Fiumicino 1, 4th–5th century AD
Country: Italy
Place: Harbour of Claudius, Fiumicino (Rome); found on land during the construction of Rome airport. Now in the Museo delle Navi, Fiumicino
Coordinates (approximate): 41.78, 12.25
Type: Roman river barge (navis caudicaria), adapted to fluvio-maritime navigation
Identified: No
Dated: 4th–5th century AD, based on the associated finds, consistent with certain features of its construction.
Beam: 4.75 m.
Keel Length: 11.11 m.
Length Overall: 13.83 m.
Number of Masts: One towing mast, stepped well forward.
History of the wreck
The Fiumicino 1 wreck, preserved in the Museo delle Navi in the town of that name, was found together with four others in 1960 during the construction of Rome airport, in a peripheral area of abandonment or storage in the harbour basin of the emperor Claudius, built in AD 42. It has recently been the subject of an extensive study by Boetto (2009), which supersedes earlier ones.1 The ship is dated by the associated finds, consistent with certain features of its construction, to the 4th–5th centuries. The wreck has undergone several processes of transport, reassembly and restoration (including some that used timbers of the period), which makes its study somewhat difficult today; nevertheless, the results provided by Boetto are of excellent quality and resolve the main earlier doubts. It is a caudicaria, adapted to fluvio-maritime navigation.2
The analysis of a large number of ancient repairs to the hull has allowed Boetto to conclude that the ship was in use for a long time. Its burden would have been 49.7 t, and it would have had a towing mast rather than a sail. Boetto considers the vessel a caudicaria on the basis of these elements and of its lines, with a high, twisted stern and a smooth, raked bow (Boetto 2009: 55) (Figure 8). A fire-branded inscription recorded during the original excavation, now lost, gives a frame the Latin epithet Trituta (“saved three times”), surely a carpenter’s mark like those documented on the Saint Gervais 3 (Liou and Gassend 1990: 234, 245, 251, 252).
Description of the site
The Fiumicino 1 is 13.83 m long and 4.75 m in beam, with a maximum height of 1.45 m on the starboard side (Figure 1). It preserves the keel, with its extension at the stern by a section of sternpost, 13 strakes of planking (7 on the starboard side and 6 on the port side), 42 frames with three futtocks, a mast step and two fragments of a stringer. In longitudinal section it has a raked, smooth bow and a more angular, pronounced stern (Figure 1). The hull has a flat section, with more angular lines in the transverse sections at the stern than at the bow.
Figure 1. Plan and longitudinal section of the Fiumicino 1 wreck (drawings: Boetto, CCJ-CNRS, after Boetto 2009: 32–33).
Cargo
Not reported.
Ballast
Not reported.
Anchors
Not reported.
Guns
Not applicable.
Iron concretions
Not reported. Iron nails and bolts were widely used in the construction (see below).
Hull Remains
The hull is preserved almost complete over its length, with the keel and its sternpost extension, 13 strakes, 42 frames, three futtocks, the mast step, two fragments of a stringer and an unconnected fragment of wale (Figure 1).
Keel, keelson and mast step
The keel (Quercus) is joined at the stern by a trait de Jupiter (rayo de Júpiter) scarf to the sternpost extension piece. Such a transition piece is very commonly found at the bow of seagoing ships, forming the stem. The keel is 11.11 m long and changes along its length. It has a rectangular section at the scarf, 13 cm wide and 17 cm high, a trapezoidal section at the point of maximum beam (12 cm wide at the top, 9 cm at the bottom and 15 cm high) and a square section at the bow, 10 cm a side, with no rabbet at all along its length (Figure 2). The extension piece forming the sternpost does have a notched rabbet (asymmetrical on the starboard and port sides). It is 2.72 m long, 10/12 cm wide and 16 cm high, and ends in a trait de Jupiter scarf joining it to the sternpost proper. Although no bolt reinforces the joint, Boetto found that three wooden treenails, acting as nails, joined the two pieces, a technique also documented in the Madrague de Giens (Rival 1991: 165, fig. 31-1).
In the Fiumicino 1 there are six iron bolts driven perpendicularly along the keel. One of them is in the sternpost extension piece, where it fastened a floor timber (now lost) to the keel, while the other five fasten the framing to the keel. Originally two of these bolts also fastened the mast step over the floor timbers. They are large iron nails driven through the outer face of the keel, in which recesses were cut so that they would not project,3 with large heads, 4 to 4.5 cm in diameter. According to Boetto they are very similar to those of the Laurons 2 wreck (Gassend et al. 1984: 91, figs. 17a and 19). The bolts are of decreasing circular section, from 2.5 to 1.6 cm, becoming square in section (1.2 cm a side) at the level of the floor timber (Figure 3). A treenail is also used as a guide and plug, and the point of the bolt was probably clenched over the upper face of the floor timbers. Their position partly blocks the limber holes of the floor timbers.
The mast step (Quercus) is preserved over a length of 2.75 m, with widths of 5 to 10 cm and a height of 15 cm. It was broken during the excavation. On its underside it has grooves to fit over the upper faces of the floor timbers, with rectangular notches, and over the half-frames, with semicircular notches. According to Boetto, its correct position (not the one in which it is displayed today) is further forward, with holes in the mast step 1.70 m apart coinciding with the forward bolts mentioned above. The upper face of the mast step has two cavities: the first, square, 5 cm a side and 3.5 cm deep, would have housed a stanchion. The second, rectangular, 15 cm long, 5 cm wide and 6 cm deep, is 50 cm further forward. It has the characteristic profile decreasing towards the bow, for stepping the foot of the mast with the possibility of lowering it aft. Its forward position led Boetto (2009: 46) to propose that this vessel could not have sailed with a square sail, and perhaps had a pole or towing mast to allow it to be towed along the river bank.
Figure 2. Changes in the cross-section of the keel, which has no rabbet (drawings: Boetto, CCJ-CNRS, after Boetto 2009: 35).
Figure 3. Longitudinal section, with associated cross-sections of the keel, at the bolts joining the longitudinal and transverse timbers (drawing: Boetto, CCJ-CNRS, after Boetto 2009: 44).
Frames
The framing consists of 42 floor timbers (Quercus) and half-frames (Quercus and, in a few cases, Quercus ilex), with the impressions of three more frames on the hull. On the starboard side, towards the bow, three futtocks were preserved. The frames in the central part of the hull have broken ends, as do the half-frames at the bow. The floor timbers, by contrast, are complete and have no connection with the futtocks. The alternation of floor timbers and half-frames is regular, except at the stern, where two consecutive floor timbers were recorded. The room-and-space (malla) is irregular, varying between 12 and 30 cm (average 19 cm). The floor timbers, of oak, have rectangular or trapezoidal sections, 6 to 10.5 cm sided and 8 to 18 cm moulded, with the smallest moulded dimension right in the central part of the vessel (9/10 cm) (Figure 4). A complex piece was found in the area of the sternpost extension, 1.40 m long, 11 cm wide and 11 cm high, cut from a quarter of a trunk, probably of pine, with traces of small nails, perhaps of copper, on one of its faces. Its angular shape led Boetto to interpret it as a crutch from the upper part of the stern.
The half-frames, 6 to 12 cm moulded and 6 to 10 cm sided, have shapes similar to those of the floor timbers and are also of oak. The distance between them over the keel is irregular: at the bow their ends touch the keel itself, with gaps of less than 4 cm, to the point of almost touching each other. Boetto (2009: 43) calls them demi-couples rapprochés, reinforcing the bow (Figure 5).
The frames have limber holes over the keel and also at the level of the third strake, where the hull is still flat, in all cases rectangular (5 × 3 cm). In the area of the sternpost extension, the floor timbers have triangular limber holes over the garboard strakes. The closely spaced half-frames also allow water to pass thanks to half limber holes cut over the keel.
The framing is fastened to the hull by wooden treenails that act as guides or plugs for iron nails. Boetto explains that this reduces the corrosion of the nails, which would otherwise be in direct contact with tannin-rich species such as oak, and prevents the wood from splitting when the nails are driven4 (Figure 6). The fastenings are regularly spaced at 10.5 cm, with a maximum of four per strake of planking. The inner and outer diameters of the treenail-plugs are between 1.3 and 1.6 cm. The nail shanks are of square section, 1 to 1.2 cm a side, and the heads between 3.8 and 4.4 cm in diameter. The upper faces of the frames, especially at the stern on the starboard side, were dressed with an adze so that the points of the nails would not project, although some of them were not cut off but bent over like a staple. The nails broke during the recovery, and their heads and any points on the upper faces of the frames were lost. The framing was fastened to the wale only by wooden treenails 1 cm in diameter.
Figure 4. Cross-sections of the Fiumicino 1 (drawing: Boetto, CCJ-CNRS, after Boetto 2009: 34).
Figure 5. Arrangement of the framing of the Fiumicino 1, with closely spaced half-frames (demi-couples rapprochés) (drawing: Boetto, CCJ-CNRS, after Boetto 2009: 42).
Figure 6. Diagram of the technique for fastening the frames to the hull in the Fiumicino 1 (drawing: Boetto, CCJ-CNRS, after Boetto 2009: 44).
Planking
Among the planking strakes that make up the hull, the garboard strakes, the ordinary strakes and a fragment of wale are preserved. The garboard strakes have a simple rectangular section. On the starboard side this strake is made of two lengths: the main one is 9.53 m long and the secondary one, at the bow, 4.37 m, with a width of 30 cm and a thickness of 4.5 to 3.5 cm; the two planks are joined by a bevelled scarf. On the port side the garboard strake is of classic type, a single plank from bow to stern, 12.51 m long, 24 cm wide and 3 cm thick. It has a thin correcting and/or graving strip at the joint with the second strake on that side, resulting from a repair.
The Fiumicino 1 has 34 planks forming 6 strakes on the starboard side and 5 on the port side, the hull being clearly divided into stern, central and bow sections. This allowed Boetto to establish that the design principle and building process were the classic ones of Antiquity, but that the process was based on the prior erection of the keel, the garboard strakes and the central strakes before the bow and stern were defined (bordage médian premier; Boetto 2009: 50). The central planks on the starboard side (Pinus pinea L., and in one case Cupressus sempervirens L.), about 6 m long, 29 to 40 cm wide and 3 to 5.4 cm thick, end at practically the same point in the hull, in bevels or in hooked joints, where two narrower planks are joined to the main central strake to close the lines of the bow and stern. Most of the planks forming the bow and stern are of oak (Quercus). During construction, corrections were made to give both sides of the hull a certain symmetry, mainly by adding a plank to the third strake on the starboard side at the bow.
The planking is generally edge-joined by the classic mortise-and-tenon technique, but there are numerous cases in which the joint has no locking pegs, more frequent in the joint between the keel and the garboard strake and in the bow area, although there are scattered cases all over the hull, as well as joints in which only one of the two locking pegs was fitted. In the case of the keel, Boetto proposes an ancient repair involving the replacement of the keel and the port garboard strake (Boetto 2009: 47), which would not be decisive for a vessel whose sphere of navigation was restricted to the river Tiber and its mouth. The replacement of strakes may be related to its use at quays.
The mortises have an average width of 7.4 cm, a depth of 4.5 cm and a thickness between 0.7 and 1 cm. The tenons (Quercus ilex) are on average 8 to 9 cm long, 4.3 cm wide and 0.6 cm thick. The spacing between mortises varied greatly, from 16.2 to 74 cm (average 41.5 cm), which shows that their role was evolving in Late Antiquity. The pegs, of holm oak and cypress and driven from the inside of the hull, have a truncated-cone profile and a faceted or prismatic section. Their inner diameter is between 1.1 and 1.3 cm and their outer diameter between 0.7 and 1 cm. The spacing of the pegs along the seams was very irregular, from 17.5 to 76 cm (average 36 cm).
On the starboard side, under the frames joined to the keel by bolts (M14 and 24), Boetto (2008: 39) found that three of the locking pegs were cylindrical (same inner and outer diameter) and that in one the outer diameter exceeded the inner. This could indicate that these pegs were driven with the frames already fastened to the hull when these planks were fitted, but according to Boetto the evidence is too limited to indicate a particular building method. Many of the pegs recorded had a characteristic mark of friction and wear against the tenon (Figure 7). The analysis of the wale, its mortises and their spacing, with the absence of locking pegs, could indicate that in the upper parts of the hull there were tenons without locking pegs.
At the joint of the strakes with the sternpost extension there are iron nails, driven perpendicularly; the presence of treenails could not be determined because of the restoration processes. On the starboard side the nails (of square section, 0.9 cm a side, with circular heads 2 cm in diameter) are set in tetrahedral recesses, and only one half-moon-shaped recess for a nail was recorded on the port side. The particularity of this technique, without parallels in Roman ship construction (but with parallels in Gallo-Roman river craft), could be related to the existence of an architectural signature (Boetto 2009: 41), certainly linked to a fluvial sphere of navigation.
Figure 7. Tenons and pegs from the hull (drawing: Boetto, CCJ-CNRS, after Boetto 2009: 40).
Ceiling, thick stuff, and wales
The internal longitudinal structure consists of a mast step and two fragments of a stringer (Pinus pinea L.), with nail holes indicating its fastening to the frames, which have no notches for fitting stringers.
An unconnected fragment of wale was found during the excavation. It is 4.52 m long, 16 cm wide and 9 cm thick, with a rounded outer face worked with saw and adze. On its sides are lines scribed by the carpenters to mark the position of the adjacent planks. It preserves two rectangular notches, 11.5 and 14.5 cm long, 2 cm wide and 1.2 cm deep. Spaced 2.10 m apart, they are related to the seating of two beams.
Caulking
Not reported.
Size and scantlings
The Fiumicino 1 is 13.83 m long and 4.75 m in beam, with a maximum height on the starboard side of 1.45 m and a burden of 49.7 t. It has a flat-bottomed hull with a keel without rabbet, a raked bow and a high, twisted stern, with a towing mast rather than a sail (Figure 8).
Figure 8. Bas-relief of the 3rd century in Salerno cathedral (after Ucelli 1950: 375).
Table 1. Scantling of the timber remains of the Fiumicino 1 Shipwreck
| Timber | Sided [cm] | Molded [cm] |
| Keel At the scarf At maximum beam At the bow | 13 9–12 10 | 17 15 10 |
| Sternpost extension | 10–12 | 16 |
| Keelson | – | – |
| Maststep | 5–10 | 15 |
| Buttresses | – | – |
| Floor timbers | 6–10.5 | 8–18 |
| Half-frames | 6–10 | 6–12 |
| Room-and-space At central frames | 12–30 (avg. 19) | |
| Planking Garboards | 29–40 (central) 24–30 | 3–5.4 3–4.5 |
| Wale | 16 | 9 |
| Ceiling | – | – |
| Main mast Ø Chock | – – | – – |
The mast step is preserved over 2.75 m; the sternpost extension is 2.72 m long. Planking and wale: sided = width, moulded = thickness.
Wood
Keel, frames and mast step: oak (Quercus); a few half-frames of holm oak (Quercus ilex). Central planks: stone pine (Pinus pinea L.) and, in one case, cypress (Cupressus sempervirens L.); bow and stern planks mostly of oak. Stringer: stone pine (Pinus pinea L.). Tenons: holm oak (Quercus ilex). Pegs: holm oak and cypress.
Bibliography
Boetto, G. 2009. L’épave de l’Antiquité tardive Fiumicino 1: analyse de la structure et étude fonctionnelle. Archaeonautica 15: 29–62.
Gassend, J.-M., Liou, B. and Ximénès, S. 1984. L’épave 2 de l’anse des Laurons (Martigues, Bouches-du-Rhône). Archaeonautica 4: 75–105.
Liou, B. and Gassend, J.-M. 1990. L’épave Saint-Gervais 3 à Fos-sur-Mer (milieu du IIe siècle ap. J.-C.). Inscriptions peintes sur amphores de Bétique. Vestiges de la coque. Archaeonautica 10: 157–264.
Rival, M. 1991. La charpenterie navale romaine. Matériaux, méthodes, moyens. Travaux du Centre Camille Jullian 6. Paris: CNRS.
Ucelli, G. 1950. Le navi di Nemi. Rome: Libreria dello Stato.
Notes
1 All the information on this wreck given here has been taken from her work published in Archaeonautica 15.
2 It is included in this work because its nautical space is, by its very function, very limited, so its shipyard must also have been in the area of Ostia.
3 Thus avoiding damage to the heads from the friction of the keel against the sand in shallow areas.
4 An explanation that Boetto takes from Ucelli (1950: 151), in his work on the Nemi ships. Without rejecting this argument, the author considers that the purpose of this technique was to improve the fastening of the frames to the hull still further, the wooden treenail acting as plug and guide for the nail. Moreover, the technique is documented in wrecks where the frames are not of oak, as in the treenails of the Jeaune-Garde B. The author believes it is related to a building technique typical of the Italian peninsula.