The Fleets and Roman Border Policy


See this text in

Through the Battle of Actium in 31 B.C. the destiny of the Roman Empire was decisively influenced by a sea-battle for the final time. Following it Rome dominated the seas; enemies with their own large fleets no longer existed. Despite this, the Roman navy was not disbanded; indeed, it was expanded further. This clearly demonstrates that the fleets did not play an inconsiderable role in Roman imperial policy. As a result, one has to ask oneself what functions the fleets performed.

The role of the fleets during campaigns

The best literary sources are to be found in Dio Cassius, Velleius Paterculus and Tacitus, who describe four expeditions of Roman fleets along the North Sea coast. This occurred in connection with the Roman attempt to conquer Germania upto the River Elbe. In 12 B.C. the Romans under the leadership of Drusus risked the first attempt; the second expedition followed in 5 B.C. under Tiberius. Germanicus undertook the two final voyages in A.D. 15 and 16, both ending in catastrophes, when tidal waves destroyed the ships. The reason for all four operations was always the same: a part of the army of occupation was brought by ship as close as possible to the theatre of war. Together with the soldiers equipment and provisions were also loaded on the ships newly constructed for this purpose. In this Tacitus (Annals 2, 6) describes the ships of Germanicus. They had wide hulls, but narrow sterns and bows. The keel was flat, so that the ships did not run aground, even by low-water. All could be sailed and rowed. Some possessed a cover to protect the cargo from wind and weather. A series of these characteristics survived upto the 4th century, as displayed by the five Roman military ships from Mainz (Mainz 1, 2, 3, 4, 5).

In later times, too, troops and their equipment were transported by ship to their place of action. An impressive witness for this is the 40 m high Trajanís Column in Rome. The whole of the exterior is decorated by a continuous band of reliefs on which the events of the First and Second Dacian wars (A.D. 101/102 and 105/106) are depicted. Scenes 33-35 show the emperor and his soldiers marching out of winter-quarters and his journey to the theatre of war in 102. One can clearly distinguish different types of ships. Firstly, there are small cargo vessels with high bulwarks and wide bows, allowing storage space for the folded tents. They are steered by two lateral steering-oars. One can recognise these in particular on the ship shown during a journey on the Danube. The helmsman sits on an elevated position, so that he can look over the heads of the rowers. These ships had only one row of oarsmen on each side; the oars rested on top of the bulwarks.

The horse transporter, however, displays other characteristics. In order to obtain room for the animals, there is no steering-oar here. The small team of rowers is responsible for steering. A conspicuous aspect of this vessel lacking in the other cargo ships is the inclined bottom at the front of the ship. The use of such inclined ends was that the ships could be manoeuvred onto a flat bank. This allowed them to be loaded from the front, which was of obvious advantage when having to load a ship with animals.

Such flat ends are especially accentuated in the ships of Type Zwammerdam. These ships have often been recorded north of the Alps and could measure upto 40 m long, as documented by the ships Zwammerdam 4 and Mainz 6. One needed a large crew to be able to navigate such long ships on a river. In Roman times only the army could provide enough personnel, so that most ships of Type Zwammerdam can be regarded as military vessels. The find-spots seem to confirm the close connection with the army; in the majority they are forts or towns in the proximity of which Roman military forces were stationed.

If this is a first indication that the ships depicted on Trajanís Column are not requisitioned civilian vessels, the final proof is provided by the ships on which the emperor and his staff are travelling. On the one hand they display the inswinging prow typical of warships, on the other hand two rows of oarsmen are positioned on each side. The disposition of the oars demonstrates that these ships had an outrigger, so that the two rows of oars dipped into the water at varying angles and, therefore, did not hinder one another.

In doing so the rowers sitting on the outside placed their oars through the rail on the outrigger, the oars of the rowers sitting on the inside lay on the bulwarks below the outrigger. The bireme warships were called liburnae and represent the typical ships of the provincial fleets.

The scenes on Trajanís Column disclose, therefore, that fleets also took part in the Dacian wars. Thanks to the inscription of C. Manlius Felix we can assume with great probability that there were two fleets in operation, the Pannonian and the German. In the inscription Manlius Felix is described as "Praefectus classium Pannonicae et Germanicae" (Admiral of the Pannonian and German Fleets), so one can safely assume that he commanded both simultaneously. The concentration of fleets stationed far apart was usual during campaigns in later times, too. The inscription of Valerius Maximianus mentions his special responsibility during the Marcommanic wars of the Emperor Marcus Aurelius (168-175). He had to secure the food supply for both the Pannonian armies. For this purpose he had under his command sections of the Misenum, Ravenna and British fleets. It is pertinent that the Pannonian fleet, i.e. the regular provincial fleet in the theatre of war, was not part of this assignment. Apparently, it was needed for other operations. These could have involved the transport of troops and equipment, as well as patrolling the Danube. After all, the army itself was involved in the war, i.e. without the fleet the Danube as the Empireís border would have been defenceless against enemy attacks.

The surveillance of river borders

In a series of Roman provinces a river forms the border of the Empire. In Europe this is the case for the provinces on the Rhine and the Danube. We know of three provincial fleets here, the classis Germanica, the classis Pannonica and the classis Moesica. Although one could conclude from their names that their areas of operations stretched over both provinces bearing their corresponding name, this was not the case. Thus, so-called military diplomas clearly show for the classis Pannonica and the classis Moesica that they belonged to the army of Pannonia inferior and Moesia inferior respectively.

As for Germania inferior one can see the same from the distribution of inscriptions and tiles, which are limited to Lower Germany. The sole exception is the Brohl Valley, which lay in Upper Germany, but the stone-quarries there were exclusively used by the army of Lower Germany. That the classis Germanica belonged to the army of Lower Germany was recently confirmed by a military diploma dating to Trajan. It was found in The Netherlands and names beside the auxiliary troops of the Lower German army the fleet as well. For the protection of the Empireís borders there was certainly a chain of forts along the corresponding rivers in all three provinces, but an effective control of the border was only possible with ships undertaking daily patrols, as the author Ammianus Marcellinus still describes for the 4th century. But Lower Germany, Lower Pannonia and Lower Moesia were not the only provinces with a riverine border with the non-Roman world. Since the fleets did not operate beyond their provinces, however, one must ask who was in charge of surveillance here.

A special feature of the province Moesia inferior provides a first indication for this. Although the province had its own fleet, the classis Moesica, there have been found in the legionary fortress of Novae tile-stamps bearing the name of the locally stationed legio I Italica in a frame shaped like a ship. Moreover, this is not any ordinary ship, but a warship displaying the typical concave prow and a sternpost decoration, such as seen on liburnae on Trajanís Column. This extremely seldom way of stamping tiles gives every reason to believe that the legio I Italica was equipped with its own squadron of warships. The tiles clearly belong to the 2nd century. This means, however, that in Moesia inferior alongside the fleet itself the naval squadron of a legion was also stationed. The reason for this lies in the area of operations of the classis Moesica. Its inscriptions and tiles are to be found only on the western and northern coasts of the Black Sea and in the Danube delta. They are absent further up river. One can assume, therefore, that the classis Moesica possessed mainly seaworthy ships with a deep draught, which were unsuitable for use on the rivers. As a result, the legio I Italica took charge of surveillance on this section of the Moesian border using more suitable ships.

But the finds in Moesia inferior are not the only pieces of evidence for naval squadrons in the legions. Confirmation is provided by information in the Notitia Dignitatum, a late Roman administrative handbook. Among other things it lists the legionary units and their bases under the command of the "Dux Pannoniae I et Norici". Some of these units consisted of liburnarii from distinct legions. At first, the word liburna meant a very particular type of small warship, but in the course of time it became synonymous with a warship in general. As a result, in the late Roman period the crews of warships are called liburnarii. According to the Notitia Dignitatum four legions were provided with such liburnarii: the legio II, legio X, legio XIV and legio I Noricorum. Except for the legio I Noricorum, which was first recruited at the end of the 3rd century, the other units are old ones, whose bases in the 2nd and 3rd century are known. Drawing conclusions from the information in the Notitia Dignitatum as to the conditions in earlier times, we have the following picture: In Noricum the Danube was controlled by ships of the legio II Italica, which since 172/173 lay in Lauriacum. At the end of the 3rd century it receivedhelp from a squadron of the legio I Noricorum in Favianis. For Pannonia superior this task was carried out by the 10th legion in Vindobona and the 14th legion in Carnuntum. In both provinces the Danube formed the border of the Empire, but, apparently, the naval squadrons of the legions were sufficient for surveilling the border, perhaps, because the border at this point was considered by the Romans as less threatened.

There is also evidence of military squadrons from legions for provinces in which a river did not form the Empireís border. For instance, in Niö one finds the funerary inscription of L. Cassius Candidus, a soldier of the legio VII Claudia in Viminacium in Moesia superior. Before his death he was discens epibetarum. During the Roman Empire epibeta is a very rarely used term for a member of the fleet. It particularly describes the marines, as opposed to military personnel endowed with nautical tasks. The inscription is generally dated to the late 2nd or 3rd century, thus belonging to a period in which Moesia superior was no longer a border province. Since, however, a river cannot be blockaded in the same manner as a land border and the Danube in particular possessed supraregional importance, the surveillance of the Danube remained a military operation even after the setting up of the province of Dacia.

A similar situation is known on the Rhine in Upper Germany. The anchor with a mark of the 16th legion, which during the Claudian period was stationed in Mainz, and the gravestone of a shipbuilder from the 22nd legion from the late 1st or early 2nd century belong to a period when rivers here, too, firstly the Rhine, then the Neckar, formed the imperial border. The same cannot be said of the two soldiers of the 22nd legion, who as optiones navaliorum superintended the legionís own wharves. Both belong to the late 2nd century, when the forward limes had already been constructed.
The assumption that the 22nd legion only possessed cargo vessels during this period is negated by one of their tile-stamps from the 2nd or 3rd century. Beneath the legionís name one can clearly recognise a warship with concave prow and many rowers. As in the case of Moesia superior, we must assume for Germania superior, too, that the Romans employed military ships on the large rivers, irrespective of the border situation, without us today knowing their special tasks.

The naval squadron of the 22nd legion became again extremely necessary in the period following 260, after the area east of the Rhine had been abandoned. This made the Rhine the Empireís border again, the west bank of which was well fortified with a chain of forts. At the confluence of important tributaries on the eastern bank there were small outposts, the so-called burgi, which could only be reached by ship. The enclosure walls were built out into the rivers and so formed a small harbour. The ships Mainz 1-5 from the late 3rd and 4th century belong to this historical context. Whereas the ships 1 und 2 sowie 4 und 5 belong to a very narrow, rapid type, ship Schiff 3 is a broad, short vessel. Both types were reproduced in accordance with the original and are exhibited in the Museum für Antike Schiffahrt in Mainz.

Reconstruction Mainz A is based upon the scientific analysis of ships Mainz 1 und 5. Accordingly, it produced in the reconstruction a roughly 21 m long, very narrow ship with a length-to-width-ratio of 8:1 which could be both rowed as well as sailed. Thanks to the favourable form these ships could reach high speeds. The keel is flat and hardly stands out from the bottom. The 35 man crew consisted of 32 rowers - 16 on each side one behind the other - two men for the sails and a helmsman. The distance between the thole-pins is 95 cm, a measure conforming to the classical tradition, according to which oarsmen should sit at such a distance, in order not to impede one another.

The remains of the prow surviving from Ship 5 suggests that the ships of this type possessed the inswinging prow of a warship. The steering apparatus is formed by a beam running transverse through the ship serving as a support for both the rudders. Because of their high speed, ships of this type served most probably as troop-ships. Possibly, the crew of such a ship also served as the guard in the small burgi on the east bank of the Rhine.

Reconstruction Mainz B embodies a broader and shorter type of ship. Its reconstruction is based upon Ship Mainz 3. The fore, which had not been preserved, was restored according to the model of a ship from the Treasure of Rethel in France. This vessel, too, could be rowed and sailed. The narrow stern and the ratio of length to width of 5:1 show it to have been a military ship.

Unlike reconstruction Mainz A, reconstruction Mainz B possesses a covered deck in the stern. The midships with an outrigger on both sides offered room for only 14 oarsmen, 7 on each side. Thus, it was much slower as a troop-ship. Most probably it was used as a patrol vessel, where speed was not so important as regular appearance, as is documented by a report of Ammianus Marcellinus. Although the literary sources are silent about how the surveillance of the border in Upper Germany was organised in the second half of the 3rd and the 4th century, the ships themselves offer conclusive evidence: it was the task of the legion at Mainz for which it was equipped with its own naval squadron.

The province Raetia on the Danube also possessed a riverine border. Here one can see that not only legions must have had their own naval squadrons, but auxiliary units, too. In 1994 members of the Museum of Ancient Shipping in Mainz were able to excavate two Roman ships (Oberstimm 1 and 2) at the auxiliary fort of Oberstimm. In their shape they resemble the rapid troop-ship from Mainz, but they are constructed completely in a Mediterranean technique. The planks display mortise-and-tenon joints, for building materials one used fir for the planking and oak only for the keel and frames. As with the ships from Mainz, they could be rowed and sailed. However, they were a little shorter and intended for a crew of only 21 or 19 men. Particularly through the narrow stern and the ratio of length to width of 6:1 they can definitely be identified as military vessels.

According to dendrochronological analysis the ships were built in the late Domitianic or Trajanic period. By A.D. 118 at the latest they had been decommissioned and used to strengthen the river-bank, as piles of a revetment from this year had pierced both ships. These, too, had been used for surveilling the river, as the construction of the Raetian limes had not been completed by the Trajanic period. Against this backdrop a patrol of the river seems very sensible. Since in this period no legion was stationed in Raetia, the ships from Oberstimm must have belonged to an auxiliary unit.

In summary, one can therefore say that fleets and naval squadrons were eminently important, both during campaigns for securing supplies and for troop transport, as well as for surveillance of large rivers, whether they formed the Empireís border or whether they represented merely regional traffic routes. As a result, even after the Battle of Actium, the Roman emperors could not dispense with their navy.


Barbara Pferdehirt

Translation by Clive Bridger, Xanten

Back to the homepage of NAVIS