In 1981-82 construction work in the German city of Mainz revealed the badly damaged remains of five Roman river craft from the imperial era which were probably in a boat scrap yard at the time of their deposition. Four of these hulls are long (approx. 21m) and narrow. While the ships appear to have had sailing equipment, they seemed to have been primarily operated by oar power, with room for thirty rowers (fifteen per side) per vessel. One of the ships preserves a rack that has been hypothesized as a mounting place for shields, presumably to defend the rowers. The sleek lines and lack of cargo capacity almost certainly identify these finds as warships, naves lusoriae. The navis lusoria was a fast, light ship used by the Romans for patrolling the Danube and Rhine river frontiers. Its shallow draught was ideal for use in rivers, where the depth is highly variable and where the ship may need to be landed on shore without the aid of port facilities. The navis lusoria appears beyond archaeological material, too. Ammianus Marcellinus makes reference to such ships being used during a Rhine campaign near Mainz by Julian in the middle of the fourth century. In this incident the boats are used to make a night landing of soldiers and poles for the creation of a palisade, suggesting the flexible nature of these boats, as well as the confidence that the Romans had in them as indicated by their willingness to use them at night. Legal evidence suggests that the Romans employed the navis lusoria in very large numbers, with one fifth century text mentioning some 225 deployed ships along one 600 km stretch of the lower Danube.
By the early 1990s serious technical work necessary for the reconstruction of a navis lusoria was being carried out by Peter Marsden using computer modeling to predict performance and to limit costly design errors prior to construction. Early estimates predicted that the maximum efficient speed of the boat was 6-7 knots (11-13 km/h), although perhaps for short bursts the boat could reach 10 knots (18.50 km/h). In its empty state the draught was estimated to be about 30 cm, and once filled about 45, helping to confirm the boat’s viability as a river warship.
Dr. Olaf Höckmann from the University of Regensburg helped put forward the idea of reconstructing one of the boats, building upon his work on the Mainz vessels. To help fund the project, the university had to partially turn to private industry since the initial estimate was ca. €100,000, and the chief donor received their company’s logo on the ship’s sail. For technical expertise, the metallurgists from Clausthal University and wood experts from FH Rosenheim were brought in. Large local oak was used for the hull, following the Mainz ships. Rather than just an object of academic curiosity, the ship was made available to the public. Launched in 2004, the ship was rowed on the Danube in 2005 and 2006 and made appearances at several public festivals, with rides on the ship being offered.
The actual reconstruction measures 21.7m long, 2.8m wide, and 0.96m high and was named Regina. Despite the vessel’s early role as a public spectacle, some time was found to run tests. A May 2005 voyage from Regensburg to Weltenburg revealed that upriver the ship was able to average about 4 km/h for several hours, suggesting that traveling 25 km upriver in a day was realistic. In June 2005 the boat was tested by the University of Hamburg, with students serving as the rowers after receiving some training. Early sprinting tests showed that the maximum speed of the boat began to decline precipitously after about four minutes of hard rowing, with the speed dropping from about 6 km/h to 4 km/h but this was improved upon later. Future tests pushed the boat’s speed up over 11 km/h, but this still fell substantially short of the 18.5 km/h estimated prior to construction. However, this was performed by a hastily trained group of university students of mixed fitness and not experienced Roman sailors, and so presumably the modern test has resulted in underperformance of the boat’s capabilities in this regard. Similar conclusions were borne out in the testing of the boat’s maneuverability. Results were highly variable, with 180° turns taking anywhere from one to two minutes, and full 360° turns taking from a little over a minute and a half to two and a half minutes. This variability was likely a result of the untrained crew, hinting that Rome performance could be significantly better.