By Jens Christian Moesgaard and Volker Hilberg

A group of coins with a distinctive cross motif are attributed to Harald Bluetooth (c. 958-987), but where were they actually struck? Over the years the research has come up with a number of different suggestions: Lund, Roskilde, Odense, the ring fortresses and Jelling. Even Wolin in Poland has been suggested, because many such coins have been found south of the Baltic. However, they are unlikely to have been produced here; instead the finds are more likely to indicate close connections with Denmark. 
 
The large Viking trading town of Hedeby near the Danevirke fortifications has also been put forward as a possible source for the coins. However in 1966 the Swedish numismatist Brita Malmer rejected this option in her thesis after a thorough study of all the coins. Not a single one of them had been found in Hedeby or its surrounding area.
 
However, over the past few decades a number of coins have been found in Hedeby! As early as 1979 an example was produced from a large excavation of the market town’s harbour at Hedeby Nor, and since then five more have been found! From 2003 metal detectors began to be used within Hedeby’s semi-circular defences. Amateur archaeologists from Bornholm were brought in because of their experience and expertise in metal detector use. Many finds have been produced, including two of Harald’s cross coins. During a small excavation three more coins were found in a burnt down pit-house. 
 
Thus with the six cross coins found at various locations in Hedeby, we can no longer say that the coins are not found here. Perhaps they were even struck in Hedeby? It would be an obvious choice as there was a long tradition in the town of coin making and coin use. In the rest of Denmark no other coin making centre is known of at this time and the coin production at Ribe is of a much earlier date.  
 
Historians have traditionally thought that Hedeby was under German control from 974 until 983. But if the cross coins were struck for Harald Bluetooth in Hedeby in the 980s it may be that we should revise this hypothesis.