Stensatte Jernalder-Kældre i Vendsyssel


  • Poul Kjærum


Stone-set iron-age cellar, stensat jernalder kælder, kælder, jernalder, stensat, cellar, iron age, stone-set, Løgten, bækmoien, dalgaard, water-reservoir, vand-reservoir, late celtic iron age, sen keltisk jernalder


Stone-set lron-Age Cellars in Vendsyssel

At Løgten Mark, at Bækmoien, and at Dalgaard near Tolne, all in the immediate vicinity of Frederikshavn, Hjørring County, excavations were carried out in the years 1956-59 of 9 stone-set cellar rooms of the same type as those previously known from Donbæk 1) and from Bækmoien 2) in the immediate neighbourhood, but otherwise unknown in Denmark.

The investigations on the Løgten site have made it possible to distinguish 2 clearly distinct types within this group of constructions, and have also permitted a tentative reconstruction.

Of the 7 cellars at Løgten Mark (Fig. 1), 3 (B, E & F, Figs. 2, 3, 4) are of a very simple form, consisting only of a stone-set room, dug just under a metre below ground surface, to which access was given by a stone-set passage leading to the chamber in a curve or else a sharp angle. To this group belong all 3 of the constructions known at Bækmoien 2) (Fig. 6), though these possess a niche built out from the back wall of the chamber, and in addition 2 of the cellars at Donbæk 6) 7) and the Dalgaard construction (Fig. Il).

The second group, consisting of cellars A, C, D and G (Figs. 5, 6, 7 & 8), is more complicated. Chamber and passage admittedly correspond to the first group (except that cellars C and G possess 2 chambers), but from the chambers run stone-set channels which lead down to a stream, in one case with a well as an intermediate link (cellar G, Fig. 9).

There are many indications that these channels were not merely constructed as drains. The cellars with channels are all placed on the slope of the ground towards the stream, in contrast to the cellars of the first group which all lie on level ground. Moreover in the case of cellars D and G features were discovered suggesting that the watercourses could be blocked, to lead water into the channels. The channel belonging to cellar G sloped towards the chamber, not away, and ends in a quite narrow stone-set chamber, which must be interpreted as a water-reservoir, and from here a narrow opening about 10 cms. wide leads into the east chamber.

2 of the cellars at Donbæk belong to this group, though the construction of one of these is somewhat doubtful as the details reported are rather meagre. In these cases, too, the channels were originally interpreted as passages 1).

Cellar G in Løgten gave a good basis for reconstruction. That it was preserved in its full height (just under 1 metre) is shown by the door lintels which still remained in situ, and by the fact that very few stones were found in the fill. In the west chamber and in the passage, about 0.5 m. above the floor, large patches of burnt clay were found, and proved in a few places to lie immediately upon carbonised remains of wattling. They are interpreted as remains of an in-fallen clay-daubed wattle wall. Below this was found a layer of sand covering the cobbled floor. On this floor was a layer of round sticks and a piece of squarecut timber, most of which lay at rightangles to the centre of the east wall of the chamber. The floor of the passage was covered with a closepacked layer of charcoal. In the chamber, above the sticks, lay small fragments of scorched birchbark, and in the same level a sticky yellowish mass covered the sticks and the charcoal layer in the whole extent of respectively the chamber and the passage; investigations suggest that this mass is completely disintegrated birchbark. The sticks, charcoal and birchbark are interpreted as remains of the roof.

On both sides of the entrance from the passage to the cellar stood two massive shaped posts, surviving to a height of about 20 cms. above the floor level of the passage. They are assumed to be the remains of a door-frame, but as one of them stands in the centre line of the chamber there is a possibility that it may have borne a centre beam. However, no corresponding post was found at the other end of the chamber, and in fact no other posts or postholes were found, either in the cellar or in its vicinity around the upper courses of the walls. Among the stones there, and only there, were, however, found a few fragments of burnt clay daub and pieces of charcoal.

On this basis the cellar is reconstructed (Fig. 13) with wattle and daub walls above the surface, following the edge of the stone walls along both cellar and passage, and with a roof covered with turf resting on birchbark sheets. The fact that the passages between the entrance corridor and the chamber, and between the two chambers were covered suggests that the roofing was not continuous between the separate sections.

In a number of the other cellars discovery of clay daub, highly disintegrated wood, and postholes at the mouth of the passage bears witness that they have possessed a similar construction.

Similarly it appears that the passages between the chambers in cellar C were covered.

6 of the cellars in Løgten field lie grouped within a very small area, and as all are contemporary it is assumed that there must have been an intimate connection between them. They may be associated in pairs, as there are 3 of each type within the group. They appear to be otherwise selfcontained constructions, in so far as they are not connected with each other, nor are in contact with buildings of any other nature. At Bækmoien one of the cellar passages gives onto the side of one of the ordinary long-houses of the period, but it does not appear that this can have been the case in Løgten field.

In addition to pottery (Fig.15), which is found in large quantities both on the floor, in the fill and along the upper edge of the stone walls, there have been discovered in these constructions a number of hammer-stones, 2 "fire-dogs" (Fig. 14, Løgten cellar B, Bækmoien cellar E and Donbæk), 1 loom weight (Fig. 14, Bækmoien cellar E), and an iron knife (Løgten cellar G).

The function of the constructions is not finally determined, but it is postulated that the first group was either seasonal dwellings or perhaps store cellars. The second group, on the other hand, is assumed to have had a special function, as is suggested by the water supply and the low door openings. What function this may have been must await further research. The old interpretation of the cellars as secret passages or defensive cellars appears to be belied both by their mode of construction and by their siting.

On all sites the cellars are dated on a basis of their pottery to a very short period in the Late Celtic Iron Age, at the transition to the Roman lron Age. They are found neither earlier nor later, and as they are only known from a narrowly delimited area in Vendsyssel it is considered that they must have arisen under influence from abroad. In this connection one may point to the lrish and Scottish earth houses or weems as the closest parallel.

Poul Kjærum.





Kjærum, P. (1960). Stensatte Jernalder-Kældre i Vendsyssel. Kuml, 10(10), 62–89. Hentet fra