Cap del Vol, late 1st century BC–1st century AD
Country: Spain
Place: Cap del Vol, municipality of El Port de la Selva (Girona), at a depth of 24 m
Coordinates (approximate): 42.34, 3.21
Type: Roman coastal trading vessel (wine carrier)
Identified: No
Dated: Between the end of the 1st century BC and the 1st century AD, based on the Pascual 1 amphorae and local imitations of Arretine ware (Nieto and Foerster 1980; Foerster 1980; Nieto 1982).
Beam: Unknown.
Keel Length: Unknown; the mast step is practically as long as the keel from stem to sternpost.
Length Overall: Unknown.
Number of Masts: One mast step is preserved.
History of the wreck
The discovery of the Cap del Vol wreck was reported to the authorities in 1978, although it is known that its cargo of Pascual 1 amphorae had been looted since at least 1967. Underwater excavation campaigns were carried out between 1978 and 1980 under the direction of Nieto and Foerster. The preliminary test trenches located remains of the hull, so efforts were concentrated in the southern part of the site (in a 2 × 3 m square). Some Pascual 1 amphorae bore stamps applied before firing, with the labels S, H, L, QFS, M. PORC and IVLI THEOPHIL, which showed that they contained wine produced in the ager of Badalona. The presence of imitations of Arretine ware, produced locally in the Augustan period and during the 1st and 2nd centuries and intended mainly for trade with southern Gaul, allowed the wreck to be dated between the end of the 1st century BC and the 1st century AD (Nieto and Foerster 1980; Foerster 1980; Nieto 1982).
In 2011 the CASC (Centre d’Arqueologia Subaquàtica de Catalunya), under the direction of Vivar, De Juan and Geli, carried out a first assessment of the state of preservation and archaeological potential of the Cap del Vol wreck. Among other matters of interest for the study of maritime trade, it confirmed that the general state of the ship was excellent and that the frames were fastened to the hull by a technique of looped internal lashings secured by wooden treenails, as Marlier had proposed in her doctoral thesis of 2003. During the 2012 excavation campaign, directed by the same team, a series of particularities in the ship’s construction were observed that allow the level of detail and systematisation to be raised from that of an architectural signature to that of a shipyard practice. This means that traces and ways of carrying out a well-known building technique by a particular group of craftsmen can be detected: there are general similarities with contemporary wrecks, above all with those of the same architectural signature, but enormous differences in detail.
After listing all these construction details, which make up a shipyard practice clearly different from that of the other wrecks known today, what arguments allow the geographical area of the shipyard to be specified? With the caution required by any argument based on the accumulation of archaeological evidence, the author considers that, of all the possible locations for the shipyard where the Cap del Vol was built, Tarraconensis is the most likely. The ship seems to be a specialised machine able to sail in various nautical spaces, including rivers and lagoons, judging by its flat-bottomed hull lines and its backbone composed of a keel and garboard strakes of particular shapes. At sea, however, its sphere of navigation seems clearly to have been coastal, carrying on the outward voyage Tarraconensian wine loaded near the packing centre itself, with Narbo as its final destination; its return cargo has yet to be determined. As part of a Laietanian commercial enterprise, frequently repeated, the Cap del Vol seems to belong to an export system in which costs were reduced to a minimum (no port infrastructure, storage or transhipment of goods were needed). Logic suggests that the enterprise was carried out by agents of the same ethnic group, including the shipbuilder, owner and sailors. Finally, the Iberian coin, possibly of Bolskan, found at the foot of the mast must be considered fragile evidence: although the circulation of such coins was restricted to the Gulf of Valencia and the mouth of the Ebro, it is very worn and was already old when used. Nevertheless, it seems too much of a coincidence that one of the only two wrecks with Iberian apotropaic coins is the Cap del Vol.
In the author’s view, the Cap del Vol is a ship whose construction could show features of cultural hybridisation between the indigenous tradition of the Iberian Peninsula and the shipbuilding of Narbonensis: techniques and knowledge that, without being native to the Iberian Peninsula, were applied by a group of fabri navales of the eastern Iberian coast, with nuances of shipyard practice without known archaeological parallel. They are in fact a hybrid reinterpretation of architectural types within a little-known and poorly understood Iberian line of shipbuilding, whose earliest exponents would be the Mazarrón boats (late 7th and mid-6th century BC) and the Binissafúller wreck (4th century BC).
Description of the site
The Cap del Vol wreck lies near El Port de la Selva, in the province of Girona, at a depth of 24 m. The 1978–1980 excavations concentrated on the southern part of the site (Figure 1); in 2012 a further sector of the hull was excavated and recorded (Figures 2 and 3).
Figure 1. Plan of the excavated part of the Cap del Vol wreck: 1, sternpost; 2, forefoot; 3, keel; 4, keelson; 5, ceiling plank; 6, hull planks; 7, floor timber no. 1; 8, floor timber no. 2; 9, floor timber no. 3 (after Nieto and Foerster 1980: 166).
Figure 2. Plan of the sector of the hull of the Cap del Vol excavated in the 2012 campaign (drawing: Geli, CASC).
Figure 3. View of the Cap del Vol wreck during the 2012 campaign (photo: Vivar, CASC).
Cargo
The cargo consisted of Pascual 1 wine amphorae, some stamped before firing (S, H, L, QFS, M. PORC, IVLI THEOPHIL), containing wine from the ager of Badalona. Local imitations of Arretine ware were also found. The cargo had been looted since at least 1967.
Ballast
Not reported.
Anchors
Not reported.
Guns
Not applicable.
Iron concretions
There is a concretion of an iron bolt halfway along the mast step, at the centre of its width. Iron nails of round section, practically unknown in ancient ship construction, were used significantly to fasten the ceiling and some frames. No bronze or copper nails are present.
Hull Remains
The general state of the ship is excellent. The hull has flat-bottomed lines with a hard turn of the bilge, suitable for crossing sandbars and sailing in shallow waters. It apparently has no bilge pump (Figure 4).
Figure 4. View of the central part of the hull after mechanical cleaning of the vine shoots and resin, with the corresponding identification labels (CASC).
Keel, keelson and mast step
The keel was joined to the sternpost by a trait de Jupiter (rayo de Júpiter) scarf, whose rising part has been lost. The sternpost is in turn made of two pieces joined vertically by tenons, the upper, curved piece being the sternpost proper. Nieto and Foerster’s initial opinion was that the main piece was the stem and that the lower, triangular piece, 63 cm long and 12 cm wide, might be related to a cutwater, to make sailing easier by reducing resistance to forward motion. Its base is heavily eroded, and its narrowest part, which should have had a notch for the joint with the keel shoe, has been lost. The discovery of the mast step and the interpretation of the orientation of its mortises for stepping the mast allowed Nieto (1982: 165) to deduce that the piece interpreted as the stem was in fact the sternpost, and that the southern part of the wreck was the stern. The author therefore proposes that this triangular piece is the sternpost heel or skeg (durmiente del codaste or alerón de deriva), as seen in the Kyrenia and Palamós wrecks. The keel is not very robust, 12 cm wide and only 6 cm high (complemented by a further 2.5 cm of shoe), with a flat rectangular section. Of great interest for comparison with other wrecks is the fact that along its length it does not seem to have a rabbet for the garboard strake, which makes for a very flat-bottomed hull.
The sternpost has a trapezoidal cross-section, with the underside narrower, and an average height of 19 cm (Figure 5). It curves as it moves away from the keel, at the same time increasing in width. The upper face of the sternpost increases from 12 cm at the joint with the keel to 16 cm at the opposite preserved end. At its proximal end a sternpost foot is distinguished, marked by a noticeable reduction in thickness with a rounded moulding, and on its underside there is a notch to house the keel shoe. It has a series of notches and recesses to form the trait de Jupiter scarf with the keel (with the help of metal nails); the lateral locking key was not found (Figure 5). On both sides of its upper face the sternpost has a notched rabbet that changes from a gentle simple bevel at the joint with the keel to a pronounced notch, allowing the strakes to meet it at a sharp angle at the stern. The edge-joining to the strakes by punicana coagmenta, with mortises 5 cm wide and spaced 8 cm apart, is improved by nails placed between every two mortises and oriented to fasten the strake to each face of the rabbet (Figure 5).
The base of the keel is protected by a shoe 12 cm wide and 2.5 cm thick, with a notch to fit the sternpost heel. It is fastened to the keel only by metal nails, with no tenon or wooden peg. Its underside is very eroded, which was interpreted as wear from beaching or passage through shallow areas. In 2012, heavy erosion of the starboard garboard strake revealed its strong rectangular section and the absence, or minimal angle, of the rabbet in a keel that is narrower than the first strakes and wider than it is high (Figure 6).
Figure 5. Exploded axonometric view of the sternpost: 1, sternpost; 2, sternpost foot; 3, after end of the keel; 4, keel shoe; 5, sternpost heel; 6, idealised section of the sternpost and heel, showing the two systems used to fasten the first hull plank to the sternpost (after Nieto and Foerster 1982: 167).
Figure 6. Heavy erosion of the starboard garboard strake, showing its strong rectangular section and the absence, or minimal angle, of the rabbet in a keel narrower than the first strakes and wider than it is high (CASC).
At the bow, the upper face of the stem at the trait de Jupiter scarf is wider and thicker, as in wrecks of its period (e.g. Tradelière or Rabiou), but it begins with a moulding without parallel among wrecks of the period (Figure 7), such as the Barthélémy B, of the same architectural signature. On the upper face of the stem, at the scarf joining it to the keel, the joint is reinforced with a nail (Figure 8). The stem has a notched rabbet (Figure 9). Near the joint with the stem, two faceted conical treenails were driven vertically from the outside into the seam between the keel and the garboard strake (Figure 10).
Figure 7. The moulding forming the upper face of the stem (CASC).
Figure 8. Upper face of the trait de Jupiter scarf between the keel and the stem. Note the change in the wood species selected and the iron nail reinforcing the joint (CASC).
Figure 9. The notched rabbet in the stem of the wreck (CASC).
Figure 10. The two wooden treenails projecting from the outside, driven right in the seam between the keel and the garboard strakes, which support a piece of unclear interpretation with no known parallels (CASC).
The mast step increases in width from 17 cm at its distal end to 34 cm in the area where there are four mortises for seating and securing the foot of the mast (Figure 11). The south face of the central hole is vertical, while the opposite face is concave. The mortises in the side faces of the main one are of similar size and trapezoidal section (Figure 12). The mortise that prevents longitudinal movement of the mast also has one vertical and one concave face, but arranged inversely to those of the main mortise. The mast step also has a 3 × 5 cm mortise to house the tenon of a stanchion.
The interpretation is that the vertical face of the main mortise faced the stern. This is justified by the mechanical stress exerted by the sail during navigation, with parallels in cases where the position of bow and stern is clearly defined (e.g. Saint Gervais 3), and by the ease it gives for lowering the mast aft, of which there are some examples in iconography. The direct parallel for this mast step is in the Plane 1 wreck, of the same architectural signature as the Cap del Vol but with marked differences. In the main cavity of the mast step, standing on edge with its obverse facing where the tenon of the mast should have been and its reverse resting against the vertical south face, a very worn votive coin was found, possibly of Saguntum1 (Carlson 2007: 319) (Figure 13), although Bolskan seems the most likely identification.
For seating the mast step on the framing, a series of grooves about 19 cm apart were cut in its underside, to spread through the hull the multidirectional forces exerted by the mast.2 On its sides there were slots for fitting the first ceiling planks, without known parallel, which in 2011 could be interpreted as longitudinal erosion. The mast step is a single timber of considerable width, thickness and length, without known archaeological parallel. It shares features of its fitting to the framing with wrecks such as the Plane 1, but not its dimensions; it shares its dimensions with the Cavalière, but not the system for housing the tenon of the mast. The mast step also serves as a keelson, being practically as long as the keel from stem to sternpost. The hold was very open, with possible stanchions only at the ends of the mast step.
Figure 11. Axonometric view of the mast step on the frames, with a groove for fitting the first ceiling strakes (after Nieto 1982: 166). The author’s own observation of the wreck in 2011 and 2012 concluded that this is longitudinal erosion caused by the first ceiling planks, either during the ship’s life or during the post-depositional process.
Figure 12. Mortises for securing the foot of the mast, and sections (after Nieto 1982: 166).
Figure 13. Reverse of the coin found in the mast step of the Cap del Vol wreck (photo: courtesy of the Centre d’Arqueologia Subaquàtica de Catalunya).
Frames
The framing shows an alternation of floor timbers and half-frames3 every 25 cm, at least from the distal end of the remains up to the mast step (Figure 1). The frames, of holm oak (encina), are of rectangular section and well finished, with average dimensions of 10 cm sided and 16 cm moulded. They are fastened to the hull by treenails placed rhythmically in pairs alternating with single treenails, and these never serve to fasten the floor timbers to the stem or the keel, although the floor timbers are in contact with them. The presence of a small groove joining the two holes of the pairs of treenails on the outer face of the strakes, together with what was interpreted as a small protective and locking wedge (Figure 14), seems to be a variant of the technique of internal lashings locked by treenails, with other parallels such as Saintes Maries 24. This allowed Marlier4 to establish that the frames were fastened by lashings secured with treenails, which placed the Cap del Vol within the same architectural signature as the Cap Béar 3.
Figure 14 shows a diagram of this fastening of the framing in the upperworks, which differs from that recorded in the bottom only in that metal nails are used alternately on the half-frames. The first two floor timbers (nos. 1 and 2 in Figure 15) have a sharper angle, since they are at the end of the ship in direct contact with the sternpost, with the limber holes at its edges. Floor timber no. 3 lies at the level of the joint between the sternpost and the keel, with a centred limber hole. It has only a gentle angle in section and there is no rabbet in the keel, which allowed the author to argue, and to confirm in 2012, that in the central part of the hull the floor timbers, and therefore the lines of the ship, are completely flat. It would thus be a flat-bottomed ship, in principle less suited to open-sea navigation in adverse conditions, but very well suited to sailing in shallow waters, taking refuge in rivers or lagoons and being beached (Nieto and Foerster 1980: 170). Figures 16 and 17 show the bow with the first floor timbers and a looped internal lashing locked by treenails.
The 2012 campaign added the following observations on the framing:
The system of fastening the frames to the hull uses a pattern of pairs of treenails and single treenails that does not match that of most ships of the architectural signature to which the Cap del Vol belongs. Some of the single treenails, or those of the pairs, had to be secured or consolidated afterwards by an iron nail with a protruding point, using the treenail as a guide plug. Blind nails (a punta perdida) seem to be present in the framing. Isolated wooden plugs were placed between the pairs of treenails fastening the frames to the hull, in the groove for the internal lashings (Figure 18). The treenails driven from the inside of the hull are often not trimmed and project above the upper face of the frames (Figure 19). Some treenails in the frames are of rectangular section and others of large diameter. In several cases the internal lashings have no groove joining the pairs of treenails on the upper face of the frames (Figure 19). In no case are active frames detected.
On the upper faces of the frames observed there is work possibly intended for external lashings that were never made, either by bevelling the corners of the upper face or by cutting grooves (Figure 20). These may be frames reused from a ship of lines identical to the Cap del Vol, or the ship may come from a shipyard where both technical traditions coexisted. The wreck provides the earliest known evidence of offset or asymmetrical half-frames (cuadernas sobrepasadas), a defining feature of the Imperial architectural family (Figure 21). Some details of frame Q115 are summarised in Figure 22.
Bundles of vine shoots were placed in the spaces between the frames (Figure 23) to keep the circulation of water in the bilge free of small objects. Parallels are scarce: only the Cap Bénat (for which an origin in southern Europe has been proposed without scientific justification), Saintes Maries 2 and Saintes Maries 24 (the last two of the same architectural signature as the Cap del Vol).
Figure 14. Diagram of how the frames are fastened to the planking strakes by pairs of cylindrical treenails. In the upperworks nails and treenails are used alternately (after Nieto and Foerster 1980: 170).
Figure 15. Floor timbers at the southern end of the wreck, showing the rhythm of the treenails (after Nieto and Foerster 1980: 171); compare, for example, with the Palamós wreck.
Figure 16. View of the bow with the first floor timbers (CASC).
Figure 17. Top view of a fastening between frame and hull by a looped internal lashing locked by treenails (CASC).
Figure 18. Top view of one of the internal lashings, locked on its upper face by a wooden plug or wedge (CASC).
Figure 19. A pair of projecting treenails without a groove joining them to protect the lashing, which has disappeared (CASC).
Figure 20. Corners of a frame cut back to allow the passage of an external lashing (CASC).
Figure 21. Off-centre half-frames in the arrangement and rhythm of the framing (CASC).
Figure 22. Summary sketch with some details of frame Q115 (drawing: C. de Juan Fuertes).
Figure 23. Vine shoots laid in an orderly way between the frames (CASC).
Planking
The planking strakes are 4 cm thick,5 with mortises 8 cm wide and 1 cm thick, 10 per linear metre (Foerster and Pascual 1982: 54). The holes drilled with a bow drill for the pegs locking the tenons are very conical, without archaeological parallel because of the great difference between their maximum and minimum diameters (Figure 24). The pegs driven from the inside of the hull are often not trimmed and project. Some of them seem to have been replaced afterwards by others of larger inner diameter, as a repair at the time of assembly (Figure 24). There are also faceted pegs of hexagonal section, with an inner diameter smaller than the usual average.
The hull has a hexagonal repair, showing the position of the tenons inserted afterwards (Figure 25). The repair tenons have no upper heel for hammering; they are of ordinary shape, with some working to give them the necessary curved profile (Figure 26).
Figure 24. Truncated-cone locking pegs for the tenons, whose diameter varies according to the direction in which they were driven. Some were not completely trimmed (CASC).
Figure 25. Hexagonal repair of the hull, showing the position of the tenons inserted afterwards (CASC).
Figure 26. Close-up of a repair tenon, which has no upper heel for hammering but is of ordinary shape, worked to give it the necessary curved profile (CASC).
Ceiling, thick stuff, and wales
The vine shoots were deliberately placed, as recorded under one of the ceiling planks (Abies sp.). In the excavators’ opinion (Nieto and Foerster 1980: 166), their function could be similar to that of a grating preventing the limber holes from becoming clogged with debris, so that water could circulate in the bilge.6
The ceiling consists of a fixed stringer alternating with removable ceiling strakes; two lengths of plank in the same strake are joined by bevelled scarfs fastened with two iron nails. The stringer is systematically nailed to each frame with three or four nails (Figure 27). Round-section iron nails, practically unknown in ancient ship construction, were used significantly to fasten the ceiling and some frames (Figure 27).
Figure 27. The ceiling of the hold, consisting of fixed stringers fastened with sets of three to four nails, alternating with thinner removable ceiling strakes (CASC).
Caulking
The inside of the hull has a coating of resin applied after the ship had been built, 0.5 to 2 cm thick, in which wood shavings could be seen together with long animal hairs reinforcing this waterproof substance. The 2012 observations suggest that the layer protecting the wood from sea moisture is a wax rather than a resinous substance (Figure 28).
Figure 28. Protective resin covering the whole interior of the hull, applied after it had been completely assembled (CASC).
Size and scantlings
No estimate of the ship’s length is given. The keel is light (12 × 6 cm, plus a 2.5 cm shoe), with no rabbet along its length, and the floor timbers in the central part of the hull are completely flat: a flat-bottomed ship with a hard turn of the bilge, suited to shallow waters, rivers and lagoons and to being beached.
Table 1. Scantling of the timber remains of the Cap del Vol Shipwreck
| Timber | Sided [cm] | Molded [cm] |
| Keel Keel shoe | 12 12 | 6 2.5 |
| Sternpost Sternpost heel | 12–16 12 | 19 (avg.) – |
| Keelson | – | – |
| Maststep Stanchion mortise | 17–34 3 × 5 | – – |
| Buttresses | – | – |
| Frames | 10 (avg.) | 16 (avg.) |
| Futtocks | – | – |
| Room-and-space At central frames | 25 | |
| Planking | – | 4 |
| Ceiling | – | – |
| Main mast Ø Chock | – – | – – |
The sternpost heel is 63 cm long. Planking: mortises 8 cm wide and 1 cm thick, 10 per linear metre; the planking is thinner in the upperworks.
Wood
Frames: holm oak (encina, Quercus ilex). Ceiling plank: fir (Abies sp.). The wood species changes at the scarf between the keel and the stem (Figure 8).
Bibliography
Carlson, D. N. 2007. Mast-step coins among the Romans. The International Journal of Nautical Archaeology 36.2: 317–324.
Foerster, F. 1980. A Roman wreck off Cap del Vol, Gerona, Spain. The International Journal of Nautical Archaeology 9.3: 244–253.
Foerster, F. and Pascual, R. 1982. El pecio del Cap del Vol. Vida Submarina 5: 47–56.
Marlier, S. 2003. Unpublished doctoral thesis. [Full reference to be completed.]
Nieto, X. 1982. El pecio del Cap del Vol. Nuevas aportaciones. Cypsela 4: 165–168.
Nieto, X. and Foerster, F. 1980. El pecio romano del Cap del Vol: campañas de 1978 y 1979. Cypsela 3: 163–177.
Nieto, X. and Foerster, F. 1982. [Full reference to be completed.]
Notes
1 Since these are Iberian coins, possibly placed following a ritual, this may indicate that the Cap del Vol and the Plane 1 were ships built in the Iberian Peninsula.
2 Where the longitudinal vectors are dominant.
3 “Continuous and discontinuous frames” in the publication (Nieto and Foerster 1980: 165).
4 Unpublished doctoral thesis, consulted with the personal permission of its author, whom the author thanks for her collaboration. The piece interpreted as a wedge is indeed one, as the author was able to document in 2012.
5 Thinner in the upperworks.
6 A technique similar to that documented by Marlier in the SM2 and SM24 wrecks.