ships of Lake Nemi, the
- Deborah N. Carlson
The Lake Nemi ships were two enormous, palatial houseboats built by the Roman emperor Caligula (r. 37–41 ce). Lake Nemi is a small volcanic crater lake just 1.8 km (1.1 miles) wide and 35 m (115 feet) deep, situated in the Alban Hills 30 km (18 miles) southeast of Rome. Attempts to recover the Nemi ships drew the attention of key historical figures across five centuries, until in 1928–1929 Benito Mussolini ordered the water pumped from the lake to expose the two wooden hulls, which were in a superb state of preservation following immersion in fresh water for almost two millennia. At a length of more than 70 m (230 feet), the Nemi ships remain the largest ancient ships discovered to date. The ships’ complete destruction by fire at the close of World War II constitutes one of the great tragedies of nautical archaeology.
Historical and Archaeological Context
Over the course of five centuries, Caesar, Tiberius, and Trajan have been variously credited with building the Nemi ships, but the attribution to Caligula was affirmed by the discovery of three sections of lead pipe stamped with “C • CAESARIS AUG GERMANIC” and found in association with the first ship. Written sources offer a number of reasons why Caligula is a “good fit” as mastermind of the Nemi ships. First, Caligula demonstrated his enthusiasm for big ships with the construction of a huge carrier to transport a 500-ton obelisk and its foundation from Egypt to Rome (Plin. HN 36.14.70)1; he is also said to have cruised the Campanian coast aboard gem-studded Liburnian galleys equipped with baths, banquet halls, and fruit trees (Suet. Calig. 37). Like many Romans of the period, Caligula was fascinated by Egypt, the cult of Isis, and Nilotic subjects; it has been suggested that the Nemi ships were a Caligulan version of Egypt’s opulent “chambered” river boat, the navis thalamegus.2 The thalamegus was a 300-foot-long (120-m) catamaran constructed of numerous exotic building materials by Ptolemy Philopator (r. 221–203 bce), described by Callixenus (as preserved in Athenaeus V.204d–206c), and the likely prototype of the vessel on which Caesar and Cleopatra sailed together (Suet. Iul. 52).
The lake on which the two Nemi ships were anchored was known to some Romans as the speculum Dianae (“mirror of Diana”) (Serv. ad Aen. 7.515), owing to the presence of the sanctuary of Diana Nemorensis, the goddess of the grove (nemus being the source of the name “Nemi”), at nearby Aricia. Epigraphic and archaeological evidence points to the worship of Isis, also a moon goddess by this time, at or near Aricia,3 emphasizing Lake Nemi’s allure to moon worshippers like Caligula, who is reported to have invited the goddess Luna into his bedroom during the full moon (Suet. Calig. 22). Archaeological exploration of the sanctuary in the 19th and early 20th centuries was imprecise and incomplete, with countless objects sold to private collectors and many sculptural pieces appearing in museums in Boston, Copenhagen, Nottingham, Palma (Majorca), Philadelphia, and Rome.
Recent excavations by Italian and Norwegian archaeologists have brought to light interesting evidence for a lavishly decorated villa on the western shore of Lake Nemi.4 The remains of this villa, which was erected in the mid-1st century bce, expanded in the 1st century ce, and abandoned in the mid-2nd century, lay between the sanctuary and the lake, creating a kind of tripartite complex.5 Excavation revealed a cryptoportico providing access to the villa from the lakeshore, as well as a number of antefixes and decorative glass panels nearly identical to those recovered from the Nemi ships; this has led to the conclusion that the most important of the villa’s four phases was Caligulan.6
Efforts to recover the Nemi ships began in the Renaissance, when the humanist Leon Battista Alberti recruited swimmers to help him haul ashore one ship, using iron hooks deployed from a raft; the plan succeeded only in ripping away portions of the wooden hull. In 1535, a desire to explore the sunken ships up close led to the invention of what has been called the first diving bell designed for archaeology.7 Three centuries later, in 1827, Annesio Fusconi deployed a version of the diving bell designed by Edmund Halley (in 1690) but capable of accommodating eight divers. Over a period of about two weeks, Fusconi’s men retrieved gilded copper nails, bricks, pieces of terracotta pipes and opus sectile pavements, a bronze column capital, and various timbers from the ship’s hull, many of which were subsequently carved into furniture, smoking pipes, and walking sticks, while numerous artifacts were sold to local nobility or to the Vatican Museum.8 Additional artifacts discovered in 1895 by Eliseo Borghi, who also located the second Nemi ship in slightly deeper water, revealed that the opulence extended to the ship’s hardware: decorative bronze fittings cast as apotropaic symbols (Medusa head and hand with forearm), wolf, lion, and panther protomes, each gripping a mooring ring in the mouth, and bronze ferrules in the form of a lion’s head which once capped the ends of the ship’s two steering oars. These and other artifacts from Lake Nemi are now in the Palazzo Massimo alle Terme Museum in Rome. In 1928, at the direction of prime minister Mussolini, four electric turbines began pumping water out of Lake Nemi via a Roman conduit designed to control the level of the ancient lake. The first ship, which lay closer to the shore, came to light in September 1929, and after continued pumping the second ship emerged from Lake Nemi eighteen months later.
Construction and Design
The recovery of the two hulls from Lake Nemi afforded archaeologists and naval historians of the 1930s an unprecedented opportunity to study specific aspects of ancient Roman ship construction.9 We now know from the excavation of Roman shipwrecks that, in many ways, the construction of the Nemi ships followed a fairly canonical process. First, a shell was formed by attaching pine hull planks to the keel, and then to one another, using pegged mortise-and-tenon joinery. Planks were connected end to end to form strakes (runs of planking) by means of diagonal scarf joints, which were nailed at the tips. Oak frames were installed on the interior and fastened to planking by means of copper nails driven through oak treenails (thick wooden dowels) and double-clenched (bent twice) over the internal face of each frame. The exterior of the hull was smeared with pitch, followed by a layer of woolen textile, and finally a 1-mm-thick skin of lead sheathing was affixed with nails driven in a quincunx pattern.
Other construction features distinguish these vessels from traditional seagoing Roman ships, but then the Nemi ships were not seagoing and were anything but typical. Their enormous size, at 71–73 m (233–240 feet) in length and 20–24 m (66–79 feet) in beam, required that the main keel be supported by four smaller sister keels, two on each side. The 10-cm-thick (4-inch) hull planking was unusually but necessarily robust, and the oak floor timbers (long frames that span the bottom of the vessel over the keels) were scarfed to futtocks (short frames extending through the turn of the bilge) to form continuous frames, as opposed to alternating floors and half-frames. Some of these frames feature peculiar chocked butt scarfs (where the butt ends of floor timber and futtock meet flush at the bottom but not at the top, resulting in a Y-shaped scarf that is filled in with an angular block or chock) at the turn of the bilge, which are characteristic of 18th-century warships.10 It has been suggested that the Nemi ships’ enormous size, consistently flat bottoms, and uniform curvature at the turn of the bilge can only have been achieved using either graduated moulds11 or the application of a system of measurement based on the pes monetalis.12 Indeed, Marco Bonino sees in the drawings of the Nemi hulls sufficient repetition of whole numbers and geometric patterns to suggest an early form of naval architecture.13 An obvious but fundamental obstacle is that current and future research on the Nemi ships must rely on architectural drawings instead of actual physical remains.
The first ship, 71 m (233 feet) in length, is sometimes described (inaccurately) as a sailing ship or merchantman, but it had no mast and it carried no cargo; it was in fact a veritable floating palace.14 Early efforts to salvage the first ship damaged or destroyed portions of the bow, but its graceful curving cutwater shape is preserved in three sections of a bronze casing that fit onto the stem. The first ship carried two wooden steering oars, each approximately 12 m (40 feet) long and originally mounted on either stern quarter. Various pieces of hydraulic equipment recovered from the first ship included lead pipes, a bronze faucet, and terracotta pipes (tubi fittili) both rectangular and circular in section. The latter were often found cemented together in pairs, giving rise to the suggestion that they functioned like suspensurae in a heated Roman bath.15 Surviving elements of the superstructure included tessellated pavements, opus sectile pavements, a marble doorsill, wooden window shutters, and roof tiles of terracotta and gilded copper. Two circular wooden turntables, the larger being almost 1m (3.3 ft) in diameter, equipped with spherical bronze or conical wooden rollers, were likely designed for showcasing heavy objects such as statuary. The domestic nature of the archaeological material indicates that the first ship was primarily a private residence that functioned as a continuation of the lakeside villa.16
The second Nemi ship, 73 m (240 ft) long, lay in deeper water, farther from the lake shore, making it less accessible to salvors and therefore better preserved.
This vessel was an oared galley, distinguished by an outrigger which is essentially a wooden framework attached to the hull that positions the fulcrum of the oar(s) beyond the gunwale (the upper edge of the ship’s side) to increase leverage. This outrigger, which was the only example to have survived from antiquity, was realized by extending the upper of two superimposed deck beams through the sides of the hull. Outside of the hull the ends of the beams were fastened to another longitudinal timber parallel to the gunwale. This junction was strengthened by long iron straps capping the beam ends and the underside of outrigger was planked. A portion of the outrigger discovered by Borghi in 1895 was believed to belong to the deck, but in fact no fragment of the second ship’s decking was ever identified. Three (of four) bronze fittings decorated with an outstretched male forearm were still affixed to the ends of the beams on the outrigger’s after face, revealing that the second ship was originally equipped with two steering oars on either side of the stern. Associated finds include a bronze railing with posts in the guise of herms, pieces of marble opus sectile pavements, gilded copper roof tiles, four fluted column shafts of breccia, and about fifty fragments of a sculpted terracotta frieze, as well as terracotta antefixes. The architectural evidence, particularly when weighed together with a partial sistrum found among the wreck material, suggests that the second Nemi ship was likely adorned with a temple and served a predominantly ceremonial function.
The ships’ two anchors represent some the largest and most interesting examples of ancient Roman anchor technology.17 One of these was a wooden anchor with lead stock, a type that was well established in the Mediterranean by the 1st century ce. The oak shank of this anchor measured a remarkable 5.5 m (18 feet) long and was topped by a ring of cordage (not metal) held fast by rope bindings that survived intact. The attached arms, also of oak, were capped with iron teeth. The associated lead stock weighed an astonishing two tons.
The second anchor was of iron with a removable stock held in place by a cotter pin—a design closely resembling British Admiralty anchors made in the mid-19th century. But the iron anchor from Nemi is doubly intriguing because its shank and lunate arms were clad in wood, a trick likely intended to increase the anchor’s surface area so as to prevent it from sinking too deep into the muddy lake bottom. When the wood was removed, it became clear that the anchor had been forged in three sections (shank and two arms), and the total weight—1,275 Roman pounds (419 kg or 924 modern pounds)—had been incised in Roman numerals onto one of the arms and again onto the stock.
Recovery and Destruction
While it is tempting to associate the sinking of the two Nemi ships with the unofficial damnatio that followed the murder of Caligula in 41 ce (Cass. Dio 60.4.5–6, Suet. Claud. 11.3), numismatic evidence from the ships suggests that they—like the lakeside imperial villa—continued in use at least through the reign of Nero (54–68 ce).18 It is intriguing nonetheless that no mention of the Nemi ships survives in ancient literature, even though there is plenty of scandalous commentary on Caligula’s extravagant lifestyle. Certainly the absence of most of the superstructure from both ships implies that despoliation began in antiquity before the actual sinking and continued in the form of sporadic scavenging for the better part of two millennia.19
When the second Nemi ship was hauled ashore in October 1932, scientists knew little about the conservation of waterlogged wood. To the credit of those involved, the enormous wooden hulls were covered with damp canvas to prevent desiccation and distortion and later coated with a plant-based tar mixture, on the advice of Norwegian researchers who had used a similar substance to treat several Viking ships. The ships were given modest cover, but by 1933 it became clear that exposure to the elements was causing their rapid deterioration, so local suppliers donated tons of cement, bricks, iron, and lumber to construct the Museo delle Navi Romane, which opened in 1936 on the north shore of Lake Nemi.
On the night of May 31, 1944, as Allied forces were pursuing a retreating German army in a final push to capture Rome, the Museo delle Navi Romane caught fire and virtually everything within it was destroyed. The only objects that survived were thousands of copper nails and those artifacts recovered by Borghi in 1895 and stored in the Terme Museum. An official report filed in Rome two months after the incident concluded that the fire was not the result of bombs dropped from Allied planes but was instead a deliberate act of vandalism by German soldiers stationed next to the museum.20 In 1999, the Associazione Dianae Lacus set out to build a full-sized replica of the first ship. In 2001, with funding from the Nemi town council, the ship’s reconstructed keel and stem were displayed next to the museum, but by 2016 the $10 million project seemed not to have advanced beyond this first stage.21
- Bellabarba, Sergio. “The Origins and Ancient Methods of Designing Hulls: A Hypothesis.” Mariner’s Mirror 82.3 (1996): 259–268.
- Bilde, Pia Guldager. “The Roman Villa by Lake Nemi: From Nature to Culture—Between Private and Public.” In Roman Villas Around the Urbs: Interaction with Landscape and Environment: Proceedings of a Conference held at the Swedish Institute in Rome, September 17–18, 2004. Edited by Barbro Santillo Frizell and Allan Klynne. Rome: Swedish Institute, 2005. Online.
- Bonino, Marco. “Notes on the Architecture of Some Roman Ships: Nemi and Fiumicino.” In TROPIS I: First International Symposium on Ship Construction in Antiquity (Piraeu 1985). Edited by Harry Tzalas, 37–53. Athens: Hellenic Institute for the Preservation of Nautical Tradition, 1989.
- Bonino, Marco. “Further Steps in the Study of the Nemi Ships: Architecture and Clues for Their Reconstruction.” In TROPIS VI: The Sixth International Symposium on Ship Construction in Antiquity (Lamia, 1996). Edited by Harry Tzalas, 99–113. Athens: Hellenic Institute for the Preservation of Nautical Tradition, 2001.
- Bonino, Marco. Un sogno ellenistico: Le navi di Nemi. Pisa: Felici, 2003.
- Carlson, Deborah. “Caligula’s Floating Palaces.” Archaeology 55.3 (2002): 26–31.
- Eliav, Joseph. “Guglielmo’s Secret: The Enigma of the First Diving Bell Used in Underwater Archaeology.” International Journal for the History of Engineering and Technology 85.1 (2015): 60–69.
- Ghini, Giuseppina. “Nemi: La villa di Caligola, le navi imperiali e il Santuario di Diana.” In Residenze Imperiali nel Lazio: Atti della Giornata di Studio Monte Porzio Catone, 3 Aprile 2004. Edited by Massimiliano Valenti, 31–42. Monte Porzio Catone: Cavour, 2008.
- Green, Carin M. C. Roman Religion and the Cult of Diana at Aricia. Cambridge, U.K.: Cambridge University Press, 2007.
- Höckmann, Olaf. “Zwammerdam und Nemi: Zur Bauplanung Römischer Schiffe.” Archäologisches Korrespondenzblatt 18 (1988): 389–396.
- Lanciani, Rodolfo. “The Mysterious Wreck of Nemi.” North American Review 162 (February 1896): 225–234.
- McManamon, John. Caligula’s Barges and the Renaissance Origins of Nautical Archaeology under Water. College Station: Texas A&M University Press, 2016.
- Moltensen, Mette. A Roman Villa by Lake Nemi: The Finds/The Nordic Excavations by Lake Nemi, loc. S. Maria (1998–2002). Rome: Quasar, 2010.
- Moretti, Giuseppe. Il Museo delle Navi Romane di Nemi. Rome: Libreria dello Stato, 1940.
- Palladino, Alessia. “Il valore ideologico delle navi di Nemi: I modelli ellenistici e la theosebeia di Caligola.” Documenta Albana 25 (2003): 27–43.
- Palladino, Alessia. “Il santuario e le navi di Nemi.” Horti Hesperidum fasc. II.1 (2012): 584–600.
- Pomey, Patrice. “Conception et réalisation des navires dans l'Antiquité Méditerranéenne.” In Concevoir et construire les navires de la triere au picoteux. Edited by Eric Rieth, 49–72. Ramonville-Saint-Agne: Eres, 1998.
- Speziale, Giuseppe C. “The Roman Anchors Found at Nemi.” Mariner’s Mirror 17 (1931): 300–320.
- Steffy, J. Richard. Wooden Ship Building and the Interpretation of Shipwrecks. College Station: Texas A&M University Press, 1994.
- Ucelli, Guido. Le navi di Nemi. 2d ed. Rome: Libreria dello Stato, 1950.
- Wirsching, Armin. “How the Obelisks Reached Rome: Evidence of Roman Double-ships.” International Journal of Nautical Archaeology 29.2 (2000): 273–283.
1. Armin Wirsching, “How the Obelisks Reached Rome: Evidence of Roman Double-ships,” International Journal of Nautical Archaeology 29.2 (2000): 274–276.
2. Alessia Palladino, “Il valore ideologico delle navi di Nemi: I modelli ellenistici e la theosebeia di Caligola,” Documenta Albana 25 (2003): 27–43.
3. Carin M. C. Green, Roman Religion and the Cult of Diana at Aricia (Cambridge, U.K.: Cambridge University Press, 2007), 57.
4. Mette Moltensen, A Roman Villa by Lake Nemi: The Finds/The Nordic Excavations by Lake Nemi, loc. S. Maria (1998–2002) (Rome: Quasar, 2010).
5. Pia Guldager Bilde, “The Roman Villa by Lake Nemi: From Nature to Culture—Between Private and Public,” in Roman Villas Around the Urbs: Interaction with Landscape and Environment: Proceedings of a Conference held at the Swedish Institute in Rome, September 17–18, 2004, edited by Barbro Santillo Frizell and Allan Klynne (Rome: Swedish Institute, 2005).
6. Giuseppina Ghini, “Nemi: La villa di Caligola, le navi imperiali e il Santuario di Diana,” in Residenze imperiali nel Lazio: Atti della Giornata di Studio Monte Porzio Catone, 3 Aprile 2004, edited by Massimiliano Valenti (Monte Porzio Catone: Cavour, 2008), 36.
7. Joseph Eliav, “Guglielmo’s Secret: The Enigma of the First Diving Bell Used in Underwater Archaeology,” International Journal for the History of Engineering and Technology 85.1 (2015): 61.
8. Rodolfo Lanciani, “The Mysterious Wreck of Nemi,” North American Review 162 (February 1896): 230.
9. The definitive reference is Guido Ucelli, Le navi di Nemi (2d ed.; Rome: Libreria dello Stato, 1950), 147–180.
10. J. Richard Steffy, Wooden Ship Building and the Interpretation of Shipwrecks (College Station: Texas A&M University Press, 1994), 72.
11. Sergio Bellabarba, “The Origins and Ancient Methods of Designing Hulls: A Hypothesis,” Mariner’s Mirror 82.3 (1996): 262–264.
12. Olaf Höckmann, “Zwammerdam und Nemi: zur Bauplanung Römischer Schiffe,” Archäologisches Korrespondenzblatt 18 (1988): 389–396.
13. Marco Bonino, “Notes on the Architecture of Some Roman Ships: Nemi and Fiumicino,” in TROPIS I: First International Symposium on Ship Construction in Antiquity (Piraeus 1985), edited by Harry Tzalas (Athens: Hellenic Institute for the Preservation of Nautical Tradition, 1989), 37–53; Marco Bonino, “Further Steps in the Study of the Nemi Ships: Architecture and Clues for Their Reconstruction,” in TROPIS VI: The Sixth International Symposium on Ship Construction in Antiquity (Lamia, 1996), edited by Harry Tzalas (Athens: Hellenic Institute for the Preservation of Nautical Tradition, 2001), 99–113; Patrice Pomey, “Conception et réalisation des navires dans l’Antiquité Méditerranéenne,” in Concevoir et construire les navires de la triere au picoteux, edited by Eric Rieth (Ramonville-Saint-Agne: Eres, 1998), 56–57.
14. According to Pliny (Ep. 8.20), it was considered sacrilegious to sail on the sacred waters of Lake Vadimon, and presumably the same was true for Lake Nemi.
15. Ucelli, Le Navi di Nemi, 160–161.
16. Marco Bonino, Un sogno ellenistico: Le navi di Nemi (Pisa: Felici, 2003), 117–153; Giuseppina Ghini, “Nemi: La villa di Caligola, le navi imperiali e il Santuario di Diana,” in Residenze imperiali nel Lazio: Atti della Giornata di Studio Monte Porzio Catone, 3 Aprile 2004, edited by Massimiliano Valenti (Monte Porzio Catone: Cavour, 2008), 40.
17. Giuseppe C. Speziale, “The Roman Anchors Found at Nemi,” Mariner’s Mirror 17 (1931): 300–320.
18. Ucelli, Le Navi di Nemi, 368.
19. Ibid., 295.
20. Ibid., 316–321.
21. Deborah Carlson, “Caligula’s Floating Palaces,” Archaeology 55.3 (2002): 26–31.