentation has continued, with the marshy surface becoming increasingly smaller. 

Results of OSL and C14 dating on sediments from Drill No. 4. (Gy. Sipos) [2]

After the Óbudai Danube became an oxbow, complex processes took place in the former riverbed, but in general, the surface-forming influence of the surrounding watercourses increased significantly, while the Danubian influence gradually decreased. Floods entered the old tributary less and less frequently, and the sediment deposited there could only be partially transported by the smaller watercourses flowing into the Danube. These watercourses also transported a considerable amount of (slope) sediment from the valleys running down from the west. As a result, the riverbed, estimated at a maximum width of 200 meters, has been steadily narrowing. One of the interesting features of this Early Holocene basin is that the upper and lower estuaries of the basin were not closed, as streams on both sides continued to 'use' the riverbed. Finally, it was only anthropogenic interventions in the 19th and 20th centuries (sluices, stream bed relocations) that eliminated the former upper and lower estuaries.

Hydrography of the III. district between the Middle Neolithic and the Middle Ages.
(I. Viczián) [2]

Danube influence continued to be exerted through the groundwater, which during floods was moving between the oxbow and the Danube through the loose, gravelly alluvium. During floods, the deepest areas, such as the Mocsárosdűlő, were saturated with water, while during low water periods the river sucked this stagnant water away. Later, human intervention, canalisation and drainage further increased this suction effect. In addition, the surface waters of the Mocsáros were fed by springs, surface water courses and direct precipitation. The combined flow of these probably did not exceed 2 cubic metres per second. In the volume Ókori táj, ókori város [4], the water yield of all springs in the area was estimated at 42000 cubic metres per day. This is equivalent to about half a cubic metre per second, to which must be added the average flow of the Aranyhegyi creek of 0.3 cubic metres per second, and the flow of other smaller streams. Over the past millennia, some of these watercourses have been buried by the city and climatic influences may have modified the water yield, so that an exact value for these can no longer be determined. By comparison, this is at most a third of the water yield of the Zala river entering the lake Balaton. In addition, it should be noted that the waters of the oxbow flowed in two directions due to the effect of the emerging watershed in the Óbuda-Danube basin. From the late Neolithic onwards, the waters of the Mocsáros surplus, the sources of the Roman baths, the Árpád spring and the waters coming from the Kert street in Békásmegyer reached the Danube in a northerly direction via the Csillaghegy Ditch, while the combined waters of the Aranyhegyi stream and the Rádl Ditch flowed in a southerly direction. It is possible, however, that the Aranyhegyi stream may have changed its course in the early period from its newly formed alluvial cone in the channel. From Roman times onwards, human influence has profoundly redrawn the hydrology of the landscape. 

Connections between archaeological cultures and sediments of the Mocsárosdűlő. (Gy. Sipos) [2]

As archaeological research uncovers more and more of the Early Holocene Óbuda Island, our ideas about the prehistoric landscape and environment are expanding and refining regarding the area of Budapest's III. district. The research even sheds light on when humans settled on this piece of land. 

The relationship between prehistoric man and the river is illustrated by the sites along the river banks, which in some periods were closer to the river and in others moved further away from it. This periodicity has long been known to climatologists because it can be linked to changes in climate. In wetter periods, when flood risk increased, human settlements moved away from the river, while in drier periods they may have even occupied the lower floodplains intermittently. 

The cutting off of the Óbudai Danube by the Middle Neolithic, facilitated access to the area, and the deepening of the main branch of the river created the opportunity for the settlement of communities of the Linear Pottery culture (c. 5500–4500 BC). The occupation of the former Danubian areas of the island of Óbuda may have begun during the period of the pottery of the Notehead ceramics style (c. 5300-5200 BC). Then, gradually, several new settlements were established, when there was only an oxbow at the site of the Mocsáros. The site of the village excavated by Zsuzsanna M. Virág at 75-77 Nánási Street was inhabited until the end of the Želiezovce group period (c. 4900 BC). During the Želiezovce period, previously uninhabited islands were also populated, such as the northern tip of the present-day Óbuda Island. This means that the present-day Óbuda Island and Margaret Island with its similar floodplain levels may have existed at the end of the Holocene. The settlement excavated on Nánási street was formed on a land surface sloping down to the east, i.e. towards the Danube, and consisted of two parts. The higher (103.6 m.B.f.) and the deeper (102 m.B.f.) parts were separated by a two-metre deep ditch. Settlement on the lower surface could only have been intermittent, interrupted by the rising water level of the Danube in the late Neolithic (4950-4400 BC) [3].  

Recent hydrological features of the III. district. I. Viczián [2]

Human settlement has ultimately accelerated the natural processes of sedimentation of the Óbuda oxbow lake. Settlements had to be provided with access through the former Óbudai Danube riverbed. Farmers were also interested in filling in deeper areas and in the canalization of creeks towards the Danube. The ponds turned into sloughs, the sloughs into marshes, the marshes into bogs, while the wetlands continued to shrink. The climatically and tectonically incised main branch of the Danube have disappeared for good after the last layer of gravel in the marsh. Danubian floods continued to inundate the area for a long time and accumulated considerable amounts of silt. Thus, one of Budapest's largest islands ceased to exist some 6,000 years ago, and the fact that record floods can sometimes still form open water in the deepest parts of the Mocsáros, does not change this.

Translated with DeepL.com (free version)

MΩMOΣ XI. Research of Prehistory Conference 

Proceedings of the conference held at the BTM (Budapest History Museum) Aquincum Museum on 10-12 April 2019. Link: https://edit.elte.hu/xmlui/handle/10831/85856?key=%C5%91sr%C3%A9g%C3%A9szeti%20tanulm%C3%A1nyok

[1] Gábor Szilas – István Viczián – György Sipos – Dávid Gergely Páll – Zsuzsanna M. Virág – Kinga Rekeczki: The Impact of Fluvial Landscape Evolution on Prehistoric Settlement Patterns along the Danube: An Interdisciplinary Environmental Reconstruction in Óbuda, NW Budapest

[2] Farkas Márton Tóth – István Viczián – György Sipos – Dávid Gergely Páll – Zsuzsanna M. Virág – Gábor Szilas – Dávid Kraus: Environmental Changes along a Former Tributary of the Danube. Interdisciplinary Research in Mocsárosdűlő (Budapest, District III) 

[3] Zsuzsanna M. Virág: Neolithic Humans and the River Danube. The Possibilities of Environmental Reconstruction in an Urban Area. A Case Study (75–77 Nánási Road, Budapest, District III.

[4] Katalin H. Kérdő, Ferenc Schweitzer, (2010) Aquincum : ókori táj, ókori város. http://real-eod.mtak.hu/4508/

Icy Flood in the Sausage Kitchen

MAGYARUL

At the end of the winter of 1893, the owner was forced to repaint his 'Wurstküche' at the foot of the Salt Barn in Regensburg's old town for the second time in a decade after the oddly shaped small building was inundated by the icy floodwaters of the Danube. Along with the renovation, Fritz Schricker had another flood sign made, this time not inside the restaurant's guest area, but outside on the western wall outside, and even had his own name engraved in ornate Gothic letters. It is a recent custom at that time, for if the owners of the restaurant had had a flood sign made after every major flood on the Danube, Fritz Schricker's 1893 flood sign - a little exaggeratedly - would not have fit on the wall. Neither inside nor outside.

Fritz Schricker's memory is preserved along with the flood level in his restaurant, which was owned by the same family until 1990. And since 1893, six more flood levels have been added to one of the world's oldest restaurants, a sausage kitchen called the Historische Wurstküche, or Wurstkuchl as it is spelled in the local dialect. The terrace at Weiße-Lamm-Gasse 3 offers a spectacular view over the Danube, the northern tip of the island of Unterer Wöhrd, with Oberer Wöhrd behind it, the Stone Bridge (built between 1135 and 1146) to the left, and the huge, four-century-old Salt Barn to the west. Despite its central location, the Wurstkuchl was a suburban building in the strictest sense of the word, being outside the Danube-side city walls built after 1320, until 1856, when the city finally succeeded in having the ancient walls, which had been maintained at great expense, demolished. However, the demolition of the wall would have been problematic in the case of the Wurstkuchl, because the back wall of the restaurant was the city wall itself, so a piece of the old Regensburg city wall still stands on this section, commemorated by a plaque. Its 'suburban' location meant that while the imperial city of Regensburg was protected by thick walls from the Danube flooding, the Wurstkuchl was flooded by practically all major floods, but the owners still considered it worth the trouble of saving the equipment and drying out the wet walls, as the oldest crossing of the Danube still standing today provided enough traffic to restart the business again and again. 

The eternal recurrence at this point on the Danube is understandable in a literal sense, just as a sensible person can build on flood plains if he calculates that the expected benefits will outweigh the expected losses. It is uncertain when the first restaurant was built on this site, but there are sources that workers on the Steinerne Brücke (Stone Bridge) used to come here for lunch almost 900 years ago. In 1293, the city council of Regensburg decided to build a wall around the outskirts of the city, extending beyond the Arnulfian walls. The southern inland section was completed by 1320, after which the river bank was fortified, with a two-kilometer stretch of city wall, divided by 15 towers. Immediately east of the Stone Bridge stood the Kräncher Tower, a round tower with a crane on top, which served the interests of trade on the quay. Obviously, the Wurstkuchl restaurant must have been built after the construction of the city wall, as this wall formed the back wall of the inn. Archaeologists have excavated the garbage dumps of the kitchen on the quayside, which show that meat dishes were typically prepared here, with poultry, beef, pork, rabbit, lamb and goat bones making up the bulk of the finds.

In 1616, the old restaurant was demolished to build the Salzstadel, but after the construction of the huge warehouse that still dominates the image of the city from the Danube, they rebuilt a new building on the exact same spot with a trapezoidal floor plan. This curious shape is explained by a special 4 m by 40 cm section of Hans Georg Bahre's 1630 drawing of Regensburg from the north, i.e. from the Danube. This engraving shows two gates on either side of the Wurstkuchl. These gates ensured the traffic circulation between the Danube quays outside the city walls and the city itself. This quay was quite narrow, as it is unfortunate if the enemy has too much space to lay siege. And in such a narrow area, it would probably be difficult for the carts to turn onto the gate without hitting the tavern, so it was obvious to knock down both protruding corners and build the tavern in a trapezoidal shape. 

The trapezoidal shape meant that the interior, which has to share a space with the kitchen, is rather small, with only 25 guests can be seated, and a total of four flood signs adorning the walls inside. There are a further four flood signs on the outside walls; two on the west and two on the north, either side of the entrance to the Wurstkuchl. It is a rare coincidence of cultural, culinary, hydrological and urban history curiosities where the history of the Danube is so dense; the eight flood signs commemorate seven floods in total, as there is a sign both outside and inside commemorating the March 1988 flood. The order of the flood plaques on the walls of the Wurstkuchl is as follows: 

1893. February 16. ice flood ~700 cm

2013. June (4.) 682 cm

1882. December 29. ice flood 670 cm

1988. March 27. (2 signs) 659 cm

1954. July 645 cm

1965. June 643 cm

2011. January 15. 627 cm

Bavarian idyll with Danubian flood signs (1988 and the bit worn one from 2011) (source)

Each of these floods could be the subject of a separate article, as these years do not always correlate with the major floods in Hungary. This is mainly due to geographical distance, as flood curves may flatten towards the lower sections due to lack of replenishment. This is especially true for icy floods, such as the one we are now discussing, the flood of 1893, which was one of the highest floods in Regensburg, while it peaked at a lower flood level in Vienna, it again flooded villages (e.g. Gerjen) in the Hungarian section due to ice jams forming in the river bends. In Regensburg, the sign for the flood that peaked on 16 February 1893 was the highest on the wall of the Wurstkuchl, but it should be added that ice floods are always separated from ice-free floods by hydrology, This is because it is not always clear where the water ends and the ice begins, and in the case of certain structures, such as the aforementioned Stone Bridge of Regensburg, the swelling effect of the ice jams must be taken into account. 

The two highest flood sign on the Western wall of the Wurstküche.

It is therefore unclear on what basis the owner Fritz Schricker put the second flood sign on the wall of his sausage kitchen. The building was presumably inaccessible at the time of the icy flood, and it was only after the flood had receded that the extent of the damage could be seen. In any case, the flood mark is at least one large span higher than the maximum ice-free flood level, which is also the record level on the Hungarian stretch of the Danube, with the exception of the water gauges at Dunaszekcső and Mohács. The 1893 flood in Regensburg, which peaked just 132 years ago, is also a curiosity in that it has already been photographed, with seven of them available on the Hochwasserschutz Regensburg website, three of which are available in higher resolution below: 

Ice jam above the Stone Bridge seen from the Oberer Wöhrd

The flooded Protzenweiher marketplace in Stadtamhof

Ice jam as seen from the Stone Bridge with the Unterer Wöhrd

One might think that modern flood defences are already capable of protecting urban facilities in the most developed states of Germany, but instead of a positive outcome, it should be noted that the largest floods will continue to flood the Historiches Wurtsküche, despite the fact that a mobile dike system has already been built along this stretch. This is because the flood defences are unable to protect the restaurant from the water pressure rising from below. This is not a problem in the neighbouring Salt Barn, where the weight of the huge building can counteract the rising water table, but this is not the case in the small Wurstkuchl, which is forced to let the Danube break into the building, as this would cause less damage. Thus, it is expected that in the centuries to come, the walls of the small sausage kitchen in Regensburg will bear further flood marks, while there will be no fear of visitors being left behind during periods of low water. 

Literature:

• https://de.wikipedia.org/wiki/Historische_Wurstkuchl
• https://www.heimatforschung-regensburg.de/2486/1/1063113_DTL1774.pdf
• https://www.regensburgnow.de/wurstkuchl/
• https://hochwasser-regensburg.tumblr.com/image/52136760733
• https://www.wurstkuchl.de
• https://www.hochwasserschutz-regensburg.bayern.de/dok-historische-hochwasser/galerie-eisgaenge.html
• https://www.regensburg.de/fm/121/hochwasser-bedeutende-pegelstaende.pdf
• https://www.hnd.bayern.de/pegel/donau_bis_passau/regensburg-eiserne-bruecke-10061007?
• Christine Schimpfermann: Hochwasserschutz im Denkmalensemble – Strategien zur Konfliktlösung am Beispiel Regensburg