Skuldelev 1

When presented to the public this high sided and beamy ship was different from the prevailing notion of Viking ships as slender war ships. Skuldelev 1 was a cargo ship with a cargo-capacity of around 24 tons which allows for transportation of bulk goods. Its building date in the years 1030-1050 AD puts the late Viking age economy into perspective. The building material was mostly pine, and dendrochronological analyses have shown that it was built in Norway. Together with two other ships and filled with stones it ended its life as part of the first phase of the Skuldelev blockage in the Roskilde Fjord in eastern Denmark. The blockage was later reinforced with two more ships, all are clinkerbuilt in the Nordic tradition. In 1983 the full scale replica "Saga Siglar" was launched in Norway, and this ship succesfully circumnavigated the globe.

The Skuldelev Blockage - excavation and research history

The excavation of the Skuldelev ships marked the start of maritime archaeology in Denmark. The blockage in the Roskilde Fjord had been known by local fishermen for some time, and in 1924 they recovered the keelson from what later became known as Skuldelev 1, and this find was reported to the National Museum of Denmark. In 1956 sportsdivers recovered a piece of framing which was handed over to the museum, and this in turn led to the large-scale excavation project.

The blockage is situated about halfway down the more than 40 km deep fjord that cuts into the island of Sealand in a North-South direction at a place where the depth is particularly shallow. This was to prevent enemy ships from attacking the town of Roskilde. Here the natural passage Peberrenden was blocked by the sinking of three ships filled with stones. Apart from wreck 1 the large cargo ship, wreck 3 a smaller cargo ship and wreck 5 a warship of medium size was sunk. This took place in the years 1070-1090 AD. Some years later in 1100-1140 AD two more ships were sunk to renew or reinforce the blockage: a very large warship initially believed to be two wrecks thus dubbed wreck 2/4 and a smaller vessel wreck 6. All ships were stripped from any rigging and equipment before they were positioned in the Peberrenden passage and filled with stones.

The archaeological work started in 1957 with the excavators wading in the shallow parts of the blockage and scuba-diving where the water was deeper. The work included mapping of the extension of the stone covered areas and removal of stones. The wreck was partly exposed where the fishermen had dug into the stone barrier to make a navigable passage. More parts of the wreck came to light and eventually another wreck was discovered, this time built of oak.

The following year the work continued and wreck 3, the small cargo ship was found and partly exposed. Also in 1958 the unexplored northern part of the blockage yielded more timbers, and this was called wreck 4. The timbers here later proved to be part of wreck 2, which thereafter has been known as Skuldelev 2/4.

The 1959 season carried on in the northern part resulting in the dicovery of yet two more wrecks, nos. 5 and 6. This year the scuba-gear was replaced by surface supply equipment, which greatly increased the efficiency of the diving. It had by then been realised that it would be impossible to raise the wrecks by diving. One of the goals of this season was therefore to establish the extent of the wreck area in order to enclose it in sheet piling. In this way the water could be pumped out and the wrecks excavated on dry land.

After the necessary funding had been provided, the cofferdam was built during the early summer of 1962 and the excavation completed during four months of work. A team of archaeologists and students were engaged in this pioneering operation, making use of methods invented especially for this project. The stones were removed manually leaving the wreck area as sandy fjord bottom. The wrecks had all been flattened out under the weight of the stones when the clinker nails in the planking corroded away, so it was fairly straightforward to excavate the wrecks from wooden gangways that could be moved as the uncovering progressed. No metal tools were used, instead the waterlogged timbers were sprayed free from sand and mud with water.

When the wrecks were fully uncovered they were mapped using photogrammetry, just as all wrecks were thoroughly photographed to record constructional details. During the excavation the timbers were kept wet at all times by means of garden sprinklers. To involve the public in this fascinating process one of the sides of the pentagonal cofferdam had been provided with a platform, from where the work on the wrecks could be seen. This platform continued in a jetty from where regular ferry traffic took place.

The recovery of the timbers had to deal with the fact that they were all waterlogged and very fragile. Framing was more consistant than the thin clinkerlaid planking. Supporting strechers of masonite proved to be well suited for keeping all the fragments of a plank together. When a piece of timber had been successfully lifted up from where it had rested for 900 years, it was rinsed and put into a plastic foil tube which was sealed in each end. In this way the wood was prevented from drying out, and all the fragments were kept together.

Constructional features of the Skuldelev 1 ship

The oak keel was entirely preserved and had a length of 12.1 m, it was rabbeted to give the garboard a fairly steep deadrise. Parts of the sternpost were preserved, and these reveal that it had a very complex design consisting of three individual pieces; the first recieved the hood ends of the first two strakes and most of the third, the second accomodated the top of third strake and strakes 4 and 5. This second part of the sternpost was a continuation of the first part and likewise unrabbeted. The third part was missing but can be reconstructed on the basis of the planking, and it made the whole arrangement quite remarkable. This was a V-sectioned timber with carved out continuations of strakes 6 to 12, so to say the upper half of a sternpost of Skuldelev 3 type.

The planking was made of pine, with some repairs made of oak. Each strake was averagely made up of four planks. The scarves were mostly of the short diagonal smooth type that prevail in Viking age and early medieval shipbuilding in Scandinavia, but in some of the upper strakes the planks were joined with long tongue scarves in particularly curved areas. Viewed in section in the reconstruction the planking divides the ship in a bottom and an upper part connected by the fifth strake which is much steeper than the previous ones. In this way it resembles a meginhufr as seen in older ships like Oseberg and Gokstad from Norway.

Framing consisted of fourteen main framestations and some additional frames in the stem and stern areas and aft of the mast. The main material was lime-wood. The beamy middle part of the ship was an open hold, and here the framing consisted of floor timbers covering the first five strakes, and bites with knees covering strakes 6 to 11 and occasionally 12. The midship framestation was immediately forward of the mast step, and here a cross beam rested on the bite-knees which ended on the edge of strake 9, and yet another cross beam on top of strake 11. This top cross beam was found at four more framestations evenly distributed in the ship. Forward and aft of the open hold the ship had been decked, the deck planks ran lengthwise and rested in notched beams on top of the bite-knees. A longitudinal stiffener ran all along the inside of the hull at the top edge of strake 11 where it also formed lodging knees for the uppermost cross beams.

As typical for ships of the Nordic tradition the ship had been rigged with a single square sail. Hull elements of the rigging was a large timber in the fore with three sockets for a tacking spar, and side frames aft of the mast step to which the shrouds were fastened. Since little of the starboard side was preserved, nothing remained of the side rudder.
As reconstructed the ship had an overall length of 16.3 m and a beam of 4.5 m. It had a height amidships of 2.1 m - altogether a sturdy seaworthy ship with a considerable cargo capacity of around 24 tons.The rig has been reconstructed with a sail area of 80 m2, practically in the shape of a perfect square.

Dating and interpretation of the Skuldelev 1 ship

Several samples have been cut for dendro-chronological dating, and a building date in the interval 1030-1050 AD has been established. It has been repaired twice with oak planks, first with wood from the Oslo region in Norway and later with wood from Scania in present day southern Sweden. These repairs have been dated to the interval 1060-1070 AD. The provenance for the building has been determined as the western part of Norway.

The ship was a cargo carrier probably of the type knorr. Later finds have revealed that cargo ships from this period could be considerably bigger, e.g., the Haithabu 1 wreck. Ships of such a size could carry 40-50 tons - twice as much as Skuldelev 1.