Roman Imperial fleets I



Drawing by Graham Sumner What you see is the Roman naval base at Velsen, just west of Amsterdam, which was in use during the reign of the emperor Tiberius. It is almost certainly identical to the fort named Flevum mentioned by Tacitus.

At the conclusion of the civil wars in 30 BC, Octavian had acquired fleets of his own and of his former enemies, totalling nearly 700 warships of all types, far more than could be afforded or, in the total absence of any opposition, would be needed. With the start of the Imperial period the whole concept of fleet organisation changed in parallel with the change in its role for the future. Having become a Roman lake, the Mediterranean and, increasingly, the Black Sea, had to be consolidated and policed. To do this, Octavian (now Augustus) and Agrippa established permanent, separate fleets, each with its own identity, commander, headquarters home base and defined area of responsibility. The system was to be formed around two main classes or fleets, based in Italy, with subsidiary fleets at strategic points around the Empire.

The first was the Classis Misenensis, based at Misenum at the northern tip of the Bay of Naples. Established by 22 BC, this was to be and remain the senior fleet of the navy and was ranked as praetorian, i.e. part of the emperor’s personal guard. The fleet’s area of operations was the entire western Mediterranean basin, but it could also (and did) project its power into the Atlantic and established a subsidiary squadron on the Mauretanian (Algerian) coast. This fleet was maintained at a strength in ships and men much greater than was strictly needed to perform its duties. As the senior fleet of the empire, it covered the western Italian coast and transported emperors, members of the imperial family and other notables; it also, importantly, acted as a training centre and a reserve of trained personnel for all branches of the service. These men could be and were sent to supplement other forces throughout the empire and even to provide the manpower for the foundation of other fleets. As an example, men from the Misene fleet were sent to establish the Classis Britannica in AD 43.

The Misene fleet remained the principal and strongest of the empire’s fleets almost to the end of the Western Empire. It was close enough to be able to be directed from Rome, as relay riders could deliver despatches between Rome and Misenum in one day. It was ideally positioned to be able to project its power across any part of the western Mediterranean basin, as well as to screen the ports of western Italy and the ends of the trade routes to the capital itself. The fleet had local facilities at various other ports and naval stations, for example at Cagliari (Carales), Civitavecchia (Centumcellae) and Aleria in Corsica. Ships were sent further afield from time to time, either for a particular mission or as a temporary detachment, inscriptions of members of the fleet having been found, for example, in Syria and Piraeus. There was a permanent detachment at Ostia and Portus, (when built); another was stationed at Rome, initially quartered in the praetorian barracks, but from the Flavians (later first century AD) until at least the mid-third century AD in their own permanent barracks in the city. There, one of their duties was to attend to the awnings that gave shade to the Colosseum.

The size of this fleet, and indeed that of all of the fleets, is unknown. The names of many ships of the fleet appear on grave stelae and votive altars and Nero was able to enrol a legion (approximately 4,500 men) from among the marines of this fleet in AD 68, later named I Adiutrix by his successor, Galba. At this time, it has been estimated, the fleet had over 10,000 sailors (whether this included marines is not said); at an average of 200 men to crew a trireme, this would indicate a fleet of about fifty ships. This is, of course, a very crude way of estimating numbers when considering a period of several centuries and a variety of ship types, each with differing crew numbers. Nevertheless, the base itself was the size of a town and the fleet establishment was many thousands of men deploying dozens and dozens of ships for most of its existence.

The second of the Italian fleets was the Classis Ravennate, established in about 23 BC at a new base built a short way south of the city of Ravenna, at the upper end of the Adriatic Sea. Slightly smaller than the Misene fleet, it was also rated as praetorian and had as its area of responsibility the Adriatic and Ionian Seas and, being adjacent to the mouth of the River Po (Padus), the navigation of that river system. This enabled the fleet to be a part of the protection of Italy north of the Apennines. The fleet could and did also operate around the Peloponnese and into the eastern Mediterranean.

Like the Misene fleet, a detachment from this fleet was stationed in Rome, again with their own quarters. The fleet’s harbour was one of the best on the Italian Adriatic coast, which has few natural harbours and the location also linked with the northern end of the Via Flaminia, a direct link to Rome. From their location, the fleet could provide rapid connections and communications with the north end of the Adriatic (through the port of Aquileia), to the eastern Alpine and upper Danube regions, or across to Split (Salonae), the Dalmatian coast (previously, with its myriad islands, a notorious haunt of pirates) and connections with the middle Danube area. In the south, stations were maintained at Ancona and Brindisi, the latter one terminal of the route to Durres in Albania (Durazzo, Dyrrhachium) connecting with the Via Egnatia through the Balkans to Thessaloniki (Salonica) and Byzantium (later Constantinople/Istanbul). There were two other stations in western Greece, to cover the Gulfs of Patras and Corinth and the Ionian Islands and passage along the western Peloponnese. Units of the fleet operated from time to time in support of the Misene fleet, especially in the third century AD, with frequent campaigns in the East.

Provincial fleets

Two other fleets were established for the eastern Mediterranean, the Classis Alexandrina and the Classis Syriaca. The former was to control the sometimes troublesome North African and Palestinian coasts and to oversee the increasingly important trade route, including grain transports, from Egypt to the West. The policing and regulation of traffic on the River Nile was the responsibility of the potamophylacia, a separate river police force organised by and inherited from the Ptolemies that had its own men, ships and bases on the river. This could be, and was from time to time, augmented by the fleet when needed; the potamophylacia was wholly absorbed by the fleet in the second century AD.

One other area of operations for this fleet was the Red Sea. The Romans did not maintain a permanent fleet on this sea, but did organise a fleet in 26 BC with personnel drawn from the Alexandrine fleet for a military expedition to what is now Yemen. This fleet had eighty warships and 130 transports, the latter being requisitioned local merchant ships. The warships, which in the absence of any anticipated opposition (there was in fact none) need only to have been of the smallest types, were ‘built’ on the Red Sea shore, presumably from prefabricated parts brought overland. Some were provided by the allied kings of Nabatea and Judaea, who also contributed forces for the venture. It is possible that the Nile-Red Sea canal was in use and that some of the ships could have been brought from the Mediterranean by this route. The canal was prone to silting if not constantly maintained; Trajan (ruled AD 98–117) restored the canal and with his annexation of the kingdom of Nabatea in AD 106, had both sides of the northern Red Sea under Roman control. Even so, there remains no evidence for any but occasional forays by the Alexandrine fleet on to the Red Sea.

The Classis Syriaca had its headquarters at Seleucia near Antioch on the north Syrian coast and was placed to cover the Levantine coast as well as the south coast of Asia Minor, also previously a notorious centre of piracy. The fleet also extended into the southern Aegean Sea and being the closest to the ever-present threat of Parthian and Persian power to the east, was instrumental in maintaining transport and communications links with the West, frequently having to transport troops to oppose threats or attacks from the east.

Although each fleet was a totally independent entity, their spheres of responsibility could and did overlap. Ships from separate fleets operated together seamlessly for differing operations, ships from other fleets being drafted to assist in major operations, such as the transport of troops and supplies for campaigns against the Parthians, or Trajan’s campaigns across the Danube.

The North African littoral of what is now Algeria and Morocco had been in a state of unrest and occasionally open revolt after the emperor Gaius (Caligula, emperor AD 37–41) had its ruler murdered. Under his successor, Claudius (emperor AD 41–54), the whole territory was brought under direct Roman rule in AD 41 and 42 and formed into the provinces of Mauretania Caesariensis (the eastern part) and Mauretania Tingitana (the western part). These campaigns were supported by the Classis Misenensis, augmented by ships of the Alexandrian and Syrian fleets. The capital of Caesarea (Caesariensis) received a naval base with its own harbour, distinct from the merchant harbour and which became home to a permanent naval detachment or squadron. This unit was made up from ships and men from the Alexandrine fleet, but was not constituted as a fleet in its own right, but remained an adjunct of its parent fleet, which was well able to supply the required ships and men as well as support, from the relative tranquillity of the eastern Mediterranean.

The squadron, although sufficient to patrol the coasts, including the Atlantic seaboard,, was not able to deal with major conflagrations and Misene ships had to intervene to suppress raiding by Mauretanians in AD 170 and 171. It intervened again in AD 260, to help suppress revolts in Africa and Numidia (part of modern Algeria). There were more peaceful interventions when naval personnel were employed to apply their abilities in civil works, for example, in AD 152, the engineer in charge of building an aqueduct at Saldae in Mauretania reported that ‘the constructor and his workmen began excavation in their presence, with the help of two gangs of experienced veterans, namely a detachment of marine infantry and a detachment of alpine troops…’

The expansion of the empire to the line of the Danube under Augustus, completed by 12 BC, engendered the formation of two more fleets for the defence of that river. The Danube was naturally divided into upper and lower parts by the Iron Gates Gorge (between Orsova and Donti Milenovac, about 100 miles (160 km) east of Belgrade), which was at that time an impassable torrent. For the new border adjacent to the provinces of Noricum, Rhaetia and Pannonia (approximately modern Switzerland, Austria and western Hungary), flotillas previously formed and used in the advance on the rivers Sava (Savus) and Drava (Dravus), were moved up to the Danube and reinforced to form the Classis Pannonica, with headquarters at Zamun, near Belgrade (Taurunum).


Roman Imperial fleets II


West Roman Triremis Vehiculum, Dromon, 530 AD


Enneris, Imperial Roman time

For the lower Danube, from the Iron Gates to the Black Sea, the Classis Moesica was formed, with headquarters at Isaccea (Noviodunum) in Romania. This location is nearer to the Danube delta than the centre of the fleet’s stretch of the river because this fleet’s responsibilities extended into the western Black Sea and the sea routes between the river mouth and the Bosporus. With the increase of Roman power into and around the Black Sea area, the fleet acquired the further duties of protecting Roman interests in the western half of the Black Sea, including the Hellenistic cities around its northern shore and the strategic link with the Bosporan kingdom of the Crimea and adjacent lands, an important producer of grain. The fleet later established a naval base at Chersonesus.

To cover the southern shores of the Black Sea and to maintain a watch over its east coasts, the Classis Pontica was formed. The Romans had had some forces on the north coast of Asia Minor since they were organised by Pompeius in the sixties BC. The last client king of Eastern Pontus, Polemo II, was ‘retired’ in AD 63 and his kingdom annexed and made part of the province of Galatia. The former royal fleet was taken over and merged with the Roman ships to form the fleet, with its headquarters at Trabzon (Trapezus), later moved to Cyzicus (on the Sea of Marmara). For once there is a little evidence of the strength of this fleet, which was noted at some forty ships. This fleet extended its influence into the south-eastern Black Sea to eradicate the spasmodic piracy there and to cover the important military supply route from the Bosporus and Danube Delta, to north-east Asia Minor, to supply Roman forces facing Armenia and the Parthians. Subsidiary bases were set up in the second century AD near Poti (Phasis) in Georgia and Asparis, near the present-day Turkish-Georgian border.

The other great riverine frontier of the empire, the Rhine, also had its own fleet, the Classis Germanica. Caesar had established the border of the Roman Empire on the Rhine in the mid-first century BC, leaving garrisons with some small craft for patrolling, at intervals along its length. Legionary bases grew into towns at, for example, Mainz, Koblenz and Xanten. In 12 BC Augustus resolved to advance the border to the River Elbe (Albis) and in that year the future emperor Tiberius and his brother Drusus crossed the Rhine. Drusus concentrated the Roman’s existing ships and boats at Bonn and formally constituted them as the Classis Germanica, with its own praefectus and administration, as for the other fleets. He also had new ships built and brought trained crews from the Italian fleets by way of reinforcement. Fleet headquarters was at Cologne (Colonia Agrippinensis) and apart from the Rhine and Moselle, it was charged with patrolling and incursions into the tributaries entering from the right bank, such as the Neckar, Main and Lippe. Seagoing ships had to be acquired to cover the mouth of the Rhine and adjacent coasts. Later, with the addition of Britain to the empire, it had to operate jointly with the Classis Britannica to maintain the essential link between the armies of the Rhine and Britain. After the loss of the embryo province east of the Rhine after AD 9, the river became the permanent border, later altered from the middle reaches up by the advance into the Agri Decumates (the re-entrant between the Rhine and Danube) between the late first and third centuries AD.

After his conquest, Caesar had left ships on the north coasts of Gaul to patrol, deter any piracy and secure the trade in the English Channel. A few such ships under local military control had been sufficient for the predominantly peaceful area but in the forties AD Claudius resolved to add Britain to the empire. The ships were formally redesignated as the Classis Britannica and reinforced by new building and by ships with crews and specialist personnel brought around from the Mediterranean by sea. Preparations for the invasion had most likely, in view of their extent, started in the reign of Gaius, but were completed by his successor, whose forces invaded in AD 43. The new fleet was augmented by ships from the Classis Germanica and was vital to the success of the invasion which depended wholly on supplies from Gaul. After the initial invasion, the Classis Britannica had to continue to grow and extend its area of operations, as the Romans expanded their area of occupation, the fleet eventually operating right around the British Isles. The fleet’s prime purpose would remain however, to maintain the essential links with the mouth of the Rhine and the armies there, as well as with Gaul and the fleet headquarters was accordingly established at Boulogne and with another, a little later, at Dover (Dubris).

There were two other formations classified as fleets which do not seem to have been permanent, but pass briefly through the surviving records, indicating that they were formed for a particular purpose, at the ending of which they were disbanded. The first was the Classis Perinthia, formed by Claudius in AD 46 to cover his annexation of Thrace, after which there is no other indication of its continued existence. Thereafter responsibility for the Thracian coast passed to the Classis Moesica.

The other formation was the Classis Nova Libyca, which appears from the scant references, to have been formed in the late second century AD to reinforce the Libyan shores at a time of unrest there. It is not heard of again beyond the mid-third century AD and either was disbanded, it’s ships returned to their parent fleets, or was lost in the great upheavals of that time.

The late Empire

These nine imperial fleets continued to operate for nearly 200 years until caught up in the upheavals of the third century AD. With the Empire hard pressed by internal dissention and external pressure, the fleets could not be maintained as before, as the first call on available manpower, resources and money was the army. Neglected and denied resources, the great praetorian fleets deteriorated to a shadow of their former selves, as did the other Mediterranean fleets. The Black Sea was progressively abandoned and the riverine fleets seriously overstretched and at times overwhelmed.

With the accession as sole emperor of Diocletian in AD 285, stability was returned to the empire, together with the need to reorganise the remains of the fleets that he had inherited. The Classis Britannica, due to its particular location and function as a mainstay of the province’s garrison and defence and in the face of increasing barbarian seafaring activity and ability, had remained the least neglected and probably the best fleet left in the empire. It was part of the command of Carausius, commander in Britain, who improved its strength and efficiency and went on to the offensive against the sea raiders. It alone continued as recognisably the classis of yore.

The rest of the fleets in the Mediterranean were reorganised into a greater number of smaller squadrons, rather than try to reconstruct the great fleets. Each squadron was commanded by a praefectus and assigned to a military district and placed under the overall command of the military commander-in-chief for that district; in so doing the fleets lost their former independent identities as classes. Thirteen such squadrons were formed and became the basis of naval organisation in the Mediterranean thereafter.

The Danube border was reorganised into four new provinces, Moesia Prima and Secunda, Scythia and Dacia Ripensis. The former Classis Moesica was also divided into four parts, one allotted to each of the new provinces and again, under overall command of the local military commander. The fleet remained independent only insofar as that part of it based in the Danube delta and responsible for the Delta and with the Thracian coast. The Classis Pannonica disappears from the record in the third century AD, but Roman naval forces on the Upper Danube are known from the fourth century AD, once again as units forming part of the border forces, rather than as an independent fleet as before.

The Classis Germanica had all but ceased to exist by the mid-third century AD, local commanders having to employ whatever ships and crews they could acquire on an ad hoc basis. On the restoration of the Rhine frontier by the co-emperor Maximian after AD 286, naval forces were again built up but again on an area by area basis and allocated to local military commanders, the classis was not reconstituted as such.

These dispositions continued to serve through to the late fourth century AD, when with the loss of territory, including increasing parts of the Mediterranean seaboard, even the squadrons lessened in numbers, especially in the west until Roman naval forces were formed from whatever ships could be amassed and crewed.

The Islands and Italy, 210–207 BC: The Romans Defeat the Carthaginian Fleet


Marcus Valerius Laevinus sailed to Rome in 210 to report that the war for Sicily was over and that a deserted land was again in cultivation. However, it is obvious that the Carthaginians had not given up and that the Romans still had to defend the islands. The consul sent praefectus classis Marcus Valerius Messalla to Africa to make a plundering and espionage expedition. He approached the coast with fifty ships before daybreak and made an unexpected landing on the territory of Utica. He ravaged it far and wide, captured many people and much booty, returned to his ships and sailed back to Lilybaeum. Livy does not relate any Carthaginian resistance during this two-week operation, so it seems that the Romans were able to sail without being stopped. The captives were interrogated and the information was reported to Laevinus:

That five thousand Numidians were at Carthage under Masinissa, son of Gala and a most impetuous young man; and that other soldiers were being hired everywhere in Africa, to be sent over to Hasdrubal in Spain, so that he should cross over into Italy with the largest possible army as soon as he could and join Hannibal; … furthermore that a very large fleet was being made ready, for the purpose of recovering Sicily.

Consequently, Quintus Fulvius Flaccus was appointed dictator to hold the elections and Marcus Valerius Laevinus continued in Sicily as proconsul. The Carthaginian naval attack on Sicily never took place. Nevertheless, the Carthaginians attacked Sardinia with a fleet of forty ships. First they laid waste to the region of Olbia; then, after the praetor Publius Manlius Vulso appeared there with the army, the Carthaginian fleet sailed to the other side of the island and ravaged the territory of Carales. The fleet returned to Africa with a large amount of booty.

In 210, the Tarentines intercepted the Roman convoy that was sailing from Sicily along the Italian coast to bring supplies to the garrison in Tarentum. Democrates defeated Decimus Quinctius’ fleet of about twenty ships with an equal number of ships at Sapriportis, about 22 kilometres west of Tarentum. The transport ships escaped to the sea. The Romans prevailed on land and kept the citadel. Here we have some rare information about how ships were collected from the cities that were under obligation to supply them. To begin with, Decimus’ fleet at Rhegium consisted of triremes and smaller ships and some quinqueremes. Livy states: ‘By personally demanding from the allies and from Rhegium and Velia and Paestum the ships due under the treaty, he [Decimus Quinctius] formed a fleet of twenty ships … in the neighbourhood of Croton and Sybaris he had fully manned the ships with oarsmen, and had a fleet remarkably equipped and armed considering the size of the ships.’

In 209, the Romans took advantage of the absence of the Carthaginian fleet as it had sailed to the Greek coast and recovered Tarentum. Quintus Fabius Maximus pitched his camp in the mouth of the harbour to besiege the city. Here we see a plan to use some of the ships as a naval siege unit, carrying artillery for shooting missiles at a long range:

Of the ships which Laevinus had had to protect his supplies, the consul loaded some with devices and equipment for attacking city walls, while some of them he fitted out with artillery and stones and every kind of missile weapon. And so also with the merchantmen, not merely those propelled by oars, in order that some crews should carry engines and ladders up to the walls, and others from ships at long range should wound the defenders of the walls. These ships were equipped and made ready to attack the city from the open sea. And the sea was unmolested by the Punic fleet, which had been sent over to Corcyra, since Philip was preparing to attack the Aetolians.

As this was taking place, the city was betrayed to the Romans by the commander of the group of Bruttians that Hannibal had put in place to protect the city. The Romans took a huge amount of booty and 30,000 slaves. Hannibal marched to Tarentum but realized that nothing could be done and retired to Metapontum; his plot to make Fabius follow him there failed.

Lack of resources became an issue in the same year when twelve Latin colonies informed that they were no longer able to send soldiers or money to support the war effort. The senate could do nothing to change their refusal but made sure that the remaining eighteen colonies fulfilled their duty.

In 208, there was another report of naval preparations being made at Carthage with the intention of blockading the whole coast of Italy, Sicily and Sardinia with 200 ships. We cannot pay too much attention to the number of ships, which we have from one source only; what matters is the fact that the Carthaginians could have put pressure on the ports and interrupted shipments, a problem the Romans had faced since the beginning of the war. Consequently, the Romans repositioned their ships. Publius Scipio was ordered to send over to Sardinia for the defence of the island fifty of the eighty ships that he had either brought with him from Italy or captured at New Carthage. The imperium of Marcus Valerius Laevinus was continued in Sicily and the seventy Roman ships there were increased with the addition of the thirty ships that had been stationed at Tarentum the preceding year. With this fleet he was to cross over into Africa and collect booty, if he thought the time was suitable. The praetor urbanus was given the task of preparing the thirty old warships that were in Ostia and of manning twenty new ships with crews, so that he could defend the coast near Rome.

The building of the fleet shows the serious intention of the Carthaginian government to continue the struggle for the islands. It should be seen as a reaction to the loss of New Carthage and Tarentum. There is no information of any Carthaginian attack taking place, apparently because the Romans did not give them any opportunity. Livy gives a frustratingly short description of what became the biggest sea battle in the Second Punic War:

The same summer Marcus Valerius crossed over from Sicily to Africa with a fleet of a hundred ships, and making a landing at the city of Clupea [Aspis], he ravaged the country far and wide, meeting hardly any armed men. Then the foragers were hurriedly brought back to the ships, because suddenly came the report that a Carthaginian fleet was approaching. There were eighty-three ships. With these the Roman fought with success not far from Clupea. After capturing eighteen ships and putting the rest to flight, he returned to Lilybaeum with a great quantity of booty from the land and from the ships.

Livy states that the Roman fleet ravaged the African coast again the following year. Despite the many similarities in stories from 208 and 207, they are not dupli cates of the same event. Marcus Valerius Laevinus was leading the fleet that sailed from Sicily and laid waste the territory of Utica and Carthage. When the Roman fleet was returning to Sicily, a Carthaginian fleet with seventy warships met them. Again, Livy does not give any details but only states that seventeen Carthaginian ships were captured, four sunk at sea and the rest of the fleet routed and put to flight. We do not know the size of the Roman fleet. It returned to Lilybaeum with much booty. Livy adds that thereafter, since the enemy ships had been expelled from the seas, large supplies of grain were brought to Rome.

This information looks like any other story of raids on the enemy territory but the significance is that now there was a Punic fleet confronting the Romans and that it suffered serious losses. Consequently, the possibility of attacking the islands and the coast of Italy was lost. The Romans had taken the edge off the new Punic campaign before the fleet had had the chance to do anything. Now we see the Romans implementing the strategy they had in mind at the beginning of the war, when Titus Sempronius Longus was sent to Lilybaeum with the mission to prepare for the invasion of Africa. Because of the failure to defeat the Punic fleet then, the Romans had to defend the islands for a decade more but these two battles made the turning-point in the war and now the Romans could go back to their original plan. The Carthaginian losses probably explain why there was no attempt to stop Scipio from crossing to Africa in 204.

The Romans awaited the approach of Hasdrubal in Italy in 207 with great anxiety. The sea route from Spain to Italy was still unusable for the Punic fleet, as it had been in 218, and Hasdrubal took the same route that Hannibal had used. However, the Roman situation was now different from 218. The Romans had had experience and time to get ready. They were informed by the Massilians that Hasdrubal had passed over into Gaul. Once in Italy, Hadsrubal sent messengers to find Hannibal and give him instructions to link up with Hasdrubal’s army in Umbria. The Romans, however, caught the messengers – four Gauls and two Numidian horsemen – who had come all the way to Metapontum to find Hannibal. Livy’s narrative is difficult to follow and we cannot be sure of all the routes Hannibal took but he moved around southern Italy to break away from the Romans. The Romans made sure that the brothers could not meet. Livy explains the Roman strategy:

For they felt that Hasdrubal must be met as he came down from the Alps, to prevent his stirring up the Cisalpine Gauls or Etruria, which was already aroused to the hope of rebellion, and likewise that Hannibal must be kept busy with a war of his own, that he might not be able to leave the country of the Bruttii and go to meet his brother.

The pressure from the Punic fleet had eased and the Romans could prepare for Hasdrubal’s arrival by transporting troops from several fronts. Livy refers to some unnamed authors and states that Scipio sent 8,000 Spaniards and Gauls, 2,000 legionary soldiers and 1,000 cavalry of Numidians and Spaniards. Marcus Lucretius brought these in ships. Gaius Mamilius sent 4,000 archers and slingers from Sicily. Slave volunteers were recalled to their standards. The senate gave the consuls Gaius Claudius Nero and Marcus Livius Salinator unrestricted freedom to fill up their numbers from whatever source they pleased, of selecting men from whichever army they liked and of exchanging and removing men from one province to another. They also resorted to a resource so far unused: the settlers on the sea coast, who had been exempt from service. Alsium, Anxur, Minturnae, Sinuessa and Sena Gallica were compelled to furnish soldiers. Antium and Ostia were still exempt. The number of soldiers thus enrolled did of course not change the total significantly but all this shows the need to use exceptional methods to find men. The consuls destroyed Hasdrubal and his army in the Battle of the Metaurus in northern Italy in June 207. Livius celebrated a triumph and Nero an ovation. After the defeat, Hannibal withdrew to Bruttium.

Crews in Byzantine fleets


A modern depiction of a Byzantine flamethrowing warship, using Greek Fire against an enemy ship (probably of the opponent Muslim fleets). In the foreground: the mechanism and the siphon of ejection of Greek fire in the interior of a Byzantine Dromon (artwork by Giorgio Albertini)

By John H. Pryor

In spite of the fact that some crews in Byzantine fleets at various times were well regarded, for example the Mardaites of the theme of the Kibyrrhaiōtai, there is little evidence to suggest that, in general, Byzantine seamen were so skilled that this gave Byzantine fleets any edge over their opponents. It is true that Byzantine squadrons managed to defeat the Russians on all occasions when they attacked Constantinople: in 860, probably in 907 under Oleg of Kiev, in 941 under Igor, and in 1043 under Jaroslav. A fleet also defeated the Russians on the Danube in 972. However, rather than being attributable to any qualities of Byzantine seamen, these victories were due to the triple advantages of Greek Fire, dromons and chelandia being much larger than the Norse river boats of the Russians, and (except in 972) being able to fight in home waters against an enemy far from home. The last is true also of the defeat of the Muslim assaults on Constantinople in 674–80 and in 717–18. In both cases, it was the advantage of home waters against the disadvantage of campaigning hundreds of miles from sources of supplies, the problems faced by the Muslims of surviving on campaign through the winter, and Greek Fire that proved decisive. The same is probably true of the victories over the fleets of Thomas the Slav in 822–3.

In general, the record of Byzantine fleets from the seventh to the tenth centuries was hardly impressive. To be sure, they did achieve some notable victories: the defeat of the Tunisians off Syracuse in 827–8, the defeat of a Muslim fleet under Abū Dīnār off Cape Chelidonia in 842, the victory of Nikētas Ooryphas over the Cretans in the Gulf of Corinth in 879 and of Nasar over the Tunisians off Punta Stilo in 880, the victory of Himerios on the day of St Thomas (6 October), probably in 905, the defeat of Leo of Tripoli off Lemnos in 921–2, the victory of Basil Hexamilitēs over the fleet of Tarsos in 956, and the defeat of an Egyptian squadron off Cyprus in 963. Against that record, however, have to be balanced many disastrous defeats: of Constans II at the battle of the masts off Phoeinix in 655, of Theophilos, the stratēgos of the Kibyrrhaiōtai, off Attaleia in 790, a defeat off Thasos in 839, the defeat of Constantine Condomytēs off Syracuse in 859, the annihilation of a fleet off Milazzo in 888, a defeat off Messina in 901, the disastrous defeat of Himerios north of Chios in 911, the defeat of a Byzantine expedition in the Straits of Messina in 965, and of fleets off Tripoli in 975 and 998.

Although the tide of Byzantine naval success ebbed and flowed over the centuries, as other circumstances dictated, nothing suggests that the quality of the Empire’s seamen was in any way decisive. Indeed, there are occasional pieces of evidence that suggest that all was not always happy in the fleets. Some time between 823 and 825, John Echimos, the ‘deputy governor’, (ek prosōpou), the acting stratēgos, of the theme of the Kibyrrhaiōtai, confiscated the properties of seamen of the fleet. After he had become a monk and taken the name Antony, later to become St Antony the Younger, he was interrogated as to his reasons for doing so on the orders of the new emperor, Theophilos (829–42). According to the author of his Life, his explanation was that they had been partisans of Thomas the Slav in his rebellion of 821–3 and were ‘hostile to Christians’, thus implying that they were iconoclasts, and that he had confiscated their property and given it to supporters of Theophilos’ father, Michael II (820–9). In spite of this explanation, the emperor initially imprisoned him and had him interrogated, suggesting that there was more to the story and that he rejected the explanation. The fleet of the Kibyrrhaiōtai had, indeed, joined Thomas the Slav, as it was also later to join the rebellions of Bardas Sklēros in 976–9 and Bardas Phōkas in 987–9, and it is clear that, at times, there must have been serious disaffection in what was the front-line fleet of the Empire in the ninth and tenth centuries.

In 880, the expedition sent under the command of Nasar, the droungarios touploimou, to counter an attack in the Ionian sea by a Muslim fleet from Tunisia was forced to a temporary halt at Methōnē by the desertion of a large part of the crews. Why they deserted is unknown, but we can be fairly sure that it was not a simple question of their having ‘lost their nerve’, as the Vita Basilii suggested.

The Byzantine warships and their tactics


Reconstruction of an early 10th-century Byzantine bireme dromon by John H. Pryor, based on references in the Tactica of Emperor Leo VI the Wise. Notice the lateen sails, the full deck, the fore- and mid-castles, and the Greek fire siphon in the prow. The above-water spur is evident in the bow, while the captain’s tent and the two steering oars are located at stern.

The typical high-seas elite warship of the empire in the period was the dromon (from the Greek dromeas, meaning `the runner’). This was a two-masted fully decked bireme with two banks of oars, one rowed from below the deck and one from above it. There were twenty-five oarsmen on each side of each deck, thus raising the total number of oarsmen to a hundred, all fully seated. The marines and the officers of the ship numbered around fifty men, while the ousia, the standard complement of a war galley (its crew excluding the marines and the officers), totalled 108 men. Another type of warship that had the same features as the dromon was the khelandion; both Ahrweiler and Pryor consider these two types of vessel to be almost identical. However, although the Greek primary sources used these two terms indiscriminately, it is interesting to mention that Theophanes identifies the khelandia primarily as horse transports. The Arabic primary sources, however, use only the term khelandion to describe Byzantine warships.

A smaller but much faster type of ship compared to the dromon and the khelandion was the galea. It derived from the same design mentality for a war ship and it had two sails (the one amidships being smaller by a third) and probably one bank of oars on the deck. Because of its speed, however, this type of ship was used primarily for courier service and, during campaigns, for the transport of orders. There is also mention of galeai being used in espionage. Other types included the supply and carrier ships like the pamphylos, which was `like a baggage-train, which will carry all the equipment of the soldiers, so that the dromons are not burdened with it; and especially in time of battle, when there is need of a small supply of weapons or other materiel, [these] undertake the distribution’.

In the non-tidal waters of the Mediterranean war galleys, like the dromons and the khelandia, would have been suitable for any sort of landing on a hostile beach, unlike the heavy and round-hulled pamphylos, which required a dock. The horse-transport units of the Byzantine fleet had been equipped with a climax since at least the early tenth century, which was a ramp used for the loading and unloading of the horses from the ship’s gunwales, either from the stern or usually from the bow. This term is mentioned in the De Ceremoniis for the Cretan expeditions of 911, 949 and 960/126 and reveals the necessary modifications to the ships when they had to carry horses, such as hatches not just to the sides but also on the decks, leading down into the holds, while further modifications would have been engineered in the hulls of the ships concerning the stabling of the horses. According to Pryor, the khelandia were indeed specialised horse transports, able to carry between twelve and twenty horses. But these must have been built differently from dromons when it comes to the dimensions of the ship’s beam, which would have been much wider to accommodate both the lower bank oarsmen and the horses. A significant structural difference between the tenth-century Byzantine transport ships and their Italian counterparts in the twelfth century was that the latter placed both banks of oarsmen on the upper deck, thus making more room for the horses in the ship’s hull.

Turning to the battle tactics of the Byzantine navy, the existence of an above-water beak in the larger warships reveals a fundamental difference between the ancient Greek and Roman naval tactics and those used by the Byzantines, at least after the early tenth century. This beak, replacing the below-water ram, possibly as early as the sixth century, indicates a change in the objectives of naval engagements, from penetrating the enemy ship’s hull below the water line to damaging the ship’s oars and upper hull and bringing it to a stop in order to board it and capture or burn it.

What is obvious in all contemporary treatises of naval warfare is the same spirit of avoidance of battle at all costs, identified as Vegetian strategy by modern historians, which characterised the Byzantine attitude towards warfare on land. The basic idea of Byzantine warfare at sea follows the simple dicta by Syrianus Magister (c. 830-40s) that `if the enemy is overwhelmingly stronger than us and a great danger hangs over our cities, then we should avoid war and overcome the enemy by wisdom rather than might’. Leo VI also strongly urges an admiral that:

You must indeed deal with the enemy through attacks and other practices and stratagems, either with the whole of the naval fleet under you or with part of it. However, without some urgent compelling reason for this, you should not rush into a general engagement. For there are many obstacles [in the workings] of so-called Tyche [Luck] and events in war [are] contrary to expectation.

When a decision to engage the enemy was taken by the senior officers, then the fleet would deploy its squadrons in several formations depending on a series of factors such as `time, by attacking the enemy at a moment when we have the winds as allies, as happens frequently with off-shore winds; place, [by using] the sea between two pieces of land, or a river, [areas] in which the numbers of the enemy are useless because of the narrowness of the sea’. The author of the Taktika provides his readers with a variety of naval formations to engage the enemy (§§50-6); the two most commonly used were the crescent-shaped and the straight line:

Sometimes [you should draw up] a crescent-shaped or sigma-shaped [i. e. C-shaped] formation in a semi-circle, with the rest of the dromons placed on one side and the other [i. e. of the flagship] like horns or hands and making sure that the stronger and larger [ships] are placed on the tip. Your Gloriousness [should be positioned], like a head in the deep of the semi-circle [. . .] The crescent arrangement should be such that, as the enemy attack, they are enclosed within the curve. Sometimes you will form the ships on an equal front in a straight [line], so that, when the need calls, [you can] attack the enemy at the prow and burn their ships with fire from the siphones.

The tactical objective of the crescent-shaped formation was for the stronger ships on the sides of the formation to overwhelm the enemy ships and then turn around and attack the rest of the formation on their exposed flanks where they were most vulnerable. Once the opposing units came into close proximity with each other they would attack the enemy ships and their crews with bows and arrows, snakes, lizards and other dangerous reptiles, pots with burning lime or tar and, of course, with Greek fire, projected either through the ship’s siphones, through small hand-siphons or thrown against the enemies in a form similar to small hand-grenades. The importance of the proper management of the preliminary missile phase was indicated by the emperor’s insistence on using the projectiles effectively, not wasting them against an enemy protected by shields, and ensuring that neither supplies were exhausted nor the crews exhausted themselves in hurling them. When the ships were close enough, boarding detachments were sent to the enemy ship and the result of the naval battle largely depended on the courage and the fighting abilities of the boarding teams. For that reason,

apart from the soldiers or the upper oarsmen, [all others] however many there might be, from the kentarkhos down to the last [man], should be kataphraktoi – having weapons such as shields, pikes, bows, extra arrows, swords, javelins, corselets, lamellar cuirasses, helmets, vambraces – especially those engaged in fighting hand to hand in the front line of attack in battle.

Finally, if we follow the writings of Leo VI, a potentially decisive weapon that came to the fore at this point of the naval engagement were the `gerania [cranes] or some similar contrivances, shaped like a gamma [A], turning in a circle, to pour either wet flaming pitch or the processed [fire] or anything else into the enemy ships when they are coupled to the dromons when the manganon is turning over them’. This technique was coupled with the thrusting of pikes from the lower bank of the dromons through the oarports, a tactic that Leo claims had only recently been devised.

The Harpax


HARPAX Roman word for the Greek harpagos or grappling hook used by the Roman NAVY; a combination harpoon and grappling iron consisting of a spar five cubits (2.25m, or 7ft 3in) long with a ring at each end. An iron hook was fastened to one of the rings, and a large number of ropes, twisted together into one cord, to the other. Fitted for use with the ballista, it would be embedded in an enemy vessel when fired, enabling the ship to be hauled in and boarded. An iron casing surrounded the spar, preventing the enemy from hacking it free.

The harpax enjoyed its greatest hour at ACTIUM, on September 2, 31 B. C., when the fleet of Octavian (AUGUSTUS) routed the ships of Antony and Cleopatra. Using the lighter Liburnian vessels, Agrippa, Octavian’s admiral, moved around Antony’s heavier ships, pinning and boarding them.

The naval Battle of Actium (Marc Anthony and Cleopatra versus Octavian) saw yet another ingenious new naval weapon, the harpax, an iron grapple hurled by a catapult at an enemy ship, which was then hauled in by a winch for boarding.

Both sides gathered large fleets and assembled legions, but Octavian, with his normal prudence, took his time. Finally, in 31 he set out with hundreds of ships and 40,000 men, landing in Greece and marching south to Mikalitzi, north of Nicopolis on the Bay of Comarus. Antony, possessing a like number of land forces, also had at his command a combined Roman-Egyptian fleet of 480 ships. The advantage rested with Antony in naval terms, because his vessels were large and heavy. Octavian, however, possessed two elements that were to prove pivotal to the outcome: his admiral AGRIPPA, and his lighter Liburnian ships, which were equipped with the HARPAX. Antony, encamped just south of Actium, nevertheless stood a good chance of victory.

The battle was really two encounters in a single day, the fierce naval conflict in the morning and a halfhearted rout on land that afternoon. The naval engagement began with the two fleets facing one another. Octavian’s force was divided into three sections – a center and two wings. Agrippa commanded the northern wing and was admiral in chief. ARRUNTIUS led the center, and Octavian was in charge of the southern wing. On the Egyptian side, Antony took command of the northern squadrons, opposite Agrippa. Marcus Octavius was opposed to Arruntius, and Savius sailed against Octavian’s ships. Cleopatra headed a reserve squadron of 60 ships behind the center of the Egyptian fleet.

The tactical advantage would fall to the commander who penetrated the other’s flanks, and here the battle was won by Agrippa. Antony fought valiantly, but the unreliable and disloyal ships of his center and south wing broke ranks. Cleopatra sailed to safety, probably signaled by Antony to do so, although the historian DIO CASSIUS dismissed her flight as the act of a woman and an Egyptian. Antony, with his own ship pinned by a harpax, transferred to another vessel and also fled toward Egypt.

Very Early Warships

The Greek Armada sails to the city of Troy in Warner Bros. Pictures' epic action adventure "Troy," starring Brad Pitt, Eric Bana and Orlando Bloom. PHOTOGRAPHS TO BE USED SOLELY FOR ADVERTISING, PROMOTION, PUBLICITY OR REVIEWS OF THIS SPECIFIC MOTION PICTURE AND TO REMAIN THE PROPERTY OF THE STUDIO. NOT FOR SALE OR REDISTRIBUTION.

Piracy was an ancient custom in the eastern Mediterranean. But none of this normally involved fighting at sea. Pirates rarely pursued merchants on the open sea because all ships carried both sails and oars and were therefore difficult to catch. (Pure sailing ships did not appear until the late sixth century BC.) The standard piratical procedure was doubtless that described in the Odyssey: the raiders beached their boats in the vicinity of a coastal town and then captured the place by land. Raiders could also blockade harbours by intercepting ships at the harbour mouth, and we hear of Levantine ports in the Bronze Age being blockaded in wartime, but as no ship could stay out to sea for very long this strategy required prior control of the coast so that the port could be besieged by land and sea simultaneously. In all these cases it would obviously have been desirable to cut off and board enemy ships at sea, but for the reason already mentioned this was difficult to do. The relief at Medinet Habu shows Egyptian ships intercepting the invading Philistines; but that was in the mouth of the Nile, and even there the feat must have required good timing.

None of the ships in the Medinet Habu relief have rams, so this device did not exist around 1200 BC. But the evidence of Greek vase paintings shows that by around 800 BC the practice of fixing bronze rams to the prows of ships so that they could be used as weapons against other ships had become standard in the Mediterranean. Owing to the lack of pictorial records from the intervening centuries we cannot say with certainty when or where this device was invented, but it seems likely that it appeared within a century or so after 1200 BC, for much of the sacking of cities at that time was the work of coastal raiders, and there was urgent need for some method of coastal defence. It is unlikely to have been invented by the raiders, as it is not in the interest of pirates to sink their prey; but after coastguards had rams, pirates of course acquired them too. The likeliest inventors of ramming were the Phoenicians, the leading seafarers of the time.

The standard warship of the early Iron Age was the penteconter, a 100-foot galley propelled by fifty oarsmen, twenty-five on each side; the word ‘warship’ is somewhat misleading, as there was no distinction between ships of war and merchant vessels, and the penteconters were equally useful for transporting trade goods (which were of small bulk at this time) and protecting them. In such ships the Phoenicians, followed by the Greeks, opened up the whole of the western Mediterranean to trade and colonization. Originally penteconters were built with only one bank of oars. The next step was the bireme, a shorter and more seaworthy vessel with its fifty oars arranged in two superimposed banks. This was in use by 700 Be; an Assyrian relief of that date shows the king of Tyre embarking in a bireme.

None of this amounted to much ‘sea power’ in the modern sense of that term; it was more like coastal power. We do not hear of sea battles before the seventh century, not even between Phoenician cities, and no big battles until the sixth, which suggests the fifty-oared galleys were for defensive purposes, to guard harbours and repel pirates. It is doubtful there were any naval tactics, which would require concerted action by a number of galleys. Real sea power had to await the invention of the trireme, a highly specialized ship with 170 oars in three banks, with more than three times the propulsive power of a penteconter, and useful for nothing but warfare. These expensive technological marvels were probably beyond the reach of a city state. They did not become common until the late sixth century, when the Persian Empire became a Mediterranean power, and the Persian king Cambyses, according to Herodotus, became the first man to aspire to command of the sea.