Gevninge – leddet til Lejre

  • Jens Ulriksen
Nøgleord: Gevninge, Lejre


Gevninge – the gateway to Lejre

Gevninge is one of many Danish villages characterised by having extensive modern housing estates built around a medieval core. The oldest part of the village, with a late Romanesque church, lies on the west side of a small river, Lejre Å, about 2 km from its mouth at Roskilde Fjord (fig. 1).

Both in the 1880s and in the 1970s, remains of human skeletons were found in Grydehøj to the west of the old village core (fig. 2). These clearly originate from burials, but the finds are undated. In 1974, remains of an inhumation grave from Viking times were found a short distance from a sunken road which, up until the 18th century, was part of the main road between Kalundborg and Roskilde. In 1979, a gilded bronze strap-end mount from the 8th century AD was found less than 200 m south of the sunken road, but it was first in the winter of 1999-2000 that settlement remains from Viking times were discovered.

The archaeological investigations
The excavation in 2000 uncovered 3600 m2 of settlement remains; these had been heavily damaged by site development in the 1960s and 1970s (fig. 3). On the basis of the evidence available, it is impossible to determine whether these represent several phases of a single farmstead or a portion of a larger settlement. The absence of any traces of structures in the northern evaluation trenches indicates that the settlement did not extend to the north of the sunken road where the graves were found. The terrain falls relatively steeply away from the excavation to the east towards Gevninge Bygade and, although it is possible, it seems rather unlikely that the Viking Age settlement extended in this direction. Relative to the topography, an extension to the south and west seems most obvious.

There is no doubt that the site should be assigned to the Viking period. House I is unlikely to be earlier than 10th century (fig. 4), whereas the rectangular pit-house belongs to the end of the same century or the subsequent one. House II (fig. 5) and the other pit-houses are – typologically – less useful for a precise dating of the site. The metal artefacts, including the strap-end mount and a handful of coins from the time of the Civil War, span the period from the 8th to the 14th century, but the majority belong in the 9th-10th centuries (figs. 9-13). Pottery is the most common artefact type and occurs as un-ornamented flat-bottomed settlement wares and Baltic ware (fig. 8). The former have typically inwardly curved rim sections, the sides of the vessels are un-ornamented and they are generally bucket-shaped (fig. 14). The Baltic ware pottery is characterised by more angular rims, which have often been finished off using a wooden shaping tool. Decoration is mostly in the form of encircling grooves, waved furrows and a series of slanting or circular impressions. Compared with the other finds from the structures, the Baltic ware from the excavation belongs in the latter part of the 10th century and up into the 11th century.

Traces of crafts were not conspicuous. In one pit-house there were several un-fired clay loom weights, while two deep postholes in the bottom of another pit-house are interpreted as the base for a loom. The distaff whorls and – possibly – the few bone and antler needles also belong to textile production (fig. 7). Iron slag, which definitely was not one of the most conspicuous aspects, originates from “fire-based” crafts. Textile production and iron working are the crafts typically seen at agrarian sites, with the former occurring most frequently.

On the basis of the buildings, the traces of crafts and the majority of the finds, the site must be categorised as an average farmstead from Viking times. The site did, however, include four unusual finds: a gold armring (figs. 12 and 13), part of a gilded bronze helmet (fig. 10), a bronze bucket and a winged spearhead. These finds give food for thought, nourished by Gevninge’s location in the landscape, combined with its proximity to the legendary Lejre.

A main transport junction
The area south of Gevninge is characterised by a series of branching streams which meet at Gammel Lejre and continue towards Roskilde Fjord in the form of Lejre Å. To the west and southwest there is an area of about 50 km2 with a more-or-less pronounced moraine landscape. Large parts of this have lain through historical times as rough ground, common and forest. This landscape type forms a very clear contrast to the area east of Lejre Å - a flat and fertile plain extending out to the Øresund and Køge Bugt. In landscape terms, this is a border area, running north-south, where crossing points had to be chosen with care. Gammel Lejre, which from the 5th to the 10th century was an important chieftain’s or royal farmstead with magnificent halls, huge long-houses and a cult site, is well-suited to the passage of east-westbound traffic (fig. 15). In the flat terrain to the east of Lejre Å, maps from the end of the 18th century and the beginning of the 19th century show no road network prior to the construction of two highways in the second half of the 18th century. These run in a straight line from Roskilde to Ringsted and Kalundborg, respectively. Between them, Ledreborg Allé can be seen; it was constructed at the same time and probably replaced a road running eastwards from Gammel Lejre. To the west of Lejre Å, the undulating landscape, with its numerous small, steep hills, small lakes, watercourses and wetlands, presented greater challenges. There was an alternative crossing point about 4 km to the northwest, close to the fjord. Today, this place is called Borrevejle, which means “the ford at the edge” (fig. 16). From Borrevejle, the road led to Gevninge and – via the sunken road to the north of the Viking Age settlement – down to Lejre Å. Here lay the ford Langvad, from where the road ran eastwards, south of Lyngbjerg Mose, towards Kattinge. The fact that the roads around 1800 led towards Kattinge is linked with the opportunity here to cross the system of watercourses and lakes which extended from Gammel Lejre and past Kor­nerup to a lake, Store Kattinge Sø, by Roskilde Fjord. At both ends of the lake there were lock bridges to allow passage. Store Kattinge Sø was originally a bay which was dammed in the High Middle Ages so that the water level today lies at +2.5 m. In Viking times the lock bridges at Store Kattinge Sø did not exist, the amount of water on the Kornerup Å drainage system was therefore less, and the possibilities for passage were decidedly different.

The road eastwards from the ford in Gevninge could well have gone via Kattinge and crossed the watercourse between Lille Kattinge Sø and the bay. Around 1800, the road continued through Kongemarken, where a Viking Age inhumation grave, a Christian burial ground from the Late Viking and Early Medieval times, as well as remains of a settlement from the same time, have all been found. From here, the road swings northwards, across Gedevad and onwards to the east to the bishop’s thorp, Bistrup, and the village of Bjerget (St. Jørgensbjerg) with St. Clemens’ church on Roskilde Fjord. Neither of these two settlements can, with certainty, be traced back to before AD 1000. It is therefore an obvious possibility that eastward traffic from Kongemarken took a more southerly route, which – perhaps – is indicated by settlement remains and stray finds between Roskilde and Svogerslev Sø (see fig. 16). In this respect, it is worth mentioning that the two stray finds from Viking times from the Borrevejle area lie in association with the old road routes. Similarly, the small hoard of silver rings from Lyngbjerg Mose was found where the road from Gevninge to Kattinge ran from about 1800.

From the above, it is apparent that there were two significant possibilities for the passage of east-west land traffic in the Gevninge-Lejre area. Both have topographic advantages and disadvantages, and identification of one as being more important than the other can be based on no more than a guess. However, inclusion of the waterways does contribute a new angle when addressing this question.

The sea route to Lejre
The Isefjord complex comprises a western and an eastern branch which both extend more than 35 km inland into Zealand. The western arm, Isefjord, is deep and wide and only has narrow passages around Orø (fig. 17). Despite the fact that Isefjord is the most accessible route from a seaward perspective, it is unlikely to have been the route taken by people travelling to Gammel Lejre. The distance over land to the Isefjord is almost three times as great as the shortest route between Gammel Lejre and Roskilde Fjord, and more than half of this distance comprises gently undulating rough ground with numerous ponds and wetlands.

Roskilde Fjord is characterised by narrow navigation channels and variable water depth, but these naturally-determined sailing conditions would not have been a problem for people who knew the fjord. The bay, Lejre Vig, is the place closest to Gammel Lejre. The sea route leading to the bay is protected by a natural feature – a transverse bar, which extends from Bognæs in the south to Selsø in the north. The mouth of Lejre Å is, in topographical terms, a well-suited site for a landing place, but there is a lack of archaeological evidence for the existence of such a feature. Given the lack of a demonstrable landing place by the fjord, attention can be focussed on Lejre Å as being Gammel Lejre’s link with the sea.

Streams and rivers as travel routes
Today, very few watercourses in Denmark appear as being navigable. A very great proportion of them no longer have a natural appearance or water flow. This is primarily due to intensive efforts during the last 200 years to drain wet meadows and fields. Any evaluation of the navigability of a watercourse in Viking times is associated with a number of variable and, in part, unknown factors. Accordingly, any conclusions are vitiated by a degree of uncertainty, not least in the case of smaller watercourses. The width and depth of the stream or river is decisive in determining the size of vessel which can be navigated. The fall and natural course of a watercourse, which in places is sharply meandering with a variable water depth, will be limiting factors relative to the size of the vessel which is able to pass (fig. 18).

The appearance of Lejre Å on maps from the 19th century can give some indication of the conditions prior to the time when drainage and water extraction were initiated. It seems that the course of the stream was relatively straight from its mouth up to Gevninge. However, at Gevninge Church there was a very sharp turn and this is still in existence. To the south of the village, the stream is considerably narrower and substantially more winding. Particularly from Kornerup and southwards towards Gammel Lejre, the course is, in places, strongly meandering. Overall, the stream has a fall from Gammel Lejre to its mouth of 7 m, which corresponds to a gradient of 1‰. The fall is not, however, evenly distributed. From Gammel Lejre, and about 1.5 km down its course, the stream falls 2.79‰, whereas the fall over the next 750 m is 1.31‰. From here to the ford in Gevninge, the fall is 0.5‰, with the last section to the mouth of the river having a fall of 0.34‰. Ole Crumlin-Pedersen has suggested that a watercourse is navigable – all things being equal – as long as the fall is less than 2‰. Alone on this basis, it is unlikely in the past that vessels sailed all the way to Gammel Lejre. It is therefore an obvious possibility that Gevninge was the place where the change was made from waterway to roadway.

The distance from Gevninge to Gammel Lejre is 3.7 km by road, as shown on maps from around 1800. The road departs from an area where Viking Age settlement has been excavated and it follows the contours of the landscape in such a way that steep passages are avoided. The route taken by this road, rather than the river, constitutes the probable link between the two places.

Gammel Lejre was not established at some random place in the landscape. With regard to resources, it was a border area between the hamlets of the Eastern Zealand plain and the Central Zealand forest settlements. In addition, it provided a satisfactory, potential crossing point east-west over the steam systems from the south. There is archaeological, legendary and historical evidence showing that Gammel Lejre was a very special place in the Late Iron Age and Viking times. This special position arose from its role as a cultic and power-political centre.

The same situation was probably the case at the Tissø complex in Western Zealand, which was established at the beginning of the Late Germanic Iron Age. Tissø lies slightly more than 6 km from the coast, and both its name and finds from the lake demonstrate the cultic significance of the site. Almuth Schülke has pointed out that the Tissø complex lies virtually on an island, with the lake to one side and wetlands and watercourses to the other. Access to Tissø was made difficult by natural barriers in the landscape which conferred exclusiveness and – not least – the possibility of controlling traffic to the settlement.

The topographically determined limitations on potential access to Gammel Lejre are not as clear as in the case of the Tissø complex. Watercourses and wetlands to the south and east form a natural border, and the rough ground of the common landscape to the west contains its own obstacles. None of these barriers was insurmountable but they could well have functioned as a border zone around Gammel Lejre. In the area of common from Borrevejle in the north to Ledreborg Castle in the south, a couple of settlements have been demonstrated along with three graves and a few stray finds from the Roman Iron Age. Similarly, in the Middle Ages there were at least five thorps here, which were later abandoned. For the central period relative to Gammel Lejre, the 5th-10th centuries AD, there are no finds from this area. It was not necessarily a conscious choice that the area lay abandoned. The same tendency to abandon poorer soils at the beginning of the Late Iron Age can be seen elsewhere, such as, for example, in Nordskoven at Jægerspris and on Halsnæs at the northernmost part of Roskilde Fjord. Neither is it unusual that areas such as these were then re-occupied for thorp settlement in the Early Middle Ages. This does not, however, change the fact that the area to the west of Gammel Lejre appears to have lain as a wilderness in Viking times. Apart from one artefact with no details of its exact find spot, there are no recorded finds from the Late Iron Age bet­ween the central site and Elverdamsåen, a watercourse lying about 10 km to the west.

Access to Gammel Lejre was obviously regulated so that approved people could enter and intruders were held at a distance. Gevninge was a link in this invisible fence. Gevninge is located where roads running east-west meet to avoid Central Zealand’s areas of hilly rough ground, and where watercourses could be crossed relatively unproblematically. Furthermore, Gevninge was a landing place and offloading point for vessels that were able to enter the lower part of Lejre Å. Larger vessels could perhaps have lain at the mouth of the stream or innermost in Lejre Vig, but from here people would anyway have been directed to follow the road from Gevninge to Gammel Lejre.

Seen in the light of this situation, Gevninge could have been the home of the Lejre King’s entrusted servant. He not only controlled the traffic through the area and access to Gammel Lejre, he also represented the Lejre king and, on his behalf, received distinguished personages and – who knows – perhaps escorted them to important meetings in the exclusivity of the magnificent hall. With this position in society, Lejre’s gatekeeper probably received gifts of golden rings, magnificent weapons and vessels from Lejre’s pugnacious king.

Jens Ulriksen
Roskilde Museum

Ulriksen, J. (1). Gevninge – leddet til Lejre. Kuml, 57(57), 145-185. Hentet fra