Shipshape 10
How to build an Anglo-Saxon Ship
Paul Constantine
The Rowing Problem
The ship was/is a rowed craft.
There has always been speculations about whether the ship could be sailed, but it is only possible to engage in such conjecture if there is little understanding of its form and its scantlings. Scantling is a word that indicates the dimensions of pieces of timber and relates to how they function within a structure. In short, whether the structure is strong enough for the job that it has to do. The ship is incredibly lightly built with wafer-thin planking and it is the wrong shape for sailing, with insufficient outboard form-buoyancy to resist the heeling power of a sail. There was no evidence of sail equipment, sail/rope wear or rigging attachments. Masts are in compression, that is, they push downwards on the bottom of the hull with great force. Putting a mast amidships, without changing the hull construction and strength would not be safe and highly likely to result in structural failure. Tholes along the sheer strake are The first oarsperson takes a seat. definitive proof that the ship was rowed.
As the ship had been stripped internally there are no further clues as to how the rowers were accommodated for their task. This generates a series of questions to be asked. Was there a deck to stand on? If so, at what height?
A temporary ply deck being tested in the half-ship model
Did the rowers stand or were they seated? Were there seats (usually called thwarts) to sit on? Did the rowers sit on the thwarts or on boxes containing their possessions? How is a thwart positioned in relationship to a thole? How is a thwart supported? If there were boxes, Height of thwart and relationship to rib? how were the boxes fixed in position?
Then there are questions about the oars, but they can come later.
Prior to the reconstructed ship being built it was easy to skip over such questions, but the higher the planks clambered up the sides, the more pressing the questions became. The questions have to be answered. Ideas about the different options needed to be tested thoroughly and decisions made.
In the early stages of reconstruction full-size half models were made and these formed the basis for experiments, with the level of the decking, the height of the thwarts and location relevant to the thole. Placing the thwarts on the ribs is the natural place to put them.
It is possible to experiment by placing a scale-sized oarsperson in the one-fifth scale model as at the top of the page. It is also of value to look at how earlier and later ship reconstructions dealt with these problems. This is an illustration of Tim Severin‘s Greek reconstruction ship Argo used in the Voyage for the Golden Fleece. The craft was of a similar size to Sae Wylfing being about 50ft overall. Colin Mudie had a hand in both Argo's and the Sutton Hoo ship design investigations. It can be seen that instead of standing or bracing their feet against the deck, many rowers have lodged their feet on the thwart immediately astern of them, to utilise their leg strength. The oars are 14ft long and they have lead couterweights close to the rower's hands to try to even up the weight of the outboard section of the oar.
How much power can be generated by rowing?
Extract from Paul Handley’s paper to RINA . (See Ship. Investigation 4) Following a very detail explanation of the displacement (weight) and shape calculations. Paul said:
Computer-prediction power curves showed that calculations compare quite closely over the expected range of speeds. The approximate predicted power is 10hp at 8 knots, 20hp at 9 knots and 30hp at 10 knots.
It is difficult to estimate the power delivered by a human with an oar, and also there is a question over the number of oarsmen, which may have been 28 or 40, depending on whether or not tholes for oars were originally fitted in the mid-section of the ship, but removed for the burial.
Tim Severin also mentions the calculation for the Argo rowers. The power delivered will vary, but we can return to that shortly.
There is no better test than to sit on a thwart, holding an oar against a thole and pulling it against the resistance of the real thing – water. To do this involved taking t
he half- hull model to the river, but of course, the half-hull model had not been built as a water-tight structure. It would not float or support a crew. A compromise situation was arranged where a structure was placed alongside the river wall and as the tide rose, taking advantage to the opportunity when the tide was at the right level to use the oars. Rowing did not manage to move the quay!
Tim Severin also wondered about the power of the Argo rowers. The power delivered will vary according to the size and fitness of the crew as well as the nature of the oars themselves. ‘Power’ or ‘energy’ can be described in terms of ‘horsepower’ or in what we might think of as ‘electrical energy’ or ‘Watts.’ Measured in Watts an average man might develop 250w, or an Olympic athlete 500w. This is a quarter or a half, of the power used in one 'bar' in an electric heater. It is minute. I can recall admiring the massive, intricately-painted, traction engines in 
the Thursford steam museum in Norfolk (left) that were once used for powering agricultural machinery or driving fairground rides, only to find that they were rated at about 8 -10hp. If you tethered 8 to 10 horses to a vehicle you would have quite a lot of power, but such feeble power in a modern car is totally inadequate. It is not quite a straight comparison as much depends on RPM. Our modern minds are so adjusted to our vehicles, yet we have little concept of the energy in the fuel that we put in them. To make that mental adjustment, imagine loading your family’s holiday cases, all possessions and the pet dog into your car, then loading the family themselves - but not yourself. Go round to the back and push the loaded car up and downhill for 8-10 miles – that should be about the equivalent of a 1lt juice bottle of fuel. That is the power compresssed into the liquid fuel. Compare it to your own 'power'.
Oars can vary enormously. Their length gives them leverage, but the longer they project outboard the harder they will be to balance and to pull through the water. The blade can be broad, but traditional working oars have blades hardly wider than the loom. Narrow blades will be less tiring and possibly prolong the endurance of the rower? The timber used can be some species of softwood such as spruce, which is lightweight and strong. The traditional hardwood is Ash, which is regarded as being able to take the load, but it much heavier. Different timbers prepared for being fashioning into oars
Reading Tim Severin’s book about sustained rowing, he points out that their constant rowing was not popular. He says that the rowing was ‘loathed’. It was very hard physical work and some form of rest had to be arranged. The full crew cannot row all the time, so some ‘watch’ system has to be evolved. He also points out that it is boring and some form of diversion had to be adopted; telling jokes or stories. He says that the most successful diversion was singing; drinking songs, hymns, traditional pub songs and folk songs were good, because of their rhythm that coordinated the strokes, but current pop songs were not so good. Their journey was more of an unrelenting adventure, rather than just a day trip. They were living on board for long periods.
Tholes. The evidence for tholes against which the oars were placed is clear. Briefly, Volume 1 has a dimensioned photograph on page 403. The account says that they can be ‘reconstructed with a high degree of accuracy’. Their bases were 3ft long and they were positioned midway between the ribs; held in place with long spikes. The main point of debate relates to the missing tholes amidships 
where the burial chamber was situated. This issue is discussed from page 413 in Volume 1. Another question to be answered relates to the gaskets or grommets. These are lashings, usually of leather, that hold the oar in place against the thole. They are needed in case the oar has to be pushed (backed) instead of pulled, to reverse the oar power. Such movement is needed when turning the boat sharply. They are attached to the boat and looped around the oar. All craft using tholes had a hole for the grommet, so it needed to be decided where the hole should be placed; through the thole, as with the Nydam and Kvalsund craft, or in some other part of the gunwale/sheer strake structure. Edwin Gifford in making Sae Wylfing used spliced rope loops, but these require the oar to be passed through them from outboard, or withdrawn from outboard due to the wider blade width not passing through the loop, which is an unsatisfactory solution, especially when coming alongside.
The Gokstad Faering
When the Gokstad ship was discovered in Norway it had a small rowing craft associated with it. In modern parlance this might be regarded as a ‘tender’ used for accessing a boat moored away from the shore. The boat was about 21.5ft long by 4.7ft wide (6.561m x 1.438m). The boat is known as the Gokstad Faering. It has been used frequently as the model for building and performance trials by various groups of academic historians. A reconstructed full-size model was used for a pilot study based in Cawsand Bay near Plymouth in June 1973 by a team from the National Maritime Museum. A report of the sea trials, authored by Eric McKee was produced detailing an exhaustive set of performance data relating to all aspects of the trial. There is no better account than this, of the exploration of the rowing process as it has clarity in its descriptions of the many simple practical test undertaken. It is too lengthy to relate here, but some of its conclusions are of great interest
The diagram above shows the relevant dimensions taken for each rower. They rowed in calm water and in rough. They checked the working and recovery strokes, the length of stroke and the action of the different oars. They recorded the speed and endurance of the rowers and set the parameters of tests that can be expected aboard the ship afloat. Rowing is not quite the simple process that at first glance it appears to be, especially when 40 people try to coordinate it.