The Gedesby ship under sail


The replica of the Gedesby ship, the Agnete, on the building slip at the

Medieval Centre, May 1995. Photo Ole Malling.

Ship-finds are good sources of information about societies of the past, because they tell us, more clearly than most finds on land, about the nature, scale and significance of communication and transport. But our knowledge of the properties of the ships of the past is patchy, and this makes it difficult to make full use of this source-material.

How much cargo they could carry, in what waters and at what times of year they could operate, and how large a crew they needed: these are central questions which cannot be resolved by traditional archaeological methods. For this purpose calculations, and better still experiments with carefully-reconstructed replicas in full scale, are a useful and even indispensable tool. Experiments of this type have been carried out by the Viking Ship Museum and the National Museum in Roskilde since 1982, and they have had considerable effect on e.g. our views on the sailing-abilities of the Viking ships.

The excavation of a small medieval wreck at Gedesby in south Falster in 1991 provided a basis for subjecting a medieval craft to critical analysis. The Gedesby ship is from the end of the 13th century and has been discussed, along with other similar ship-finds from Lolland-Falster, in earlier editions of this Newsletter (nos. 1, 2 and 3).

The reconstruction process

Shortly after the excavation, the Medieval Centre at Nykøbing Falster proposed that the National Museum and the Medieval Centre should jointly make a replica in full size of this well-preserved ship-find. And just such a project, of joint analysis and reconstruction, in fact came about, in the years 1992-95, after all parts of the wreck had been documented in the Wood Preservation Unit of the Conservation Department in Brede in 1991-92.

The first stage was the building of a half-model of the hull in cardboard, 1:5, cut out in accordance with the shape of the planks; then a whole model on the same scale was built out of lime-wood. The few pieces of the ship which were missing from the hull were added on the wooden model, and this model also formed the basis for the reconstruction of the internal arrangement with half-decks fore and aft, the rigging and the rest of the ship's equipment. The model was a creditable representative of medieval ships in an exhibition held in 1993-94 which reviewed the achievements of Danish archaeology over the last 25 years, and since then it has come to occupy a permanent position in a local-history exhibition at Bøtø Nors pumping station, near the finding-place of the original ship.

The building of a replica in full scale was begun in the Medieval Centre in 1993, and because of the extensive work already done on the models most of the problems had already been settled, so the work progressed steadily, using the original wood-working techniques. On 1 June 1995 the ship was named Agnete - after an old Danish folksong about Agnete and the Merman - and it was launched ready for rigging and trials.

The replica sets sail and picks up speed in the waters where the

original ship once sailed, 700 years ago. Photo Werner Karrasch.

Sailing-trials

The sailing-trials took place from 28 August to 2 September 1995 in the sea near Gedser, on the landward side of the notorious Rødsand sandbank. There was a strong wind which created waves up to 1.5 m high, so the ship's seaworthiness was well and truly tested. Another good reason for this choice of location was that the original ship must certainly have sailed in this area, about 700 years ago; as was the case with the trials of the Viking ship replicas, we tested the Gedesby ship in its home environment.

The test equipment consisted of a Danish-made colour-track-plotter, linked to a GPS navigator, an anemometer on an escort-boat, and a VHF radio, a secretary with a pencil, and a log book for the trials. During the sailing the track-plotter drew a coloured stripe on its screen to mark the ship's true course over the sea-bed, taking into account deviation and set of the current where applicable. The secretary had to be alert and nimble to work the plotter, set 'event-markers' on the course-line, e.g. to mark a turn or a change in the wind, to note down the wind-reports coming from the radio from the anchored escort-boat, and to record the speed of travelling and the exact time and number of the event-markers. The whole process was saved on disk so that individual elements of the trials and events could later be screened for analysis.

The Gedesby ship is 12.5 m long and has a 5.0 m beam. Its draught is just over a metre, depending on the load. In all we sailed with a displacement of about 13.2 tons during the sailing-trials. The weight of the hull, rigging, equipment and crew came to a total of 6,565 kg. The weight of the stone ballast was initially 5,820 kg, but we added a further 835 kg to give better directional stability in the strong winds towards the end of the period of the trials. For the purposes of comparison, the 20 oxen weighing about 200 kg each which the Gedesby ship would probably have had room for would only have weighed 4 tons, so a certain amount of ballast would no doubt have been added as well. All this was propelled by a square sail, 60 m2, of linen cloth. The stability, as could be expected of a loaded cargo vessel, was reasonably good. The first reef, which reduced the sail area by at least 10 square metres, did not have to be made until the wind-force reached 11 m/s ('strong wind'), and the next at 'moderate gale' force, 14-15 m/s.

With the wind coming in across the stern the highest speed achieved under the trials was measured at 8.1 knots (almost 16 km per hour) in a wind-force of 13 m/s and with one reef in the sail. Obviously, the speed is lower in more normal wind-conditions, e.g. around 4-6 m/s, which would give about 4 knots. In general the fact is that when sailing with the wind one sails faster with a stronger wind. This holds true until a certain point, when because of the water-resistance the speed cannot be increased further, but we did not reach that point during these trials. According to common theories, in the case of the Gedesby ship this point should lie around 8.5 knots.

Click here to se graph illustratring off-wind sailing abilities of ship replicas

Close-up of the Agnete during sailing-trials, with the sail being reefed.

Photo Werner Karrasch.

Against the wind the picture is not so simple. To start with, the speed is lower, and secondly one has to proceed by tacking in zigzags. At the same time, the wind-pressure causes a deviation which can be significant in the case of a ship with such a relatively slight draught as the Gedesby ship has. In order to judge how well a ship has tacked in given wind- and sea- conditions, one calculates the VMG (Velocity Made Good), which is a numerical value for a ship's effective progress directly into the wind. The VMG is valid; i.e. one can completely ignore deviation etc. and make comparisons with the VMG of other ships in the same wind- and sea- conditions.

The Agnete's VMG ratings lie between 0.3 knots at the worst and 1.5 knots at the best. When it comes to tacking, it is the case that the upper limit of how fast one can sail against the wind is reached relatively fast, because the wind-resistance and deviation increase sharply with an increasing wind-force, and at the same time the tractive force of the sail is considerably reduced when it is reefed. In the case of the Agnete it also looks as if the VMG is at its best, between 1.2 and 1.5 knots, with a wind-force of 9-10 m/s, but begins to fall to under 1 knot in wind over 11 m/s, i.e. just at the point when the sail should be reefed. If two reefs are made, the rate falls even further as a result. We have two results in 14 m/s with a double-reefed sail. The worst rate, of 0.3 knots VMG was obtained the first time we had double-reefed, and is probably the result of lack of concentration in an unfamiliar situation. The following day, when there was also a moderate gale-force wind, we were better prepared and the results were more than twice as good: 0.7 knots VMG.

Click here to se graph illustrating tacking abilities of replicas

These measurements show that across the board the VMG results for the medieval replica Agnete are lower than those for the Viking trading-ship replica, the Saga Siglar, with 26 tons displacement (Skuldelev 1) and Roar Ege, with 7 tons (Skuldelev 3). This is because the Agnete's sail cannot be aligned as closely fore-and-aft during tacking as is possible on the Viking ships, because of the Agnete's low breadth/length ratio - 1:2.5, compared with 1:3.2 for Saga Siglar and 1:4 for Roar - which results in the shrouds getting in the way to a greater degree. This means that the ship cannot sail closer to the wind than 65-70 degrees, while the other two ships can reach about 60 degrees.

Rowing the Gedesby replica turned out, not unexpectedly, to be somewhat laborious. But with the 6 very long oars projecting out between the shrouds, as can be seen on several old illustrations, it proved possible in fact to manoeuvre the Agnete fairly successfully around the harbour in a light wind. In calm weather it must be possible to reach a speed of a little over 2 knots over a short course, but against the wind little headway can be made by rowing. The oars are no doubt only for positioning and harbour manoeuvres in confined and very sheltered waters.

It was a positive surprise to discover that the Agnete could 'turn on a sixpence' in virtually all conditions. This was true of changing tack against the wind, which is often a problem for square-sail ships, and of veering. During tacking most ships avoid veering as far as possible, because it results in loss of 'altitude' and speed. But for the Agnete the turning area was not larger than about 30 m, which meant that she could actually come about just as fast as in tacking, and moreover was still going at speed, whereas during tacking a ship lies still immediately after turning. Both types of turning took only about 50 seconds at best; this was the fastest we have to date recorded for a square-sail ship. The reason for this must be the ship's waterlines, which are very rounded compared with the longer, slighter curves seen in many of the Viking ships. The fact that the Agnete has a stern rudder in contrast to the side-rudder of the Viking ships has no effect, according to our experience, on the turning properties. The Gedesby ship replica would probably turn just as fast with a good side-rudder. In spite of the rounded water-lines, which might perhaps have resulted in imbalance or a 'hard rudder' under sail, the balance was very good. The Agnete was neither weather-helm nor lee-helm, but could be steered by a single finger on the tiller.

Sea-worthiness is no doubt the most prominent of the Gedesby ship's virtues. If a ship can make 0.7 knots against the wind and current with a double-reefed sail in a moderate gale, there would have to be exceptionally bad weather conditions, possibly combined with physical damage before it could be wrecked or swamped. During our trials the waves were up to 1.5 m high, and under those conditions it was a pleasure to see how the Agnete's very flaring sides simply threw the sea back and down during tacking, so that not even spray fell on board. At the same time one could clearly see that the ship's flexible construction distributed the pressure from the waves over larger areas of the ship than those directly hit by the first impact, and this significantly reduced the strain on the forepart.

Even the rigging seems well-reconstructed and sturdily-made. It was well and truly put to the test in the gale-force winds. As already mentioned, the full-sail carrying capacity could be rated high, since the sail did not require reefing until 11 m/s. The Viking cargo-ship Roar Ege, which is only half as large, in relation to its displacement, has to be reefed at 8-9 m/s, while the Saga Siglar, double the size of the Agnete, can wait until 14-15 m/s. One is often compelled to reef slightly earlier, however, to ease the labouring of the ship in high seas.

In conclusion, we already know that in comparison with the two Viking trading-ships the Agnete can be said to be entirely able to compete in relation to speed in the water, but in tacking there are differences. The Agnete's course made good in relation to the wind, as it is recorded on screen taking account of deviation, lies within the area 71-82 degrees from the direction of the wind. The same calculation has given between 64 and 72 degrees for Roar and between 65 and 68 degrees for Saga Siglar, in all cases with the worst figures being obtained with reefed or double-reefed sail. All three square-sail ships achieve their best VMG at 9-10 m/s, but the Agnete's is not quite as good as those of the others. Finally, it looks as if the Agnete is by far the most sluggish in slack weather. Perhaps the ship should have had slightly more sail, but it should be stressed that we have too few test results in light wind to be able to judge this. In order to obtain more statistically reliable results the Agnete should be subjected to at least a further week of controlled sailing-trials - preferably in more normal Danish summer weather with relatively light wind-conditions and perhaps with e.g. the Roar Ege participating as a sparring partner.

Publication

The sailing-trials, along with the reconstruction-work and the quantities of experience gained during the building of the Agnete will be described in a report to be published in Danish by the National Museum's Centre for Maritime Archaeology in 1996. This report will primarily be an account of the reconstruction and of the aims of the experimental-archaeology project, its methods and results. An English publication will be produced in the series Ships & Boats of the North, in 1998. This will to a greater extent be a presentation of the Gedesby find itself, of its context and of parallels to it, but it will also include the results of the sailing experiments with the replica.

In parallel with this a corresponding process has been initiated to prepare an account of the Skuldelev find and to publish it along with the experiments which have been carried out with the replicas of those ships. In this way the Centre for Maritime Archaeology will be able, in the coming years, to make important contributions both to the discussion of methodology with regard to the viability of experiments in archaeology and to the debate concerning the historical development of ships and seafaring.

Jan Bill & Max Vinner