Return to Nemi

Guido Ucelli and the Recovery of the Nemi Ships

Le Navi di Nemi was published by the Italian State Library in 1940, the same year that Italy entered World War II and the eighteenth since Mussolini's accession as prime minister in October 1922, a date that is commemorated on the book's title page: Anno XVIII E. F. [Era Fascista]. Written by Guido Ucelli, an engineer and manager of the company that produced the electric pumps used to drain the lake, his report on the excavation of the Nemi ships (with the acknowledged help of experts) is the most important primary source for their recovery.

It is a heavy book, printed on thick coated stock and weighing more than five pounds. The first edition was limited to just 500 numbered copies. A second numbered edition of 1500 copies was published in 1950 (curiously, on lighter-weight paper), which was reprinted in 1983 in an edition of 2000 copies. Scarce even at the time and now more than eight decades old, Le Navi di Nemi likely is to be found only in the stacks of a large university library. Some of its illustrations are presented here, most of which are from the second editionthe reprint of which was referenced by McManamon, who has written an important contemporary account. Photographs showing the construction of the Museo delle Navi Romane are from the first edition.

In July 1944, just seven weeks after retreating Germans had destroyed the museum, Ucelli and his wife were arrested for sheltering Jews and helping others to escape to Switzerland. In a prisoner exchange, he was released after three weeks and his wife, who was at risk for deportation, two months later. Again free, they resumed their humanitarian activity. In recognition of his contribution to the recovery of the Nemi ships, the honorary name "Ucelli de Nemi" was bestowed upon him.


Fig. 132

In this topographic drawing, the two ships are shown just offshore, sunk in water ten to twenty meters deep. The ancient emissary below Genzano that had maintained the level of the lake was connected to a pumping station with pipes extending to a floating platform just offshore. On the sloping plain at the northern edge of the lake, where the Romans likely had constructed the ships nineteen hundred years earlier, a museum was built to display them. Between it and the town of Nemi are the ruins of the Sanctuary of Diana.


Fig. 50

The construction firm Riva Milano provided the four massive electric pumps that were used to lower the water level of the lake by about twenty-two meters, draining off at least forty million cubic meters of water and consuming two million kilowatt hours of electricity. Mussolini himself (with Ucelli at his side) switched them on in October 1928, the month that Il Duce had assumed power. Over the next four years, it repeatedly would be the occasion for one phase or another of the project being completed. 


Fig. 61

In March 1929, after the water level had been lowered by about five-and-a-half meters, the stern of the first ship (prima nave) appeared. But the drop had caused a section of the lake bank to subside and two of the pipes feeding the pumping station collapsed in a jumble.

Fig. 64

Two of the pumps were moved to an auxiliary station, with sections of flexible pipes feeding those on shore. Floating on the surface, the raft readily adjusted itself to the dropping water level. After several months, the pumps again were turned on and in June a beam cap, a cast-bronze wolf's head, was discovered in the mud.

Fig. 69

The massive hull still is partially submerged on the shore of the lake.

Fig. 74

By September almost the entire ship had been exposed, the bow, recently underwater, buried under a protective layer of mud.

Fig. 77

Spectators were invited to visit the site and inspect the ship from an elevated walkway. As the water continued to recede, work continued on removing tons of mud from beneath the ship, sieving it to recover and catalog all that was found, including more bronze animal heads.


Fig. 84

With the receding water level, the sloping lakebed again began to sink, this time shifting the ship. Here, the bow is being elevated back to a horizontal position.

Fig. 83

With the hull now supported by a framework of braces, earlier damage to the stern, which was in shallower water, is apparent.


Fig. 85

By February 1930, the water level had dropped by almost fourteen meters and the second, larger ship (seconda nave) was revealed, exposing a section of deck railing decorated with a double herm. But then a violent storm sank the auxiliary pumping station and, until it could be recovered and rebuilt, work was suspended and attention instead directed to moving the prima nave. By that summer, a wooden hanger had been covered in canvas and set around the ship.

Fig. 81

In October, the cradled hull was winched along iron rails up from the lakebed to its temporary shelter on shore, where the ship would remain until it could be moved to a more permanent location.


Fig. 91

New, more powerful pumps were installed and by June 1931, the entire hull of the seconda nave had been exposed. But, under the hot summer sun, the wood was at risk of drying too quickly. To prevent warping or cracking, it was saturated in a solution of pine tar, linseed oil, and turpentine—the evaporating solvent leaving behind a protective sealant.


Fig. 187

This is one of the two bronze hands capping a longitudinal timber that supported the housing for the side (or quarter rudder in a Mediterranean context). The first had been found by Borghi in 1895. There also were gilded roof tiles and fluted columns.


Fig. 96

Workmen filling push carts excavate more mud from beneath the hull of the seconda nave which, because it was in deeper water and covered by muddy silt, had been spared the depredations of salvage and so survived more intact.


Fig. 98

In August 1931, there was further subsidence. To prevent any damage, pumping again was suspended and the water allowed slowly to rise, giving the impression that the ship was floating on the lake. One can see some of the surviving through-beams that projected beyond the hull. Extending on either side over the water, this large projecting frame (apostis) supported an outboard deck for rowers. By mounting the tholes or oarlocks outside the hull, there was a more efficient platform from which to pull the oars and greater leverage for longer ones. The additional space also allow rowers to sit in banks on each side of the vessel. The smaller but more palatial prima nave did not have apostes and either was towed about the lake or maneuvered by means of its sweep oars.


Fig. 100

In March 1932, the pumps were turned back on and, with the mud removed, the seconda nave was fitted with its own wooden cradle preparatory to being moved.


Fig. 149

By October, both ships were out of the water, the prima nave still sheltered beneath its canvas hanger and the seconda nave now covered in planking. Work on the museum that was to be their permanent home would began the next year.


Fig. 102

Eventually, the ship was winched over a wooden platform some five hundred meters to its permanent home ashore.


Fig. 269

The anchors were found at some distance from their respective ships, either because a sudden storm may have snapped the cables or they had embedded themselves too deeply in the muddy lake bottom to be raised without snapping the cable. The anchor from the prima nave had a removable stock (so it could be separated and stored flat) that was held in place by a cotter pin, anticipating a design (the Admiralty Anchor) that would not be reinvented until 1852. The intention is that the stock (the crosspiece that runs perpendicular to the shank) rotate the anchor so that an arm is forced into the bottom. Forged from three iron bars (the stock, shank, and arms), the anchor was sheathed in wood, with only the pointed bills at the ends of the flukes and the shank exposed. Deep grooves show where ropes and metal bands once had secured the wood exterior. It is an unusual construction and may been intended to allow a greater surface area to gain purchase but not too deeply. Or the wooden casing may have protected the iron core or the ship itself. Possibly, it simply imitated the appearance of a more traditional wooden anchor. The weight, which is stamped on the stock, was 1,275 Roman pounds libra (about 925 lbs., the abbreviation for pound having derived from its Roman antecedent).


Fig. 275

The anchor from the seconda nave was more typical of the time, when the boats were sunk about AD 40. The heavy oak arms are tipped with iron and secured to the eighteen-foot shank with stout pegs. So that the wooden anchor would sink, the fixed stock (the crosspiece that cants one of the arms to assure that its fluke will dig in) was cast in lead, with rope bindings securing the anchor ring. Remarkably, these lashings and the thick anchor cable were intact when recovered from the lake.

The anchor itself was said to have been invented by Eupalamus (Pliny, Natural History, VII.lvi.209) and the anchor with two arms or "teeth" (bidentem) by Anacharsis, who was one of the seven wise men of Greece (Strabo, Geography, VII.3.9).


Fig. 108

There always had been a plan to display the two recovered ships, but funding did not permit construction to begin until 1933.


Fig. 109

By October 1935, the building had been completed, except for a section in the front, which was left open so that the two ships could be trundled inside. It had been seven years since Mussolini had turned on the pumps that began to drain the lake. The prima nave was moved into its gallery in November, the seconda nave, which had farther to travel, in January 1936.


Fig. 113

Work continued on museum even after the ships had been moved inside.


Fig. 108

The newly inaugurated Museo delle Navi Romane ("The Museum of Roman Ships") opened on April 21 (the traditional day of Rome's founding), 1939just four months before Hitler's invasion of Poland and the beginning of World War II.


Fig. 111

Here, the elegantly curved bow of the prima nave has been reconstructed.


Fig. 112

One of the ship's two massive quarter rudders has been placed into position, together with its supporting beams and bronze protomes. Another oar almost twelve meters long, which was discovered under the hull on the starboard side, lies unobtrusively in front. Capped with a gilded bronze lion's head, it caused a sensation when discovered—as had its companion in 1895.


Fig. 114

The seconda nave on display in the west gallery.


Fig. 115

The outrigger beams supporting the rudders on each side of the seconda nave were capped with bronze forearms, the up-turned thumb and fingers that pointed away from the ship serving an apotropaic function in averting or warding off evil. Three such bronzes were found, which suggests that they were two such pairs of rudders, mounted at both ends of the hull. These four quarter rudders permitted the huge ship to reverse course without having to turn around.


Fig. 116a

Two spiral staircases allowed visitors to view the ships from an elevated gallery or visit a terrace on the roof of the museum.


Fig. 116b

The original bronze protomes were displayed in the upper gallery. During the war, they were transferred to the National Roman Museum for safekeeping, where they now are in the Palazzo Massimo alle Terme.


Fig. 320

The eastern side of the museum, its wall pockmarked by shrapnel.


Fig. 321

The ashes of the prima nave in the east wing, with only the metal framework that supported the wooden hull remaining.


Fig. 322

The ashes of the seconda nave in the west wing, after the museum was destroyed by fire on the night of May 31, 1944.

A commission at the time attributed the destruction to the retreating Germans, who earlier had evicted the Italian custodians and their families and established a gun battery close to the museum and then, once operative, fired so many rounds that American soldiers were able to determine the emplacement and unleash their own artillery barrage. Almost two hours after shelling had ended, fire broke out simultaneously in both exhibition halls, completely destroying the wooden ships, which were made more flammable still by the preservatives that saturated the wood. Only those movable artifacts that had been transferred to Rome nine months earlier were not lost.

The catastrophe was briefly reported by The New York Times on June 10, 1944 under the headline "Nazis Burn Galleys of Ancient Romans." And that was the conclusion of the official commission established to determine its cause: the German garrison, who were the only ones who had access to the ships, deliberately set the fires. It abandoned the position on June 2 and, when the Americans arrived several days later, there was nothing that could be done.

A recent book, however, has suggested that the official report was a cover-up and that the artillery shelling by U.S. forces that had forced the Germans to retreat actually caused the fire—which was the belated effect of shells having pierced the roof, dropping hot shrapnel down on the treated wood below, which smoldered until eventually erupting into flames. But the extinguishers, hydrants, and hoses that could have contained the blaze were never utilized by the Germans. Rather, the fire seems to have been a wanton act of arson intended to eradicate a part of Italy's cultural heritage, piqued perhaps by the armistice that the country had signed less than nine months earlier.

It had been more than a decade since the pumps first were turned on and water began to drain from Lake Nemi, and only five years that the recovered ships were displayed in their purpose-built museum before, in a single night, all that time and treasure was turned to ashes.

In 2020, the town of Nemi announced that it was seeking compensation from Germany for the willful destruction of the two ships.

References: Le Navi di Nemi (1940, 2nd ed., 1950) by Guido Ucelli; From Caligula to the Nazis: The Nemi Ships in Diana's Sanctuary (2023) by John M. McManamon. A series of photographs and short film clips also can be viewed at Archivio Luce (https.//www.

 Return to Top of Page