"Between Harold and Svein there was war all the days of their lives."
Adam of Bremen, History of the Archbishops of Hamburg-Bremen (III.12)
With only thirteen oars on a side and a crew of about twenty-eight (including a steersman and lookout), Skuldelev 5 was the smallest of the Viking longships—a snekke or snekkja and the type that most closely resembles the larger Ladby ship, which had sixteen oars on a side. Almost fifty-seven feet long and eight feet wide but less than four feet deep amidships, there were only seven strakes on a side (the same number as the Ladby ship, although an eighth one has been hypothesized). Along each side at the waterline, marking the division between the bottom of the ship and its sides, was a stronger fourth strake (Old Norse meginhufr) to which the floor timbers were attached. Over these ribs and braced against the strake by sturdy knees, beams (ON bitis) supported a deck of loose boards, above which was another array of crossbeams, held upright by stanchions and secured by knees, that served as a thwart or bench for the rowers. Stringers that ran along the upper edges of the strakes provided additional support.
All this is visualized more readily by the reconstructed cross-section of the hull from Crumlin-Pedersen. On the upper edge of the sheer strake (the uppermost one), a small section of shield-rack is preserved, which allowed shields, held high by their bosses, to be slotted between the rail and the side of the ship. But they are positioned in such a way that the oarports are obstructed, which suggests (on Skuldelev 5 at least) that the shields were placed outboard only when the ship was under sail or riding at anchor. Unlike the other planks, the sheer strake was not reinforced, either by stringers or knees, so it could be more easily replaced when worn or damaged.
Skuldelev 5 had been built about AD 1030 and repaired thirty years later, with some of its upper strakes taken from other ships. The ash sheer strake came from a boat in which the ribs had been closer together, making it necessary to close the old square oarports and cut new round ones, as one can just discern in the photograph. But it may be, too, that suitable oak planks simply were not available, as one of the pine strakes was forty-six feet long. The reused wood and extensive repairs, both to the keel and bow section, suggest that the ship may have been part of the leiðang (leding), an obligation to provide ships built and manned locally when called upon by the king in time of war (levied ships usually were larger, however, with twenty oars on a side).
Excavated in 1962 (when a coffer dam was built around the sunken ships), Skuldelev 5 is one of five that were scuttled off Skuldelev to block the approach to Roskilde at the head of the fjord, an important trading town where, in AD 980, Harald Bluetooth had constructed a stave church, the first in Zealand. The ships likely were sunk sometime in the decade between 1060 and 1070, during the intermittent war between Harald Hardrada, king of Norway, and the Danish king Sweyn II over control of Denmark. It was in AD 1061 that Harald, blockaded by Sweyn's fleet, dragged his ships across a neck of sand that divided the Limfjord from the North Sea and escaped to fight again the next year, when he defeated Sweyn at the night-long Battle of Niså (Saga of Harald Sigurtharson, LVIII–LXIII). Harold died in 1066 but civil strife in Denmark did not really end until the death of Sweyn's eldest son and successor, Harald Hen, in 1080.
Oak was especially favored in the construction of Viking ships. A hardwood, it has a high tannin content (the word itself derives from the Old High German tanna, meaning "oak" or "fir tree," as in Tannenbaum) which offers protection from bacteria and fungi. And when split along the axis of the tree, it is both strong and flexible and less likely to warp. Trees first were split in half and then again radially (like a pie) into quarters, eighths, sixteenths, even thirty seconds until a suitable wedge-shaped plank was obtained. An oak tree several hundred years old and a yard in diameter could yield at least sixteen such cloveboards and, with skillful cutting with adze and side axe, half again more. No more wide than the radius of the tree, all have the advantage of being approximately the same size. Oaks large enough to provide planks half their diameter, of course, were not as numerous as smaller trees, which were split in half tangentially and only the center two planks used, each one the diameter of the tree itself.
However the tree was cleaved, the intention was to follow the fiber of the wood—and not cut across it (saws were not used). Radially, all the planks follow the grain and have the same cross section, tangentially, only the ones at the very center of the log. Unlike the planks hewn from the trunk, "grown" timber from the limbs of the tree was used for the stem and stern posts, as well as for the ribs and knees that braced the sides of the ship. By using the naturally curved shape of a branching limb, the grain of the wood still could be followed and the shaped pieces remain light but strong.
Too, only the heartwood was used, the outer sapwood being too soft and susceptible to rot—which makes dendrochronological dating of this and other Viking ships approximate, as only the annual rings (when the tree was felled) can be counted. Those of the sapwood have to be estimated and added to that number. For example, there may be about thirty sapwood rings in a tree one to two-hundred years old.
Deciduous trees such as oak, which flower and lose their leaves in the fall, are characterized by vessel pores that conduct water and sap. More specifically, oaks are ring-porous; the wood that forms early in the spring has larger vessels than wood created later in the year, which is more dense and slightly darker in color. It is this difference in growth patterns that shows itself as pronounced annual rings.
In spite of tannin in the bark of the tree, oak does rot. Long exposure in moist soil, peat bogs, or bodies of water is destructive as well. Bacteria begin to break down the cellulose in the walls of the cells and, as they begin to deteriorate, starches, sugars, and salts are dissolved and leach away. Even the stronger lignin, which gives the cell walls their rigidity, begins to decompose and the wood becomes increasingly porous and permeable. Eventually, the lumina inside the cells become so saturated that only the water itself and what lignin remains preserve the shape and structure of the wood. If suddenly exposed to air, the evaporating water causes the weakened cell walls to collapse completely and the wood to shrink or warp, often disastrously—as happened to one of the Oseberg dragon heads.
Waterlogged wood can be conserved, however, if water in the cell lumina can be replaced before the wood begins to dry. Once the pores have been filled with some other solution, the excess water can be removed. This first was attempted in Denmark where, in the mid-nineteenth century, alum dissolved in hot water was used to displace the water in the wood, which then was air dried and permeated with linseed oil. To reduce the sheen of the oil and further protect against moisture, the surface was treated with a matte varnish. The water was absorbed by the crystallized alum, but the wood itself became dense and brittle as a result, the acidic solution and crystals further breaking down the cellulose and lignin. And, because distribution tended not to be uniform or penetrate thoroughly, there were voids and cracks in the wood, which further destabilized it. Alum also is hygroscopic and so has the potential to reabsorb moisture from humid air and revert to a solution; that does not seem to have been a problem in earlier conservation. Ring-porous wood such as oak seems, too, to have been better conserved with alum.
Other techniques, such as glycerol or an application of creosote and linseed oil, were used in the Gokstad and Oseberg ships (although individual objects were treated with alum). In 1959, a hundred years after having been synthesized (about the same time that alum first was used), there were the first applications of polyethylene glycol (PEG) to waterlogged wood. Soluble in water or alcohol, PEG is designated by its molecular weight or mass, lower weights being more liquid and higher ones increasingly waxy. Their application depends upon the degradation of the wood which, if less damaged, is immersed in a watery bath or, if more deteriorated, a waxy one. PEG with a higher molecular weight is required to provides the bulk necessary to support the fragile cell walls (Skuldelev 5, for example, was treated with PEG 4000). These steps also can be combined, the waterlogged wood first impregnated with low molecular mass PEG, which penetrates deeper into the core of the wood (but adds less strength to the cells and is more hygroscopic), and then PEG with a higher mass to stabilize the more degraded outer layers. (To prevent this potential reabsorption of water either by alum-conserved parts or those treated with PEG, the relative humidity in the Viking Ship Museum is rigorously controlled.)
When Roskilde 6 was discovered, it had been under water for almost a thousand years, the cellulose having been almost completely replaced by the water of the fjord. To prevent the cells from collapsing when excavated, the timbers of the ship were kept submerged until they could be cleaned and documented. Then, over a period of several years, they were placed in increasing concentrations of PEG 2000 that soaked into the wood, gradually displacing the water with a solidified wax. Once strengthened, the infused wood was put in a vacuum chamber and freeze dried, the remaining water frozen and the ice converted directly to water vapor, which was evacuated —leaving behind dry, stable wood.
References: The Skuldelev Ships I: Topography, Archaeology, History, Conservation and Display (2002) edited by Ole Crumlin-Pedersen and Olaf Olsen (the definitive study); "Conservation of Roskilde 6" (2013) by Kristiane Strætkvern, in Viking (exhibition catalog) edited by Gareth Williams, Peter Pentz, and Matthias Wemhoff, pp. 234-237; "Conservation of Waterlogged Wood" by D. W. Grattan and R. W. Clarke, in Conservation of Marine Archaeological Objects (1987) edited by Colin Pearson, pp. 164-206; Vikingeskibsmuseet, Roskilde website (which says that Skuldelev 5 was constructed about AD 1040 and "shortly after AD 1024"); "Navigating Conservation Strategies: Linking Material Research on Alum-Treated Wood from the Oseberg Collection to Conservation Decisions" (2018) by Susan Braovac, Caitlin M. A. McQueen, Malin Sahistedt et al., Heritage Science, 6(77) (open access); Adam of Bremen: History of the Archbishops of Hamburg-Bremen (2002) translated by Francis J. Tschan.
