1The Puck 2 wreck was discovered in 1979 during archaeological excavations conducted in the medieval harbour of Puck. The site was discovered in 1977 by a group of amateur divers. The first excavations were supervised by Wiesław Stępień from the Puck Regional Museum. The site was identified as the remains of an early medieval harbour dated from the 9th-13th century AD, extending over an area of nearly 12 ha. It ­contained harbour constructions, numerous artefacts and several wrecks. In 1990-2006 archaeological works were conducted by the National (then Central) Maritime Museum in Gdańsk and the University of Nicolaus Copernicus in Toruń. The archaeological teams were able to record the surface features of 6 ha, open several test trenches and excavate the discovered wrecks. Until now five wrecks (a logboat and four lapstrake vessels) have been found in Puck medieval harbour. The first wreck (P1) is lying on the eastern side of the site. It was documented in 1983 and is still in situ, having been secured with geotextile. The 1983 recording revealed that it is a wreck of a lapstrake vessel built in the clinker tradition, assembled with iron rivets and caulked with animal hair. According to Stępień, the aft part of the wreck has been uncovered, revealing the stem, framing system and traces of repairs caulked with textiles, which was ¹⁴C dated to 1245 AD (Stępień 1998, p. 45). Dendrochronology performed in later years yielded different results: the boat was probably built after 1380 (Ważny 2013, p. 47). The next wreck (P3) lay in the western part of the documented area. It was partially uncovered and documented in 1983 (Stepięń 1998, p. 47). In 1990, when the National Maritime Museum in Gdańsk resumed exploration of the site, the wreck was fully excavated, documented and salvaged (Pomian, Litwin 2009, p. 423). In 2018, after conservation it was reassembled in the museum’s Shipwreck Conservation Centre in Tczew. The P3 wreck ­consists of an 8.3 m-long keel, fore stem, floor timbers, a few thwarts and planking assembled with wooden pegs and caulked with moss. The width of the longest beam (2.8 m) and the overall proportions of the vessel indicate that it was a small cargo carrier. The wreck has been dendrochrono­logically dated to after 1155 AD (Ossowski, Krąpiec 2003, p. 271-292). During the preparations for salvage of P3, another wreck (P5) was found. It was lying just 0.5 m from the bow of P3, buried deeper into the seabed. It was partially uncovered, revealing the remains of a large lapstrake vessel assembled with wooden pegs and caulked with moss. The planking was well preserved and yielded very precise dendrochronological dates, fixing the felling of the trees to 1248 AD (Litwin 1995, p. 142). It was decided to leave the wreck in situ and secure it by burying it under the sediment. Interestingly, the P3 and P5 wrecks were filled and surrounded by many loose wood shavings, which may suggest that they were ready for repair works (Pomian, Litwin 2009, p. 423).

1. The P2 wreck: exploration and salvage

2The P2 wreck lay in the central part of the archaeological site. It was situated on what appears to be a submerged early medieval coastline reinforced with fascine (Pomian 2002, p. 131). Traces of burning on the keelson, remains of flax inside the hull and the fact that the wreck was attached to the seabed with wooden posts may indicate that the vessel had been partially submerged and used for soaking flax (Pomian, Litwin 2009, p. 423). The wreck was first found, explored and published by Wiesław Stępień, who was unable to complete his research (Stępień 1987). He managed to identify several construction features of the vessel and to take samples for ¹⁴C dating (555 AD, and later 810 AD), which turned out to be contro­versial (Pomian, Litwin 2009, p. 422). However, Stępień was able to provide the first reconstruction of the vessel (Stępień 1998).

3Exploration of the P2 wreck was resumed in 2003 and 2005 by a National Maritime Museum team led by Iwona Pomian, head of the Underwater Archaeology Department. Prior to these latest explorations, in 1999 samples for dendrochrono­logical analysis were taken from the planking. The results placed the vessel in the first half of the 10th century (Ważny 2001). The wreck was excavated and documented underwater in 2005 by preparing drawings as well as photo and video recording. It was decided to raise the wreck by dismantling it underwater into individual timbers, which were then transported to the conservation workshop, documented in 1:1 scale and treated with PEG (Pomian, Litwin 2009, p. 424). Several planks and structural timbers were in very bad condition and they disintegrated into small pieces, hence the total number of salvaged construction elements rose to 195.

4The wreck as found can be described as the remains of the forward half of a long, narrow vessel. It consists of an almost 12 m-long keel with bow extension and a keelson with a mast step. Stems are missing, and the keel is torn off at approximately two thirds of its original length. Seven strakes of the starboard and four of portside planking are preserved, assembled with wooden pegs and caulked with animal hair. The hull has been stiffened with a framing system consisting of partially preserved floor timbers, thwarts with stanchions and knees. The hull was longitudinally strengthened by stringers preserved in six fragments.

2. The post-conservation 3D documentation

5In 2012 the Underwater Archaeology Department of the National Maritime Museum in Gdańsk was equipped with a FARO ScanArm for the purpose of digital recording shipwreck timbers. The first documentation project carried out in the museum was the recording of timbers from the Copper Wreck in 2013 as part of an attempt to create a digital reconstruction of the hull. All 189 timbers from the Copper Wreck were documented using laser scanning producing point clouds, which were then processed into surface models The acquired models were processed, simplified and assembled in order to recreate the shape of the hull, and then used for hydrostatic calculations performed by experts from Gdańsk University of Technology (Żrodowski 2014).

6In 2016 the current authors began the documentation of timbers from the P2 wreck in hope of achieving similar results. It was decided to supplement the point cloud models achieved by laser scanning with three-dimensional wireframe drawings made in Rhinoceros 3D software, according to the methodology developed during the Hjortspring boat (Hocker 2003) and the Newport ship projects (Jones 2011). On such drawings each timber is seen as a set of faces documented one at a time, and the data on all features of the given element has to be collected manually by the documentation device operator. Both methods supplement each other: laser scanning allows the recording of every feature of the object’s surface, though the post-processing stage is very time consuming; the contact method is more rapid, but also less accurate, and the outcome is more human dependent (Jones 2011, p. 115). During recording it became clear that the documentation of long and thin planks is very troublesome with laser scanning as the object tends to bend after being turned around. Therefore laser scanning is more suitable for more robust objects like framing, whereas the contact method works well with planking. Overall, 161 cons­truction elements of the P2 wreck were recorded: all of them with the contact method, and 102 with the optical method. Among these were the keel with bow extension and the keelson, 37 fragments of floor timbers, eight fragments of stringers, ten knees, 12 thwarts, 88 fragments of planking including over 3 m-long planks, as well as small chunks and eight other elements, such as stanchions, rowlocks and pieces of shield rack. The few dozens of smallest pieces of wood of unidentified function were excluded from digital recording and only documented photographically.

3. Model assembly and description of the wreck

7After recording all structural elements, the processes of assembling the model could be started. The chosen environment was the Rhinoceros 3D 5.0 software. The software provides useful tools for importing different types of 3D objects, their post-processing, scaling, transformation, rotation, etc. The first stage was to digitise the in situ drawings, which served as the basis for the location of the keel assembly, keelson and frames.

8The preserved keel assembly of the P2 wreck consists of an 11.87 m-long T-shaped keel and the 1.3 m-long bow keel extension. The keel has been slightly twisted and bent in its aft part during a millennium underwater and had to be digitally straightened. The framing has been assembled according to the in situ documentation and field inventories. There are fragments of 15 floor timbers preserved in 35 fragments. The most prominent element of the P2 wreck is undoubtedly the keelson with a mast step, located between frames Nos 9 and 12: its length is 2.79 m, and the mast step is located 8.1 m from the fore end of the bow keel extension. It was assumed that frame No. 11, being the widest and located just behind the mast step was a central frame of the vessel and therefore a total keel length (with fore and aft extensions) was estimated at 17.8 m, which results in a total number of 23 frames. The missing elements of the framing system have been added to the model by copying and mirroring the preserved parts. Both stems of the vessel are missing and therefore their real shape remains unknown. Proposed reconstructions of stem shape have been prepared by J. Litwin and W. Stępień (Pomian, Litwin 2009) and they served as the basis for the two versions of the assembled model of the P2 wreck. The keel and the framing system components have been deformed by a millennium of resting below the seabed sediments. An attempt was made to reconstruct the original curvature of the floor timbers in order to recreate the hull as faithfully as possible. The height of three preserved thwart stanchions served as the determinant for the frame height: the length of preserved thwarts determined the width of the framing. There are 12 fragments of thwarts and ten knees preserved. With the help of the field documentation, it was possible to arrange them with the corresponding frames and to prepare the reconstructions of missing parts. Fortunately, one of few preserved knees was found still nailed to the thwart on frame No.10, closest to the mast step, which provided the cross section of the midships part of the vessel (fig. 1). The final elements of framing are represented by poorly preserved pieces of longitudinal stringer, fitted above the thwarts and on the knees.

9After the placement and assembling of the framing system came the planking. All planks were made of oak. The starboard planking is in much better condition than the port, where the individual planks had to be assembled from a large number of chunks. The planks are on average 1.5 cm thick and 22-27 cm broad (while Strake 3 is narrower at only 15 cm). The only plank preserved completely (i.e. with both scarfs) is 3.26 m long, while the longest assembled strake (starboard garboard strake) is 12.81 m long and consists of four planks, joined together with simple transverse scarfs, approximately 7-8 cm long and connected with pegs and small rivets. The planking was joined with treenails approximately 1.5 cm in diameter placed every 9-10 cm, with an overlap of approximately 4 cm and caulked with animal hair. The planks are mostly oval in cross section with the exception of Strake 5, which has a deep groove through the middle part that may have indicate the waterline level. Several planks bear incisions left by frames. The third portside garboard plank bears traces of repair: a small, 43 cm-long fragment was treenailed to the plank with an additional row of treenails. Only a small piece of a sheer strake is preserved. There are no traces of the upper railing, although W. Stępień mentions such a feature as visible underwater in the 1980s (Stępień 1987, p. 152). There are two pieces of a shield rack and two rowlocks preserved, which indicate the military function of the vessel and its means of propulsion.

4. Hydrostatic characteristics of the reconstructed hull

10With the preserved planking assembled (fig. 2), the overall hull shape could be reconstructed and the hydrostatic characteristics of the hull assessed, according to the methodology developed for the Drogheda boat and the Newport ship, with the use of the Orca 3D plug-in (Tanner 2013a, 2013b, 2017). The edges of preserved planks were elongated alongside the frames to create the fair lines of the vessel (fig. 3). The unknown details of joining the planking with stems were not recons­tructed in the model. For further simplification of this lapstrake planking, a plain hull using the Orca 3D plate tool was created. In order to estimate the weight of the vessel and therefore its light displacement, weight properties of each construction element (both preserved and reconstructed) were defined within the Orca 3D weight and cost properties tool. Most elements (keel assembly, majority of framing, planking) are of oak, while a few frames, knees and stringers are birch, alder and pine (see table 1). The wood species were analysed by Irena Jagielska from the conservation laboratory of the National Maritime Museum in Gdańsk.

Table 1: Weight properties of structural elements of the P2 wreck

Element	Material	Weight
Keel assembly (with keelson)	oak (690 kg/m3)	207 kg
Floor timbers, knees, thwarts, rowlocks	oak (690 kg/m3)	162 kg
Five floor timbers, stringers	alder (510 kg/m3)	216 kg
One floor timber	birch (610 kg/m3)	5 kg
Planking	oak (690 kg/m3)	897 kg
		1487 kg Total

11The lightweight of the vessel and the overall shape of its hull allowed for calculation of its draught and loading capabilities, with the use of the minimum freeboard (F) condition after the medieval Icelandic Law (Tanner 2017, p. 173), which in case of the P2 vessel gives 32 cm (F=2D/5). The results of hydrostatic calculations performed by the Orca 3D software are given in Table 2. Three conditions were assumed: a lightweight condition, with the vessel empty of crew and provisions; a loaded condition with 31 crewmembers (each weighing 70 kg, 30 sitting on rowing benches and one standing) and 1500 kg of ballast, provisions and equipment; and a fully loaded condition with the 32 cm freeboard (fig. 4).

Table 2: Hydrostatic characteristics of the P2 vessel

	Lightweight	Loaded	Fully loaded
Length overall	21.35 m	21.35 m	21.35 m
Beam	2.16 m	2.16 m	2.16 m
Midships depth	80 cm	80 cm	80 cm
Waterline length / wet area	18.6 m/ 26.4 m2	19.1 m/38.5 m2	19.3 m /42.3 m2
Displacement	1487.4 kg	5157.4 kg	7094 kg
Draught	21.5 cm	40.1 cm	48 cm
Freeboard	58.5 cm	39.9 cm	32 cm
Crew number and weight	0	31 men = 2170kg	31 men = 2 170 kg
Ballast, equipment, provisions, etc.	0	1500 kg (including 600 kg of water and provisions for 5 days)	3437 kg (maximum loading capability)
Max heel angle	35	20	15
LCB	10.6 m	10.5 m	10.6 m
VCB	14.5 cm	26.6 cm	31.4 cm
VCG	49 cm	57.6 cm	51 cm

5. Conclusion

12As the hull reconstruction and the results show, P2 was an extremely long and light vessel, with a beam to length ratio of 1:10 and roughly 1500 kg of lightweight. Powered by a maximum crew of 30 oarsmen, the boat could also be a very fast one. It was undoubtedly used for military purposes by a local chieftain ruling over the Gulf of Gdańsk region, maybe during the struggle with the Polish Piast dynasty for control of these areas in the late 10th century (Śliwiński 2013). However, the propulsion characteristics of rowers and hull resistance still remain to be tested. Also, the presence of the keelson with a mast step raises many questions about the sail propulsion characteristics and the seaworthiness of such a slender vessel, which bears traces of both Scandinavian and southern Baltic (Slavic) shipbuilding traditions. This documentation and reconstruction programme was the first to be carried out on an early medieval wreck from the National Maritime Museum in Gdańsk collections. Recording and reconstructing other lapstrake vessels from this time period and region will provide us with the basis for comparative studies and undoubtedly enrich our knowledge about shipbuilding in the Gulf of Gdańsk area.