SEARCH

SEARCH BY CITATION

Keywords:

  • masts;
  • rudder-stocks;
  • heel;
  • mast-ladder;
  • scarf;
  • mortise-and-tenon

Abstract

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References

This paper presents both archaeological evidence and technical features that allows the identification of two of the wooden objects discovered in 2001 at the site of a 1st-century-AD shipyard in Olbia, Sardinia, as masts. The structure and dimensions of the objects are presented and compared with similar artefacts which have been interpreted as rudder-stocks in order to establish their specific and distinct features. Possible retaining and lowering systems for the masts, and evidence of wooden mast-ladders are discussed through examination of archaeological parallels and Roman iconography.

At the 2012 ISBSA conference in Amsterdam, the presentation of some of the wooden artefacts discovered at the port of Olbia, Sardinia, Italy in 2001 (Fig. 1) (Riccardi et al., forthcoming) stimulated an interesting discussion over their correct identification. While Edoardo Riccardi and Virgilio Gavini, who found the objects, suggested that they are incomplete parts of two masts (A1 and A2) and four rudder-stocks (T1, T2, T3 and T4), some scholars thought that all six should be identified as rudder-stocks1. Here the dimensions and features of the two sets of artefacts are compared, followed by a discussion of how specific features would enable A1 and A2 to be used as masts through comparisons with both other archaeological material and iconographic evidence.

figure

Figure 1. Location map and the area where the timbers were found in the port of Olbia. (Drawing F. Tiboni, V. Gavini)

Download figure to PowerPoint

Excavations were carried out in rescue conditions prior to the construction of a tunnel in 2001 (Riccardi, 2002: 268). The site, characterized by the remains of many wooden elements from the structures of laid-up ships, all dating to the Neronian/Vespasian era, has been interpreted by the excavator as part of a shipyard, and dated to the second half of the 1st century AD (D'Oriano et al., 2002: 1256). Archaeologists also uncovered the remains of a probable slipway, and many tools and artefacts undoubtedly from a wooden-shipbuilding context (D'Oriano, 2004).

At the time of discovery, different interpretations of the six wooden elements discussed here were proposed through the analysis of their general shape and structure, and in the light of their chronological affinity (Riccardi et al., forthcoming).

A1 and A2 (Fig. 2)

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References
figure

Figure 2. a) Drawing of A1 (V. Gavini); b) A1 as found on site (E. Riccardi); c) A2 as found on site. (E. Riccardi)

Download figure to PowerPoint

Only A1 has been fully recorded as, unfortunately, A2 was partially damaged during its removal from site. Being almost identical, a single description of the two finds can be made. A1 is preserved for a length of c.7.87 m, is broken at one end, and has a diameter of 0.42 m. The timber has a very particular section: for the first 1.34 m it is octagonal, after which it is circular to its end. At c.2.32 m from the base, A1 has two rows of diametrically opposed carved mortises arranged in a regular staggered pattern, c.0.26–0.29 m apart. The foot of A1 presents a semi-circular central tenon, with a circular hole in its centre (diameter c.0.09 m), and angled shoulders giving the base of the timber, either side of the tenon, an ‘arrow’ shape. Half-way up the octagonal section, in the same alignment as the circular hole, there is a rectangular passing mortise measuring c.0.20 × 0.11 m.

A1 and A2 are interpreted as masts with the semi-circular tenon and arrow-shaped scarf acting as the heel, the rectangular mortise taking some manner of support or partner and the two rows of staggered mortises used to join a mast-ladder.

T1–4 (Fig. 3)

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References
figure

Figure 3. T2, T1 and T4 from the port of Olbia. (Drawing V. Gavini)

Download figure to PowerPoint

Only T1, T2 and T4 are sufficiently well preserved to have been fully studied. From a technical point of view, they are each made from a single piece of wood, have a rectangular mortise in the wider part, and a marked, even taper along their lengths. The section of the timbers changes from circular in the wider part, to oval towards the narrower end, creating flat surfaces on opposing sides. The three timbers have lengths of 5.23–9.95 m, and diameters not exceeding 0.32 m, tapering to 0.12 m. Two rows of mortises, not more than 0.4 m distant from each other can be noted carved into the opposing flatter sides of each timber. The mortises are not aligned in pairs but are staggered, with the distances between them decreasing towards the narrow end. In fact, the final two mortises are almost touching in one corner, and occupy almost the whole width of the timber.

Only in one case, (T4), is the wider end of the timber entirely preserved, permitting observation of the rectangular mortise, which measures 0.18 m by 0.08 m. T2 appears to be broken at just this point, while the wider end of T1 is also broken. This might suggest that these artefacts broke at this point in antiquity and were thus abandoned.

In interpreting these artefacts as rudder-stocks, with reference to other known archaeological evidence (Uccelli 1940: 168, fig. 179), it is considered that the rectangular mortise housed a tiller, the two rows of staggered mortises held pinned tenons connecting the rudder-blades, while the staggered system would reduce weakening of the stock. The distinct taper allowed the timber to be narrowed to the width of the rudder-blades at the lower end. Moreover, the break seen in T2 suggests that the rectangular mortise, the juncture with the tiller, was a point of maximum structural stress and a significant weak point.

Comparison of T1–4 and A1–2

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References

Having described the two classes of finds, we propose here a comparative analysis, as objections to our identifications have been based on two elements: the idea that A1 tapered towards one end, and secondly that A1 presents some features that make it similar to the rudder-stocks described.

Taper

While both sets of finds show some tapering, it should be stressed that of A1, as seen now displayed at the Olbia Museum, is largely a consequence of erosion of one side of the artefact as the result of the attack of micro-organisms while buried, and does not represent the original form of the artefact, or the will of the shipwright. Moreover, initial analysis of the artefact during excavation included establishing the degree of taper by measuring the circumference at regular intervals along the length of the timber. This showed that A1 does not taper towards one end, but has an almost regular profile, with a slight swelling in the centre of the preserved section. From this point to the two ends, the tapering recorded on A1 is less marked than that seen on T1, T2 and T4. If for the sake of argument, A1 were to be considered a rudder-stock of similar form to T1–4 but taking into account its shallow taper and greater diameter, it would have measured at least 16 m in length. Indeed, attempts at reconstructing A1 as a rudder-stock (Riccardi et al., forthcoming) based on the proportions of known Roman rudders has led to estimates that it would have been almost one-and-a-half times as long as that of the giant Nemi ship (Uccelli, 1940: 168, fig. 179) (Fig. 4).

figure

Figure 4. Reconstruction of rudders from Olbia and Nemi and a hypothetical reconstruction of A1 as a rudder-stock. (Drawing V. Gavini)

Download figure to PowerPoint

Profile

The detailed profiles of the two classes of objects differ greatly. T1, T2 and T4 all taper evenly from the end with the large mortise towards the staggered mortises, while also becoming oval in section. This is a common feature of rudder-stocks which taper to the width of the rudder-blades, (Mott, 1997: 41–53, Uccelli 1940: 168–70). In contrast, A1 and A2 have a more complicated profile: from the tenon end the diameter first expands and then decreases, creating a sort of entasis effect, often seen on Classical columns. The timbers do not have an oval shape, nor do they taper to the extent that they might match rudder-blades. Furthermore, the two flat-worked strips along the alignments of regularly spaced mortises occupy only a third of the timber, and this does not vary.

The octagonal shape of the first section of A1 has neither archaeological nor iconographic parallels among Roman rudders. Furthermore, the eight plane faces on this section of A1 and A2 would not seem a useful feature on a rudder-stock, as they would prevent it from turning smoothly, allowing no more than four turning angles for the rudder.

Rows of mortises

The way the rows of mortises are arranged also differs between the two categories of objects. While for the first group (T1–4) the distance between them decreases towards the narrower end, and from the start of the sequence, in the second group (A1–2) the distance is fixed at c.0.26–0.29 m. Thus, although a decreasing pattern could theoretically have been present in the missing portions of A1 and A2, this does not seem a convincing suggestion, even considering the proposed original dimensions of A1 and A2.

Other mortises and tenon

The presence of the two holes carved in the octagonal sections of A1 and A2, in contrast to the single mortise found in the circular section of T1–4, makes it difficult for us to propose a close parallel or correspondence between the two classes of artefacts. Even if the rectangular holes are similar in shape and size, in A1 and A2 they are carved at c.0.70 m from the end of the timber, while in T4 at only 0.30 m; the presence of the second smaller circular hole through the tenon at the base of A1 and A2, as well as by the presence of the semi-circular tenon itself would deny their having the same function.

It has been suggested that the circular hole could have served as a mortise for a second tiller or as the mortise for the main tiller, and that the rectangular hole served to fix the rudder-stock to the hull. This does not seem to be borne out by a functional analysis of the artefacts. As we have seen, the rudder-stocks discovered at Olbia seem to have been broken and abandoned in antiquity: these breaks occurred at or near the tiller-mortise, underlining the fragility of this point. By contrast, A1 and A2 are in perfect condition at both the rectangular and circular holes. The position of the latter, which passes only through the width of the semi-circular mortise at the base of the object, makes it an unlikely candidate as the tiller-mortise as it would have been considerably weaker than one which passed through the entire width of the timber. A survey of the known iconography (Casson, 1965; Basch, 1987; Mott, 1997) has failed to identify, the presence of a similar structural element carved in the middle of the base of the artefact on any of the Roman rudders portrayed.

Although similar in some aspects, detailed analysis thus confirms quite separate functions for each class of artefact.

Structural analysis of A1

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References

To clarify our identification of A1 and A2 as masts, A1 is now briefly analysed from a structural and functional perspective, with comparisons to other archaeological finds.

As we have seen, A1 presents a semi-circular tenon, a sort of lug or tooth, carved in the centre of its octagonal base. From a technical point of view, the semi-circular profile would have facilitated lowering, allowing the mast to rock forward on the heel once released from the mast-step. In fact, this semi-circular shape could have performed better in this respect than the quarter-circle of the heel of the Dramont E mast, which allowed mariners to remove it from the step with a lifting and sliding movement. Further, a semi-circular tenon would have entered the mast-step more easily if lowered from an upper-deck, as suggested by Roman iconography.

The profile of the heel (Fig. 2) is not a simple T-shape but is arrow-shaped, with angled shoulders; the base on both sides of the tenon is hollowed out at an angle which converges towards the centre of the mast. This complicated scarf, can be compared with mast-steps discovered on Roman wrecks from the port of Olbia, in particular, the R5 Sud wreck (Fig. 5). In this case, the mast-step has been cut into the keelson in a complex way: on both sides of the central mortise, the upper face of the keelson has been carved to create an ‘arrow-shaped’ profile into which a mast-heel shaped in the same way as A1 would fit. Unfortunately, the R5 Sud wreck has not yet been fully examined.

figure

Figure 5. The mast-step of R5 Sud found at the port of Olbia. (Photograph E. Riccardi)

Download figure to PowerPoint

Although the semi-circular tenon would not have served to counter the lever movement of the mast under sail, analysis of many mast-steps permits us to suggest that this function was usually played by the vertical shape of the after face of the mast-step. According to Santamaria (1984: 110) the Dramont E mast-step, which has three vertical sides, is designed to avoid the mast becoming detached through pressure of the wind on the sail. However, even in this, the only case in which a mast-step has been found with the mast in place, Santamaria underlined that the mast-step could not have been the only support, and that this role must have been played ‘by others elements found on the wreck-site’ (1984: 111–14).

As for the Olbia masts, it seems plausible that the rectangular mortise carved half-way up the octagonal section in the same axis as the circular hole in the centre of the tenon could have taken a transversal wood, or partner, inserted to avoid accidental vertical or oscillatory movement. However, the weight of the timber would itself have gone some way to retaining the mast in the mast-step, as well as its passage towards the deck of the boat.

It should be underlined that while the Dramont E wreck is dated to the 5th century AD, a period characterized by the increasing importance of framing in Roman shipbuilding (Pomey et al., 2012), the Olbia mast is dated to the Neronian/Vespasian era. Thus, it might be supposed that Roman shipwrights did not rely only on a non-structural keelson to retain the mast in position. Furthermore, even if the mast diameter (0.42 m) allows us to propose a height of c.12–15 m and consequently a great weight, it is possible that fitting the mast-heel tenon into the mast-step mortise might not have been enough to prevent movement. The archaeological elements do not permit us to exclude that this kind of mast was used with sister-keelsons and a mobile mast-step.

The possible use of a mobile mast-step placed over two sister-keelsons to support such a mast cannot be ruled out as the date (Neronian era) and the place (Western Mediterranean) are compatible with this technical solution (Pomey et al., 2012). In which case, we would expect to find additional elements used to fix the mast other than its weight and heel. These elements could have been similar to those described by Isidoro (Origines XIX II, 2 ‘Parastatae, stipites sunt pares stantes, quibus arbor continetur’) or those found in the Calanque de l'Ane wreck in Marseille (Ximenes-Moerman, 1998: 301–2, figs 2–3). It might be suggested that the octagonal shape of the lower end of the mast was easier to fix to a support than a round-sectioned timber, and the mortise in the middle of the octagonal section might relate to the use of a parastatae as mentioned in the Classical texts. Furthermore, we cannot exclude that the mast was secured at the point where it passed through the deck.

The presence of the circular hole seen on the Olbia masts between the tenon and the mast-base could be for a sort of key, although there is no comparable evidence for such a device for the Roman era. More probably, the hole permitted the passage of a rope used to lift the mast-base above the deck when lowered; the position shown, for instance, in a relief at Salerno Cathedral dated to the 3rd century AD (Casson L., 1965: pl. VI–I) (Fig. 6).

figure

Figure 6. The relief in the Salerno Cathedral. (Photograph F. Tiboni)

Download figure to PowerPoint

Iconographic evidence

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References

The Salerno relief has other elements which are reflected in the Olbia artefacts and other Roman mast finds. It shows a lowered mast placed above the deck of a Roman oneraria during unloading. The artist has represented the front-side of the mast; thanks to this, we can clearly see a tenon, carved in the centre of the foot and measuring 1/3 of the total width. This 1/3 ratio is also found on the Dramont E wreck mast (Santamaria 1984: 109, fig. 3) and the Olbia masts. Unfortunately, this representation does not allow us to discern the shape of the tenon, but does not exclude a semi-circular shape.

It is very interesting to note that triangular steps are placed on opposite sides of the Salerno mast. These steps, probably used to climb the mast to manoeuvre the sail, are represented on other Roman images dated to the first centuries AD, such as the ‘Mosaico delle Corporazioni’ of Ostia, a tombstone from Isola Sacra dated to the Trajan period, and the stele of the so-called ‘Navis Caudicaria found in Rome and dated to the 3rd–4th centuries AD (Casson 1965: 31–9). According to this iconographic evidence, we can argue that steps were not carved directly into the mast but rather were joined to it. In the Jerusalem ship (Bennet, 1974; Broshi, 1977), a drawing found in the Church of the Holy Sepulchre dated to the c.4th century AD, we cannot identify the shape of the steps.

The steps could have been joined to the mast individually (Riccardi et al., forthcoming) or, more probably, as a complete ladder. In fact, Domergue's discovery of ladders with triangular steps, carved from single lengths of wood in the mines of Mazzaron and Pedreras Viejas (Domergue, 1967: 44; 1990: 134, 428–9, pl. XVII-d) might suggest that similar ladders were joined to the mast using the rows of mortises. The dimensions of the ladders found in Spain, as well as those found on sites such as Ercolano or Pompeii (Adam, 1988: 217–21), support the idea that the mortises found on the Olbia masts are part of a similar system. The ladder found in Bottega 9, Insula Orientale IIA in Ercolano has a step-height of c.0.20 m (Adam, 1988: 220–1, figs 480–1) while the ladder found in the Casa del Fauno, Pompeii, has a step-height of c.0.26 m (Adam 1988, 220, fig. 479), the triangular steps of the Spanish wooden ladders are about 0.26–0.27 m in height. These measurements fit well with the 0.26–0.29 m space between mortises found on the Olbia masts, and permit us to suppose that the tenons used to fix the ladder to the mast would have been aligned with the steps. as the point of the maximum stress.

According to the Salerno relief, the mosaic from Ostia, the Navis Caudicaria from Rome and the tomb from Isola Sacra, Roman wooden mast-ladders appear to have alternating steps on each side of the mast, which might explain why the mortises seen on the Olbia artefacts are not aligned. In particular, the two images from Ostia and Isola Sacra seem to suggest that the steps on the port and starboard sides of the mast maintain a fixed distance between them, alternating on each side, as seen in the Olbia masts. Of course, we cannot be sure that the artists' primary aim was an accurate rendition; however we cannot deny that the carpenter who created the Olbia mast had decided to give his mortises a regular pattern, avoiding knots and guaranteeing strength.

Finally, it is very interesting to note how the Roman stele from Salerno and the Navis Caudicaria stele from Rome both present masts with a complex profile, suggestive of an octagonal shape. This particular shape, which is present in the basal section A1, was also found during the partial excavation of the mast of the Albenga 1 wreck, dated to the 1st century AD, even if it has as yet been neither completely excavated nor fully published (Gianfrotta and Pomey, 1980: 251). Furthermore, the presence of at least eight possibly irregular faces can be supposed for the mast of the Dramont E wreck, based on the drawings of the mast-step published by Santamaria (Santamaria 1984: 109, fig.3) which show that even if the foot is almost square, the corners are then worked to create more faces.

Conclusions

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References

The lack of comparative material, which can only be overcome with further research, prevents us from being more specific or solving some of the problems linked, for instance, to the retaining systems in use during the Roman period, the analysis of the archaeological and iconographic evidence presented above allows us to identify artefacts A1 and A2, found in a shipyard in the port of Olbia, as masts belonging to Roman vessels dated to the Classical era and abandoned, possibly when their upper ends were broken. Moreover, the suggestion that the two objects are rudder-stocks can be rejected through close analysis of their structure and dimensions, and comparison with available iconographic and archaeological evidence for both masts and rudder-stocks.

Furthermore, although the wood from which A1 and A2 are constructed (evergreen holm oak, Quercus Ilex) is native to Roman Sardinia, and despite the intriguing evidence of the matching shapes of the A1 mast-heel and the R5 mast-step, it cannot be conclusively argued that they were built by local shipwrights. However, they were found together in a shipyard dating to the Neronian/Vespasian period and might have been stocked as spare components or for reuse.

Note
  1. 1

    At ISBSA 2012 in Amsterdam the paper written by Riccardi, D'Oriano and Gavini was read by F. Tiboni. The reader agreed with the interpretation given by the authors and decided to write the present paper with them in order to specify the archaeological basis of the interpretation.

References

  1. Top of page
  2. Abstract
  3. A1 and A2 (Fig. )
  4. T1–4 (Fig. )
  5. Comparison of T1–4 and A1–2
  6. Structural analysis of A1
  7. Iconographic evidence
  8. Conclusions
  9. References
  • Adam, J. P., 1988, L'arte di costruire presso i romani. Milan.
  • Basch, L., 1987, Le muséee imaginaire de la marine antique. Athens.
  • Bennet, C. M., 1974, The Jerusalem ship. IJNA 3.2, 307309.
  • Broshi, M., 1977, The Jerusalem ship reconsidered. IJNA 6.4, 349356.
  • Casson, L., 1965, Harbour and River Boats of Ancient Rome. The Journal of Roman Studies 55.1–2, 3139.
  • Domergue, C., 1967, La mine antique de Diógenes (Province de Ciudad Real). Mélanges de la Casa de Velàzquez III, 2991.
  • Domergue, C., 1990, Les Mines de la péninsule Ibérique dans l'Antiquité romaine, 127. Rome.
  • D'Oriano, R., 2004, Relitti di Storia: lo scavo del porto di Olbia, in M. Giacobelli (ed.) Lezioni Fabio Faccenna: Conferenze di archeologia subacquea, 6374. Bari.
  • D'Oriano, R., Riccardi, E., and Gavini, E., 2002, I relitti del porto di Olbia. L'Africa Romana XIV, 12491262.
  • Gianfrotta, P. A. and Pomey, P., 1980, Archeologia Subacquea. Milan.
  • Mott, L. V., 1997, The Development of the Rudder. London.
  • Pomey, P., Kahanov, Y., and Rieth, E., 2012, Transition from Shell to Skeleton in Ancient Mediterranean Ship-Construction: analysis, problems, and future research. IJNA 41.2, 235314.
  • Riccardi, E., 2002, A ship's mast discovered during excavation of the Roman port at Olbia, Sardinia. IJNA, 31.2, 268269.
  • Riccardi, E., D'Oriano, R., and Gavini, V., forthcoming, Masts and rudders from the Port of Olbia, Proceedings of the XIII ISBSA, Amsterdam 2012: Ships and Maritime Landscapes. Amsterdam.
  • Santamaria, C., 1984, Le pied de mât de l'épave ‘E’ du cap Dramont (Saint-Raphaël, Var). Archeonautica 4, 107114.
  • Uccelli, G., 1940, Le Navi di Nemi. Rome.
  • Ximenes, S. and Moerman, M., 1998, Fouille de l'épave de la Calanque de l'Ane. Archeonautica 14, 299302.