Navigation sidebar:



  1. Setting of the Dünsberg
    1. Topography
    2. Trade routes
    3. Other monuments
      1. La Tène graves
      2. Burial mounds
      3. Roman sites
    4. Springs
    5. Iron ore
  2. Overview of previous research
    1. Descriptions
    2. Excavations
    3. Early Interpretations
  3. Settlement features
    1. Ramparts
      1. Topmost rampart—banks w-y
        1. Gates 21-23
      2. Middling rampart—banks o-s
        1. Gates 14-20
      3. Bottommost rampart—banks a-n
        1. Gates 1-13
      4. Table: Gates
      5. Strahlenwälle
    2. Sources of water
    3. Platforms
  4. Development of the site
  5. Footnotes

Having examined the general interpretation of oppida, this chapter is taking a look at the available information about the Dünsberg in particular, and of its surrounding area.

Setting of the Dünsberg


Northwest of Gießen (Hesse), we find several hill formations, the highest (497.5 m) and largest is the Dünsberg (see figure Grave Mounds). This hill is a widely visible landmark for the region and its top is occupied by the oppidum.

The Dünsberg is one of the outposts of the Rheinisches Schiefergebirge. The ground consists mainly of silicious slate, but in the east of rubble and loam. The hill has several foothills: the Kleiner Dünsberg with 385 m in the north and in the west the Vorderer and Hinterer Eulenkopf (Reeh, 2001: 3). The distance to the river Lahn in the south and east is about 9 km (Schlott, 1999: 8).

Trade routes

The river Lahn might be seen as one of the trade routes for the site. Reeh (p. 292-298) argues that some old ways are passing the Dünsberg, for example the Weinstraße, leading from Mainz via Gießen further to the north; the Rennweg coming from Koblenz, passing the Lahn north of Gießen and leading further to the west, the Hohe Straße coming from Cologne passing Herborn and continuing to the southwest; and the Koblenzer Straße, which is beginning near Koblenz and ends in Wetzlar. These roads pass the oppidum in some distance, but they would still provide accessible trade routes for it.

Other monuments

La Tène graves

To the west of the Dünsberg in the Krodorfer Forst, about 1 km from the lowest fortification, several La Tène D2 graves were found. The graves were discovered, because of 10-15 cm high rectangular or round banks, which are enclosing the cremation graves. These features, called Grabgärten1, stand out from the forest ground, and were already partly destroyed by grave robbers. (Schulze-Forster, 1997; 1998). More Grabgärten exist in the Krodorfer Forst, but were not excavated. The phenomenon is well known in the regions surrounding the River Lahn and the River Rhine. The Grabgärten date to the Late La Tène and the Roman Period (Römische Kaiserzeit) (Schulze-Forster, 1997). The cemetery in the Krodorfer Forst contains about a dozen cremations, which fits into the prevailing picture of Late La Tène funerary practice of small dispersed cemeteries (Schulze-Forster, 1998).

Burial mounds

Besides these graves, which are contemporary with the late settlement phase on the Dünsberg, we find a concentration of grave mounds in the region. Most of them are south or south west of the Dünsberg (see figure Grave Mounds). Reeh (p. 45) lists 42 grave mounds2, of which 7 are under 60 cm high while the remaining 35 are over 1 m high. Only one of the grave mounds has undergone modern excavation, unveiling finds from the Late Bronze Age (Reeh, 2001: 45). Reeh (p. 45) also states that the mounds appear to be close to the roads in the area.

Roman sites

Only 6 kilometers to the south (see figure Grave Mounds) is the site of Waldgirmes—a Roman civil administrative site, which was built just 10 years after the destruction of the oppidum on the Dünsberg, and was given up 9 years later3 (Rittershofer, 2000).

Also to the south runs the Limes. At its closest—when enclosing the fertile area of the Wetterau—it is only 18 km away from the Dünsberg (Reeh-2001: 4).


A great amount of springs have their source on the Dünsberg. In the west they appear at a height of 360 m, but in the north and east they are closer to the foot of the hill. It is likely that there once were more springs which also bore a greater amount of water than they do nowadays (Reeh, 2001: 14). Three of the springs have been incorporated into the oppidum, they are the Schulborn to the north, the Grinchesweiher to the east and the spring near the Hinteren Eulenkopf in the west (see figure Dünsberg Map) (Reeh, 2001: 142).

Further springs have Strahlenwälle pointing towards them, these are the springs east of the Kleiner Dünsberg and north and east of the Schulborn (Reeh, 2001: 142, fig. 42). The streams east and south of the Grinchesweiher and north of the Schulborn are flanked by ramparts, while the spring of the Wilsbach in the south and the springs in the west have no Strahlenwälle associated with them.

The springs had a wooden frame to keep the water clean. Some of these wooden structures were excavated in 1907. One of the biggest is the Schulborn. In the administrative report from 1907 (Verwaltungsbericht, 1907), which mentions its excavation, it is said:

daß hier ein großer Wasserbehälter angelegt war, dessen Pfosten samt der in Falzen eingelassenen Bohlenwand infolge des schlammigen Bodens größtenteils noch erhalten war. Dieser Anlage scheint eine etwas ältere und kleinere, auch anders orientierte, vorausgegangen zu sein.

Herrmann mentions that the bigger structure was 13.08-13.50 m by 4.20-4.55 m and that not only the smaller (and supposedly earlier) basin but also several wells were overlayed by it. Even though the chronological sequence (the bigger structure being younger than the smaller one) might be questionable. Only a modern excavation and, if possible, dendrochronological examinations could clarify these questions.

Nevertheless following the excavations a model of the Schulborn was built for the Museum in Wiesbaden (Reeh, 2001: 144).

In 1908 further springs were excavated. All of them contained wooden structures similar to the ones found in the Schulborn,

nur waren die Wandverkleidungsbohlen bei den geringen Maßen der Bassins nicht eingelassen, sondern durch gegenseitige Verspannungen in Pfosten gehalten. (Ritterling, 1910: 356)

The Grinchesweiher was examined in 1909, but not completely as the installation proved to be too extensive. Again the report about the excavation is short and not very detailed (Reeh, 2001: 146).

Iron ore

In the west of the Dünsberg iron ores were mined. This certainly happened in the middle ages and during modern times, but the resources might have been important in earlier times as well. Dehn (p. 260) thinks that there could be a connection between the iron ore deposits and the ramparts, which include the Kleinen Dünsberg together with the Strahlenwälle (R and S) near the Hinteren Eulenkopf.

Jacobi (p. 34) also considers early use of these ore deposits. He regards the hoards of iron ingots and the great amount of tools as indications for iron working and iron production in the area, whereby the finished goods might have played an important role in trade.

Slag has been found in the vicinity of the oppidum, especially in the west and the north (Reeh, 2001: 105), but no attempt of dating these finds has been made so far.

Overview of previous research


The oldest references to the oppidum on the Dünsberg are descriptions of borders dating to the 16th century4 (Reeh, 2001; ,Schlott, 1999). Descriptions of the ramparts occurred in the 19th an early 20th century (Reeh, 2001):

Dieffenbach described the two higher banks after a survey, 1878 all three banks are mentioned and roughly measured by Gareis and Zöppritz.

In an article about a millstone found on the Dünsberg Dickorè also denotes the kind of ramparts encountered on the hill, and mentions the Schulborn.

The Dünsberg was mentioned by Cohausen in his article about defensive walls of the region. 1904 the forester K. Behlen gives the fullest description of the hill. Besides his depiction of the banks and the identification of the Strahlenwälle, he suggests that the oldest walls were on the top of the hill, and the youngest ones on the bottom. He also detects the house platforms, and draws the first plan of the fortifications on the Dünsberg.

Behlen reports about some weapon and bone finds from the Dünsberg.


First excavations by the Landesmuseum Wiesbaden were undertaken in the years 1906-09 and 1912. Ritterling and Brenner conducted these excavations. Trenches through the banks revealed almost everywhere a pointed ditch in front of the bank5, but the inner structure of the banks remained unclear (Dehn, 1986; ) Reeh, 2001). Excavations at the Schulborn encountered two phases of the wooden lining, which also had different orientations. A model of them was constructed by the museum of Wiesbaden. The Grinchesweiher contained wooden structures similar to the ones observed in the Schulborn (Dehn, 1986). Besides these two major water reservoirs other springs and cisterns were excavated, which displayed the same kind of wooden linings (Reeh, 2001). Finally several platforms were examined, revealing their artificial nature, posthole settings and drystone walls (Dehn, 1986; Reeh, 2001). In 1916 a plaster model of the Dünsberg with its ramparts was built in the Landesmuseum Wiesbaden. Unfortunately it was not very accurate on behalf of the banks (Dehn, 1986; Reeh, 2001). Otherwise no publication followed the excavations, and some of the records were lost during the two World Wars. This and the fact that the excavation standards were not very good in these times makes the remaining records almost useless for modern research.

In 1951 a cremation was found 50 m outside the ramparts of the Dünsberg to the east. It dates to La Tène D1, the urn contained (besides the human remains) two broaches, parts of a chain and a ring (Schlott, 1999; Jacobi, 1977). Dehn assumes that this is not an isolated grave but rather part of a bigger, so far not excavated, cemetery6.

In 1965 trenches for cables were cut, the finds made during this period are only partly published (Schlott, 1999).

Preceding the construction of the television and radio tower a rescue excavation on top of the Dünsberg was undertaken in 1974. An abundant amount of Late Bronze Age pottery and a bronze hammer were found, but no structures. Still it is likely that there was a habitation site of this period on top of the hill (Dehn, 1986; Jacobi, 1977) . Only a preliminary report of these excavations was published (Schlott, 1999).

In 1977 once more trenches for cables were installed on the Dünsberg. The trenches passing through the banks did hardly yield any new information. Only the topmost wall (banks w-y) seemed to be constructed of wood and rubble, while the others displayed no noticeable structure at all (Dehn, 1986; Janke, 1981).

Since 1999 further excavations have been carried out under the supervision of Rittershofer. So far two trenches have been opened, cutting through the bottommost wall (bank g, see figure Dünsberg Map), covering an area in front of gate 7. Perpendicular to this the second trench cuts through one of the Strahlenwälle (M) (Rittershofer, 1999). Some of the inner structures of the banks could be revealed. Interestingly two phases could be demonstrated for the rampart near gate 7, and the assumption of a battlefield in front of the same gate could be verified (Rittershofer, 2000)7. In addition to that some occupational remains behind bank g, inside the oppidum, were discovered (Nickel, 2001).

Early Interpretations

Certainly the earliest interpretation for the fortification on the Dünsberg was given in 1613 (Reeh, 2001: 50), where it is said that the Dünsberg is the site of a

Schlosses oder Castells [...], wovon viele nicht unwahrscheinlich muthmasen, daß es die alten Catten zum Schutz gegen die Römer und anderer feindliche Überfälle angelegt, andere aber noch wahrscheinlicher glauben, daß hier die Römer, nachdem sie die Catten überwunden, eine Burg gehabt, um dadurch dieses tapfere Volck leichter im Zaum zu halten.

Reeh also mentions an interpretation by Wigand from 1851:

Es wird die Ansicht vertreten, daß sich dort auf dem Kleinen Dünsberg das von Germanicus errichtete “castellum in monte Tauno” (Tacitus Ann. I,56) befunden habe. (Reeh, 2001: 50)

Settlement features

In the following the numbers and letters for the gates, ramparts and Strahlenwälle are according to figure Dünsberg Map. A table showing the differing nomenclatures for the gates can be found in appendix Nomenclature.

The three mentioned plans can be seen in appendix Plans, for easier comparison.


The ramparts consist of tree concentric rings, each of them being composed of the actual bank and a ditch in front of it. They also display a trough on the inside from where material for the construction of the wall has been taken (Dehn, 1958; Reeh, 2001)8.

Generally it is assumed that the topmost rampart must be the oldest, inspired by the Late Bronze Age material retrieved on the top of the hill, the first phase of construction of the rampart is generally assigned to this period (i.e. around 800 BC) (Herrmann, 2000; Rittershofer, 2001). The middling rampart is assigned to the time of 500 BC9, while the bottommost rampart dates to the Late La Tène period.

Topmost rampart—banks w-y

The banks w-y seem to be the best preserved of the three rings. Their height is still 6-8 m on the outside. On the inside the hill is marked by a huge depression from which material for the wall has been taken (Dehn, 1958). The wall must have consisted of the local stone which was filled in to a wooden frame. Reeh states that the length of the bank is 1010 m and calculates the volume of the wall to be about 505000 m3. Traces of where a fire destroyed the wooden structures can be seen superficially in the northwest (Dehn, 1958; Janke, 1981). Janke describes the inner structure of the rampart as follows:

Der Wall schichtweise aufgebaut. Er muß einige Male ausgebaut worden sein. Auf der Außenseite umfangreiche, starke Brandeinwirkung zu beobachten. Nach einem starken Brand muß ein Ausbau erfolgt sein. [...] In drei Lagen, etwa 1 m übereinander, sind eindeutig quer zum Wall verlaufende Holzkohlestreifen (Balken) zu beobachten (Eiche). Im Inneren sind zum Teil lockere gröbere Schotter zu sehen.

The construction described could be of the Preist-Altkönig type, at least this was my assumption when creating the model.

Gates 21-23:

While Dehn sees only two original gates (22 and 23) in the topmost circle, Reeh observed three gates (21-23). Dehn emphasizes gate 23, which consists of overlapping ramparts forming the entrance. The southern end is higher and more massive than the other one, making the whole entrance appear higher. Reeh refers to gate 22 as a simple gap but highlights the fact that a ramp had been built to give easier access to the gate10. The entrance 21 is shallower and thinner than the other two gates. Instead of a 3 m deep cut through the rampart it is only 1.4 m deep. The gap between the parapets on either side is only 11.2 m, which is 2 m less than for the other entrances11. Nevertheless Reeh is convinced that this entrance is necessary for access to the Schulborn and he mentions traces of a path leading from gate 19 up to gate 21.

Middling rampart—banks o-s

The enclosed area is 21 ha, the length of the ramparts 2308 m. An average hight of 6.1 m with a volume of 62100 m3 is estimated by Reeh. The banks follow more or less the shape of the hill, including a plateau-like extension in the east. Dehn thinks that this plateau might not have been included into the middling rampart originally, and that the banks t, u and v could be the remnants of the former layout12. After having visited this part myself, I would rather say that this is unlikely. First t, u and v are very shallow ramparts in comparison to o-s, and then bank p would have to bridge over a very steep slope to join up with bank v. In all other places the ramparts approximately follow the isolines.

According to Reeh banks t, u and v are only 0.6 m high and 175 m long with two gates (19 and 20). Dehn observed only banks u and v, together with their gate 20.

East of gate 14 bank o makes a sharp bend leading uphill and then downhill again. There is also a small bank sticking out in the direction of the Schulborn. Dehn initially assumed that this might be the remainder of a rampart leading down to the Schulborn and enclosing it within the fortification. Reeh on the other hand mentions that after surveying this area he could find neither any traces of banks nor of levelling work undertaken there.

In the excavation reports from 1907 it is mentioned that at one part of the rampart the 4 m wide pointed ditch displayed two points (Verwaltungsbericht, 1907). The banks were again clearly layered, but no further features could be observed (Janke, 1981).

Gates 14-20:

While Reeh sees six gates in the middling rampart Dehn counts only four of them.

Gate 14 is a simple gap in the defence, a short distance from the Schulborn (Reeh, 2001). Gate 15 appears on Dehn's plan (Dehn, 1958) as having one end of the rampart turned inwards—like a Zangentor, neither Reeh, nor I have seen this when visiting the site. Instead Reeh proposes that this is yet again a simple, gap-like gate.

Gate 16 is emphasized by Reeh, because he thinks that the right side of the gate is flattened out. I had the impression that the ends of the rampart were turned inwards. This was pronounced on the northern side of the gate, in the south the marks were much shallower.

The left side of gate 17 was built higher than the right (Reeh, 2001). The next gate mentioned by Reeh is not on the figure Dünsberg Map. He assumes that this is rather a modern gap than an original gate. Gate 18 is also not certain, and Reeh seems rather inclined to call it a gate out of the necessity to have a gate to the west rather than being really convinced by the features in the landscape.

While gates 19 and 20 are not further mentioned by Reeh, Herrmann describes gate 20 as a gate with overlapping ends.

Bottommost rampart—banks a-n

These ramparts enclose an area of about 90 ha (Dehn, 1958), the Kleine Dünsberg in the northwest is incorporated into the walls, which gives them a length of 3628 m. Dehn suggested that maybe the original plan was to incorporate the Hinteren Eulenkopf into the fortification as well, but concedes that the two banks leading towards it might also be counted to the Strahlenwälle.

As an average the walls are still 3.8 m high from the outside. Their modern volume is 46800 m3 (Reeh, 2001). Two meter deep ditches were found in front of the banks in 1906/7 and 1999 (Verwaltungsbericht, 1907; Rittershofer, 1999), which were about 4.50 m wide (Janke, 1981).

The construction of this bank is known in greatest detail through the excavations in 1999-2001. The construction follows roughly the Kelheim type. The main difference is that no drystone walling was used for the front, only few stones have been found so far, which were situated in the lower parts of the walling, otherwise the front seems to have been made up by condensed slate rubble intermixed with loam (Nickel, 2001). The second difference is that we have clear indications of the anchorage for the palisade-like front. Every second upright beam is held by three smaller beams which acted as anchors. The anchors were clearly visible in the field as they were preserved as hollows, which would connect with the palisade beam at an angle of 63 degrees. The front of the wall was slightly leaning inwards (similar to the walls at Kelheim (Leicht, 2000))13.

The rubble and earth inside the wall appear layered, the beams on the face of the wall were about 50 cm in diameter, the anchors about 20 cm. At the place of the excavation, two phases were visible. Remnants of the first wall are visible further to the north, its 2 m deep ditch was at a later stage filled up and in its place the new wall was built (Rittershofer, 1999). The latest results show that the course of the second wall in relation to the first has been changed: while in the first phase the ditch was running northwest to southeast, the second wall runs from west to east. Further excavations would be necessary to see the dimensions of the change in course (Nickel, 2001). In the second phase the wall does not have a ditch in front of it (Nickel, 2001).

Gates 1-13:

Reeh recognizes thirteen gates, while Dehn counts seven, even though he regards gate 9 as a possible location for an eighth one. According to Dehn gates 1, 2, 5 and 8 are gates with overlapping ends, often with one end being raised and more massive in build (similar to gates 23 and 17), while 7, 10 and possibly 13 are Zangentore. Reeh describes gate 1 as a gate with overlapping ends, the western side is 1.4 m higher than the other. He also proposes a second entrance, which he calls gate 1a, which is 38 m to the west of gate 1 and of the simple gap type.

Gate 2 is a wide opening. The eastern side of the overlapping ends is 6 m higher than the western side and also considerably more massive (Reeh, 2001). Gates 3 and 4 are simple gaps (Reeh, 2001), both were not recognized by Dehn. Another gate with overlapping ends is gate 5, here the southern side is 3 m higher and more massive than the northern (Reeh, 2001). Gate 6 is once more a simple gap, but gate 7 is a true Zangentor, as already mentioned by Dehn. According to Reeh the ramparts turn in funnel-like to form the gate structure, this would mean that we have a Type 2 Zangentor after van Endert.

Gate 8 has the form of overlapping ends, with the eastern end raised 1.9 m above the western (Reeh, 2001). One of the most interesting gates is gate 9. The ramparts swing outwards before turning in funnel-like to form the gate. This is also a quite outstanding feature in the rampart14 on the survey-plan produced by the FH-Frankfurt in 198415 (see figure Platforms and Ramparts). On the other hand the short bank in front of this gate (Strahlenwall Q), as described by Reeh, was not observed in this survey. Gate 10 has a similar outline to gate 9 concerning its outward swinging ends which then form the gate structures. It clearly belongs to the type 2 Zangentoren. Gate 11 is marked as a gate by the FH-Frankfurt, but was not considered by Dehn. Reeh describes it as a simple gap in the rampart, the same accounts for gate 12.

Gate 13 is described by Reeh as a simple gap, while Dehn regards it as a possible candidate for a Zangentor. In this case I would go with Dehn's interpretation even though the situation is very difficult to judge, because this area is derogated by the modern road leading past it.

  simple gap overlapping ends Zangentor
topmost rampart 22, 21 23  
middling rampart 19, 18, 16?, 15(Reeh), 14 20, 17 16?, 15(Dehn)
bottommost rampart 13(Reeh), 12, 11, 6, 4, 3, 1a 8, 5, 2, 1 13?(Dehn), 10, 9, 7
Total 11 (14) 7 3 (6)
Overview of the gate types


One of the most prominent features of the site are the banks enclosing the Grinchesweiher. Dehn is convinced that these banks belong to two phases. The older one consists of walls H and D, which are meant to be L-shaped, the younger banks E and I cross the older banks southeast of the Grinchesweiher. While banks E and D are almost parallel, F and G leave a wide gap. They are 125 m apart when they finally join the main wall. The total area enclosed by these banks and the bottommost wall amounts to 3.5 ha (Reeh, 2001). A further bank (F) stretches for 50 m along the the stream, which is fed by the Grinchesweiher and several other springs in its vicinity. Perpendicular to it, coming from the northeast, are two further banks: C and G, which appear to be associated with F but do not join up with it.

The Strahlenwälle A and B lead from the southwest to the northeast. They start at gate 1 and 2 respectively and lead to nearby springs16.

Bank J does not start at a gate, but it leads to and possibly through an area with several springs17. The survey-map of the FH-Frankfurt shows J linked up with bank N, which leads together with banks M, f and e to a completely enclosed area (see figure Platforms and Ramparts). This area is free of platforms (Reeh, 2001), so it was probably not used for living. Strahlenwall K is L-shaped and follows one of the streams near bank J.

Close to gate 7 two Strahlenwälle have their beginning: L and M. L runs from the northwest to the southeast, while M extends to the southwest. In its lower half M is adjoined to N, which has a course from northwest to the southeast. Further south of N and M Strahlenwall O begins, it is leading from the northeast to the southwest. Strahlenwall P is only on the map of the FH-Frankfurt (see figure Platforms and Ramparts), it is situated south of g running from northwest to the southeast. Strahlenwall Q was already mentioned in relation to gate 9. It is very short and blocking the entrance to gate 9 from the east (Reeh, 2001). Extending towards the Hinteren Eulenkopf are the Strahlenwälle R and S, their common source lies near gate 10. They are not only blocking the entrance to gate 11 from the south, but they also enclose springs and seem to embrace the Hinteren Eulenkopf from two sides. This is an area which is rich in iron ores and was largely exploited during the middle ages and in modern times (Reeh, 2001). Strahlenwall T begins close to gate 13 and then follows the line of bank l. On Dehn's (1958: plate 29) plan the bank swings to the northwest, but on Reeh's (2001: 54) map the bank appears to follow the shape of the Kleinen Dünsberg to the northwest. The survey of the FH-Frankfurt loses the bank while it is still close to bank l18.

Strahlenwall U is quite short but it seems to point from bank m towards a spring in the north.

Coming from the Schulborn a short Strahlenwall (V) follows the stream to the north, while Strahlenwall W begins near the Schulborn in the west and runs towards Strahlenwall A.

Concerning the use of the Strahlenwälle Schlott writes:

Die sogenannten Strahlenwälle haben bisher keine schlüssige Erklärung gefunden; vielleicht bilden sie die Ansätze noch nicht erkannter, völlig verschleifter Wälle oder den aufwändigen Beginn weiterer (Sieldungs-?) Areale, die durch Gebücke oder Palisaden begrenzt wurden (?). (Schlott, 1999)[9]{

Reeh generally thinks that the Strahlenwälle had defensive purposes, only the area enclosed by f, M, N and J is thought to be an enclosed meadow, while Q, L and S might have been used to help bringing cattle into the oppidum.

Sources of water

Some of the sources of water inside or close to the oppidum were already mentioned. They are the Grinchesweiher, the Schulborn and several springs in the vicinity (see \ref{subsec:springs).

Reeh mentions cisterns, and includes them into some of his plans (see also figure Platforms and Ramparts).

Bemerkenswert sind noch die im Bereich des oberen Ringwalles auffallenden Vertiefungen. Auf der Westseite, innerhalb der Wehrlinie, befindet sich eine solche Anlage von 5 m im Durchmesser. Talwärts hat sie eine niedrige Umwallung. Ebenso befinden sich auf der Nord- und Westseite, unterhalb des Walles, ähnliche Vertiefungen. Bei letzterer konnte eine Steinsetzung in Lehm gelagert festgestellt werden. Vermutlich handelt es sich um ehemalige Regenwasserzisternen. (Reeh, 2001: 60)

In his second part of the book he just mentions one cistern to the south of the summit. Another supposable cistern is between the highest rampart and the middling rampart in the south. Reeh also identifies a cistern at the outer edge of bank t, and another one just in front of gate 13. None of these presumable cisterns has been excavated, and we cannot be sure what they really are. Especially in the case of the cistern in front of gate 13 caution should be taken, because this is not the ideal place for a cistern, as it would block the passage19.


Both Reeh and the FH-Frankfurt map show some of the platforms on the Dünsberg, but both are incomplete20. Together they record about 800 platforms, which are mainly situated between top and bottom rampart21. Concentrations of occupation can be found in the east and south between the middling and the bottommost rampart, almost no platforms are on the Kleinen Dünsberg, in the area between gate 1 and gate 2 and on the top of the hill22. It is also remarkable that many platforms are located outside the ramparts, especially in the west and near the Grinchesweiher (Reeh, 2001).

Very little has been published about the excavations of some of the platforms undertaken in 1912. Bremer mentioned the excavations and that La Tène pottery and metal artefacts were found, but only one of the excavated platforms could be identified, which is in the east of the oppidum (Reeh, 2001). Further information about the excavations comes from Schumacher, who mentions that platforms in the south and west of the Dünsberg have been excavated, revealing posts of square huts, and drystone walls (Reeh, 2001). The platforms were formed by cutting into the hill on one side and using the rubble to extend the terrace on the other. The platforms have a slope of 60 cm to 1 m from one side to the other, thus enabling drainage after rainfall (Reeh, 2001).

During the excavation in 2001 another platform was cut. It had not been visible from the surface, but three postholes, ceramics, fired clay and charcoal made it clear that this must have been a habitation site. Besides that a trench for drainage of the platform could be discovered (Nickel, 2001).

Development of the site

The archaeological work and some of the surveys 23 undertaken have been described in the previous sections. Additional finds were made illegally with metal detectors and some of these “collections” could be accessed by researchers.

The oldest find from the Dünsberg is a sherd dating to about 3500 BC (Michelsberger Culture), but a settlement in this time is improbable (Dehn, 1986; Bender, 1999). The same accounts for the Bronze Age, to which a wheel-headed pin dates (Jacobi, 1977). The first real settlement phase accounted for by abundant finds is the Late Bronze Age (Urnfield Culture). At that time the top of the hill must have been occupied, pottery and two copper axes have been found 24. Generally the construction of the first rampart is assigned to this period, even though there is no excavation data to support this view (Dehn, 1986; Jacobi, 1977), but it can be said that hillforts in this time were no uncommon settlement type (Jacobi, 1977).

The finds dating to the Hallstatt period consist of a few sherds of pottery only, no metal finds have been made so far. It is certainly questionable, whether the site was occupied in this period at all (Dehn, 1986).

The second settlement phase began in La Tène B2. The finds are concentrated on the eastern spur of the hill, which is enclosed by the middling rampart (Jacobi, 1977). Further finds are situated near gate 8 (Schlott, 1999). It is assumed that the second rampart belongs to this phase, making the Dünsberg part of the Early La Tène settlements, which are common sites from the Mittelgebirge to Bohemia, i.e. the first oppida (Jacobi, 1977). It is also thought that mining and iron smelting were important features for the oppida in this time, giving them an economic basis (Jacobi, 1977). While Jacobi proceeds on the assumption that the Dünsberg was settled continuously from La Tène B2 onwards 25, Dehn believes that the hill was occupied anew in La Tène C2.

The bottommost wall was certainly begun in La Tène C2 and from this time on the oppidum flourished. The Grinchesweiher and the Schulborn were incorporated into the fortification (Dehn, 1986), and the settlement area was extended to outside the boundaries of the ramparts (Jacobi, 1977; Reeh, 2001). Finds from this period include imported bronze vessels, tools and weapons. Iron production and processing took place (Jacobi, 1977). Through the tools we know that different crafts were carried out and also agricultural tasks (Jacobi, 1977; Schlott, 1999). Interesting is the great amount of weapons found on the site (Schlott, 1999).

Two questions were and are still discussed. One is the relation of Celtic and Germanic finds on the D\"unsberg and the second is the end of its occupation. Different interpretations have been offered:

Schlott mentions that in 1917 Anthes thought that the Dünsberg was still occupied in Roman times. In 1930 Kutsch concluded that the Dünsberg was one site in a line of Germanic fortifications against the Romans, finally, with Domitian, the settlement found its end. A thesis which has been dismissed by Dehn, on grounds of missing finds from the site to support it. Dehn himself held the opinion that there was no proof for continuing occupation on the site during the first century AD. He emphasizes the Celtic finds, but mentions that there is also Germanic pottery present at the site. Jacobi states that

die spätkeltische Siedlung ein gewaltsames Ende gefunden hat, das mit dem Feldzug des Drusus nach Germanien in den Jahren 11/10 v. Chr. im Zusammenhang steht. (Jacobi, 1977: 38)

Taking the Roman finds as an indication he argues that the settlement could have not existed after the campaign of Drusus against the Chatti, especially as only 30 km away Drusus constructed the Roman fort of Rödgen.

Schlott criticizes Jacobi for not taking into account that some of the weapons date to La Tène D1 and others to D2, but otherwise agrees with his hypothesis.

Other critical voices have been raised. Schlott mentions Polenz, who claimed that no settlement continuity was needed, and that the Roman finds could date to later times. Mildenberger comes to the conclusion that the weapons dating to La Tène D1 belong to a conflict between Celts and Germans, and that in D2 another conflict between Germans and Romans took place. Schlott also mentions the interpretation of Spehr, who thinks that the weapons do not represent the remainders of a battle, but are indeed remnants of ritual actions.

Schlott himself points out that there are two possible reasons for the concurrence of Celtic and Germanic finds on the Dünsberg. One explanation would be that the Celtic occupation comes to an end at the end of La Tène D1, and later Germanic settlers from the Elbe region arrive. The other possibility would be that no hiatus occurred, but that the Celtic population mixed with Germans from the north. Schlott obviously assumes that no battle took place between Celts and Germans.

Schlott also gives new information on coins found on the Dünsberg and the neighbouring Heidetränkoppidum. Both sites are the centers for a special coin type 26. The coin distributions almost exclude each other, and Schlott wonders whether this might be taken as a proof for a Celtic settlement in the Heidetränkoppidum, and a Germanic population on the Dünsberg.

Generally the end of the occupation on the Dünsberg in 10/9 BC has now been acknowledged, especially after the recent excavations on the battlefield in front of gate 7 (Herrmann, 2000; Rittershofer, 2000; Rittershofer, 1999). The exact relations between Celtic and Germanic occupation remain unclear.

The here outlined previous research outlined here gave a sufficient grounding for some of the reconstruction, but still further correlates were necessary to built a coherent model, as will be seen in chapter Methodology.

1 Grave garden

2 12 to 14 burial mounds immediately south of the Dünsberg have been completely destroyed by ploughing in the early 19th century. They are not included in this number.

3 This was after the defeat of the Roman army in the Teuteburger Forest (i.e. Kalkriese).

4 Staatsarchiv Marburg, Königsberger Salbücher 1569 and 1588, pages 391; 495.

5 In case of the middling rampart (banks o-s) the ditch had two points (Reeh, 2001: 79).

6 As mentioned before, in the Late La Tène small dispersed cemeteries are the rule (see section La Tène).

7 The great amount of germanic/celtic and roman weapons and chariot parts found in this region strongly speak for a battlefield (see also Schlott, 1999). Unfortunately the soil is too aggressive for bones to be preserved, but several horse's teeth have been found.

8 Most of the more recent descriptions and articles of the Dünsberg like Dehn, Bender, Schlott and Herrmann are actually based on the description by Dehn. Reeh delivers an independent account, but is not an archaeologist by training.

9 I do not see a reason for this as allegedly there are very few finds from this period.

10 This has already been mentioned in the excavation reports from 1906-12.

11 The terrain in this region is also very steep, which would make it rather difficult to enter.

12 This is one of the assumptions which gets more and more “certain” the more often it is repeated by other authors throughout the literature, even though the statement was quite carefully phrased by Dehn in the beginning.

13 In my reconstruction I assumed that a facing of wood (or wattle) would be necessary to give the front the necessary strength, even though no finds which would support this theory were made. Known examples of ramparts with wooden fronts are the Staffelberg, Bern-Engehalbinsel, Limberg and the Kegelriß (Leicht, 2000: 137-138).

14 The form of the gate is marked quite accurately on the plan by the FH-Frankfurt, but is not shown as a gate.

15 This plan is so far unpublished.

16 The length of Strahlenwall B is 125 m.

17 In the plan the letter J denotes the banks north and south of the spring area.

18 I took the freedom to extend the line of this bank in figure Platforms and Ramparts, being guided by Reeh's drawing. To view the original plans refer to appendix Plans.

19 What springs to mind is the pit in front of the eastern gate in Manching (see Endert, 1987).

20 A merged picture of both recordings is shown in figure Platforms and Ramparts.

21 Reeh gives numbers for his recorded platforms on page 137.

22 Reeh (p. 107) mentions a comment by O. Vuge, which points out, that on the top of the hill was a fort in 1759. Further destruction through the modern buildings on top of the hill can be assumed.

23 Besides the surveys undertaken by Reeh, others preceded the excavations in 1999. A geomagnetic prospection helped to determine the most promising areas for the excavation, and data from boreholes has been assembled. In 2001 the FH-Frankfurt undertook a survey exercise in a corridor from the top of the hill down to the excavation site, in which more borehole samples were taken. Some of the platforms sketched by Reeh were revisited. (Rittershofer personal comment and Nickel.)

24 Jacobi assumes that at this time metalworking took place on the Dünsberg.

25 Even though he mentions that it is not clear how many finds really date in to the Middle La Tène period, Jacobi is convinced that the Dünsberg—in contrast to other hillforts in the area, which break off at the end of La Tène B2—has a settlement continuity.

26 Forrer 352 is found in the Heidetränkoppidum, while Forrer 351/351a has its greatest density on the Dünsberg.