45 (287)

Figurę 23 Parts (basicaUy ninth century) ofthe central settlement at Hedeby. Notę the brook with larger Street Crossing on a bridge (centrę), other, and especially smaller streets, houses with fireplaces and doorsteps, property boundaries and wells (black dots). (After Schictzcl)

fortresses of King Alfred in the late ninth century occur in a list giving the area (in number of ‘hides’) of provision and support of manpower for the defence of the sites against, among others, the Danes.⁴⁸ If thcsc numbers are applied to Danish conditions, the environment and agrarian Systems not being too different, the following appears. The central settlement of Hedeby, covering some 24 hectares, would nced about 6(X) square kilometres if the ‘hide’ around 900 A.D. is of about the same size as in the eleventh century. Arhus, roughly equivalent to the area of Aggersborg, the largest of the ‘Trelleborg'-type fortresses would have needed up to 250 square kilometres for support with its less than ten hectares. The figures are uncertain, but it is striking ho w well the 600 square kilometres for Hedeby correspond to the area, north-east of the town, most densely settled according to the distribution of graves.⁴⁹ But in addition, the upkeep and defence of the 1 )anevirke walls would require morę land in the same area, beyond the 600 square kilometres for the town itself. Besides, Hedeby may have received supplies by sea.

The findings are in accordance with notions about the population spread round Arhus, for instance, and the size of the early parishes, showing that the towns received supplies from very wide areas and were fully integrated with the surrounding province (Chapter 4 C). We have also noted town products and trade goods to be distributed among the rural settlements in return for supplies (Chapter 4 D).

D. Archaeology and the town

Hedeby has a central role in the study of Viking Age towns, their economy and social development.⁵⁰ Relatively large areas have been excavated, although they only make up five or ten per cent of the total settlement. The eighth-century site, continuing into the ninth, lies to the south of the later walled area, which was unoccupied until about 810, at the time ot King Godfred (according to dendrochronological dating) (Figs 22-3). In the tenth century the south settlement (‘Siidsiedlung’) gave way to graveyards and ditches, probably connected with the construction of the semicircular wali around the central area to the north. The earthen town-rampart, framing some 24 hectares, 550 and 600 metres wide in respectively east-west and north-south, is today between five and ten metres high and displays several phases ot construction. Entrances with gate-towers were in the north and south, and possibly in the south-west too, where a brook enters the settlement, bringing sweet-water and carrying refuse to the harbour basin and the smali inlet. The wal] is difficult to datę exactly, but at some stage it is linked with the Danevirke main wali by the so-called connection wali, no later than mid-tenth century, according to dendrochronology (cf. Fig. 21). The latter ram part has also