Discovery and Excavation

In September 1992 the Dover A20 project was drawing to a close. For well over a year, archaeologists from the Canterbury Archaeological Trust had been monitoring roadworks linking the port of Dover with the new channel tunnel and motorway links with London. Ironically, the first land link with the continent for 8,000 years was to indirectly bring to light the earliest example of sea transport yet found.

The discoveries of the A20 project were abundant and important, though their full potential has not yet been realised. A deep shaft excavated for a pumping station, some 200m inland, was proving to be the climax of this long-running project. A section of the medieval town wall, its stones rounded by the battering of the sea, overlay the massive squared timbers of a Roman harbour mole. The recording and recovery of these timbers was seen as our last important discovery, and we prepared to wind down the project.

The shaft was excavated still deeper, however, and routine monitoring continued. The Field Director, Keith Parfitt, noticed some fragments of timber disturbed by the contractor's machine bucket lying in sediments underlying the Roman harbour timbers. After closer examination, he realised that he had found part of a prehistoric boat. It was midday, Monday 28th September; Keith managed to negotiate a temporary halt to the contractor's work schedule. Telephones began to ring.

By 11.30 the next morning, a team of specialists was assembled on site, including experts from English Heritage, who had immediately made funds available to recover this unexpected find. We were initially given 3 days to excavate the boat, later extended to six days. We worked from first light until late into the night, with water gushing into the cramped cofferdam, only 6m long and 5m broad. After recording, the boat was cut into sections and lifted by crane on Saturday 3rd October, to be stored in a temporary water tank provided by Dover Harbour Board. The last fragment, running into the section, was recovered in about 20 minutes on Sunday 4th.

We had recovered the mid-section of a prehistoric boat, probably of Bronze Age date judging by the technological parallels with the Ferriby river boats, recovered half a century before. It was a discovery of European, if not world-wide importance. The decision was made to sink another shaft, slightly larger at 7m long and 5m broad, immediately adjacent to the original discovery. After a week to recover, on Monday 12th October, we were back down another hole, now with a slightly better idea of what to expect! Miraculously, the second shaft straddled one end of the boat, crucial to our understanding of the form and function of the original vessel. This time we had nine days grace. We cut the boat into sections and lifted on Monday 19th October, leaving Tuesday to complete palaeoenvironmental sampling.

The other end of the boat lay near to standing buildings, and given the potential costs and the possibility it had been destroyed by the later Roman timber construction, it was decided not to attempt to recover this end.

back to top

Analysis

We have a 9.5m length of a perfectly preserved Bronze Age boat, about 2.3m wide and dated by 14C to about 1,600 BC. The section recovered consists of four main planks of oak. Two flat planks form the bottom, each carved out of half logs, leaving upstanding cleats and rails allowing its jointing with other boat timbers. These bottom planks were joined together along a central butt joint, with transverse timbers and wedges hammered through the cleats and central rails. Curved side planks were stitched to the bottom of the boat with twisted yew withies. These side planks also possess side cleats carved out of the solid wood. The timbers forming the end of the boat splay into a Y-shape, intricately carved from the main planks. This originally would have held a carved wooden board, reminiscent of a modern "punt". There was clear evidence for the presence and some dimensions of this missing end board; there was also evidence for at least two other main structural timbers. On the top of the curved side planks (or "iles") was another row of stitches, cut through in antiquity. There were clearly two further side planks, and the boat had been deliberately dismantled (albeit partially) when it was abandoned. She had been made waterproof by pressing in a "stopping" (possibly animal fat) into the stitch-holes and along the seams, where the stopping was overlain by pads of moss wadding, compressed and held in place by long thin laths of oak under the yew stitches. The boat had clearly been used extensively. Tool marks on its bottom (outboard) surface were differentially worn away, suggesting it had been beached regularly on a sand or gravel shore. The main timbers had split and were repaired by stitching wooden laths over the damage.

The boat was made from at least 3 straight-grained oak trees, well over 300 years old when felled, grown in dense woodland with few side branches. The original logs were probably 12m long and 1.1m in diameter, weighing about 12 tonnes each. Analysis of the tree rings shows that the two bottom planks (each carved out of half a log) were from two different trees, probably "wolf trees", out-competing their neighbours in early growth to grow fast above the forest canopy. Such trees are extinct in Western Europe today. The withies, unusually made of yew, a tough durable material rather difficult to work with, were about 1.4-1.8m long and between 6-16 years old. They probably derived from a coppiced stand of fastigated yew, now extinct in England.

The boat timbers displayed well-preserved toolmarks, though there were no clear examples of recurring tool signatures. There was particularly good survival on the bottom of the boat, suggesting the boat was less than 10 years old when abandoned, though they may have been protected by matting or dunnage of some kind. A minimum of five tool types could be deduced from the toolmarks, including adzes, axes, chisels and gouges.

An experimental reconstruction of a 3m section of the boat in August 1996 allowed us to better understand the sequence of construction of the boat. Study of the rays and rings of the original timbers revealed in the cross-sections through the boat showed that they had been substantially compressed during its long burial, by as much as 50% in some instances. The reconstruction took this compression into account, helping us to visualise the appearance of the original vessel. Using facsimiles of local contemporary tools, we found that the toolmarks created by modern woodworkers matched almost exactly those on the original boat. Indeed, the inexperienced woodworkers easily achieved a finer finish on the reconstruction than on the original. Whilst there can be no questioning the expertise of the Bronze Age woodworkers, this does seem to imply that they were not seeking a fine finish to the timbers. Perhaps they were more interested in function than display?

The original boat

Study of the hull has led us to believe that we have recovered the majority of the boat. It probably had a vertical transom end, making the Y-shaped end the bow of the vessel. A minimum reconstruction suggests the boat was about 11.7m long, with a beam of 2.26m and a height from her bottom of 0.8m. She probably weighed around 2.3 tonnes.

Various permutations of crew and cargo have been studied using this proposed hull form. With a crew of some 16 paddlers (together with four bailers!) The boat could easily achieve 5 knots and made way in winds of Force 4 with waves up to a metre high. Even adding 3 tonnes of cargo, reducing freeboard to about 0.45m, the boat could still have travelled in Force 3 winds with waves 0.6m high. Five knots appear to have been the maximum speed for maintained paddling; even a small crew of four paddlers (and one bailer) could have propelled her at up to 4 knots. Above five knots, however, the drag created by the hull form rises steeply, so that even a speed of 5.5 knots would have been impossible to maintain.

The boat appears to have been suitable for carrying a significant cargo at a reasonable speed, in fairly rough weather, possible over long distances. On a good day, with favourable tides, she could perhaps have covered 30 nautical miles with continuous paddling.

Analysis of the soil samples taken during excavation suggests the boat was abandoned in a completely freshwater environment, probably a shallow braided river channel some distance from the sea. The valley floor was probably cleared, perhaps with some settlement nearby; the deposits immediately surrounding the boat contained charred cereal grain and other domestic refuse. The absence of oak pollen may suggest the boat was not built in the Dour valley.

The ile planks

The ile planks were to prove even more difficult. Not only did the iles have to match with the adjacent ile timbers, the small section of base plank, attached by the stitches on each one, had to align with the relative cut of the base plank already on the cradle. In addition, the ile had to sit at the correct angle, or as close as possible, to keep the strake line consistent. It was decided that the complete side should be offered up at the same time so that each piece could be directly related in situ.

Supporting the southern end

Time soon became increasingly short; the re-assembly process was but one part of the programme to establish the gallery, whose opening ceremony had already been organised. The complex southern end section had still not been put into place. The final mechanism for supporting the piece was a subject of much discussion. Ultimately a one-piece welded support bolted to the existing spine plates seemed to be the most suitable answer. Struts extending back to the first available rib would help to offset the weight distribution. A wooden mock-up was prepared to form the template for this complex support.

As the newly-made pieces were returned, each rib was bolted into position with relatively few problems and trimmers added. All of the temporary wooden formers and legs were gradually removed to leave the boat supported solely on the cradle. The end of the spine had to be cut off as the design had changed. A cap, made of the same material, was fitted to conceal the alteration and the new end section was bolted into place. The re-assembly of the main boat pieces was complete.

back to top

Lowering the boat

It now remained to lower the boat to its display level. The height of cradle suitable for the re-assembly process was clearly unsuitable for presenting it to the public; many people, particularly children, would be unable to see the inboard surfaces of the boat. The entire re-assembled boat needed to be lowered by 0.5m This was an intensively nerve-wracking procedure; the re-assembly team had spent ten months positioning the boat pieces and customising the cradle support. Any mistake, even if the timbers themselves escaped damage, would require the procedure to be gone through again.

The method adopted was essentially quite simple. Four steel girders were manufactured with a bracket welded on their upper surface that allowed them to be bolted to the cradle spine adjacent to the main support columns. These were supported on stacks of timber. Simple scissor jacks were bolted to either end of each girder, which were then set atop further stacks of smaller timber blocks. When all was in place, the cradle spine was unbolted from its supportive columns.

It was essential that the boat was lowered very slowly, being kept absolutely level at all times. A trial run was carried out prior to the re-assembly to evaluate the method, the cradle weighed down with steel plates and scaffold boards to simulate the weight of the boat. This proved highly successful, and many useful lessons were learnt. It was a very different exercise lowering the boat itself. A team of eight volunteers was assembled, each charged with manning one of the jacks. The cradle was plastered with spirit levels along the spine and on the transverse ribs so that any change in angle of the boat cradle could be monitored as it was lowered. (During the trial run, teacups were filled with water and placed atop paper napkins in saucers at various points along the cradle. Any jolting or tipping would result in water spilling and dampening the napkins. They stayed dry).

On a chant from the project manager (Peter Clark), each of the volunteers raised the jacks one turn (clockwise). The boat was now freed from its supportive columns, resting only on the eight jacks atop their stacks of timber. The re-assembly team now replaced the supportive columns with new ones, only 0.5m high. They then unscrewed and removed the topmost timber block from the main stacks. Then, again following a chant, the jacks were lowered, one turn at a time (anticlockwise) until the steel girders rested on top of the next main timber block. The jacks were slackened off, a timber block removed from each of the small timber stacks, and the strain taken up once more. This process was repeated until the main spine rested on the new, shorter supportive columns.

Finally, the boat reached its final position. The cradle spine was bolted to the new supportive columns, and the steel girders, wooden blocks and jacks removed. It is difficult to describe the feeling of relief felt by the re-assembly team as we stood back from the display case; in spite of all our fears, everything had gone well. The entire process only took 70 minutes to carry out.

Completing the case

We could then address the final stages of re-assembly; fitting small pieces of timber, such as sections of laths, wedges and transverse timbers, and reconstituting stitches that had become separated from the boat pieces. We tried very hard to bring the boat back to the condition in which it was found. After a final surface treatment of the re-assembled timbers, the job was complete.

The contractors glazed the case with a series of toughened glass panels, 19mm thick and sealed with clear mastic. This allowed a check of the environmental control system to be undertaken, after which the case was sealed behind a protective timber hoarding as the rest of the gallery was fitted out.

The Dover boat has now reached a safe haven after her long journey down the millennia. She must surely enter her fifth millennium as one of Europe's greatest prehistoric discoveries.

Acknowledgements

As might be imagined, the boat owes a debt to a great number of individuals and organisations. Particular note should be made here to the staff of English Heritage, Dover Museum, the Canterbury Archaeological Trust, P & O Ferries, Dover Harbour Board, the Heritage Lottery Fund and the members of the Dover Bronze Age Boat Trust.

Peter Clark, Project Director
Canterbury Archaeological Trust
92a Broad Street Canterbury Kent CT1 2LU UK
e mail: pete.clark@canterburytrust.co.uk

Success for the Dover Boat at the
British Archaeological Awards 2000

The success and importance of the Dover Bronze Age Boat Project was recognised at a ceremony in the Great Hall of Edinburgh Castle in November 2000, where the project was given not one, but two major awards!

The British Archaeological Awards have many different categories. The finalists were welcomed by Lord Montagu of Beaulieu, and the awards presented by HRH Prince Hassan of Jordan. The boat project won the ICI award for the best archaeological project offering a major contribution to knowledge, along with a beautiful bronze plaque, which is to be displayed alongside the boat in Dover Museum. The project was also highly commended for the Virgin Holidays Award for the best presentation of an archaeological project to the public.