It is ironic that, at the very time Rome established its naval forces on a permanent footing with fixed bases, large-scale naval warfare became obsolete, at least for the next couple of centuries. Actium and the destruction of the Egyptian fleet led to the reduction of the last remaining kingdom in the Mediterranean with any significant naval forces; the newly created Roman imperial fleets patrolled the seas, dealt with pirates and raiders, provided support for land operations and worked the velarium on the Colosseum. The hypothetical army of the military surveyor Ps.-Hyginus does contain marines, but for the purposes of route clearance and road building rather than any maritime role. None the less, the few fleet actions that occurred in our period illustrate many of the same concerns relating to deployment that we see in land battles. Naval battles were more likely to be influenced by the vagaries of weather and wind than those on land, so there could be some delay before conditions allowed a battle to take place, and there was also a much greater random factor than existed in land battles. At Actium Antony was greatly outnumbered by Octavian and so risked being outflanked and his ships taken from both front and rear. As with a landbased battle he made use of the terrain, deploying as close inshore as he could, with his wings protected by the shallow waters that Octavian’s ships could not enter.
As in land engagements missiles played an important role in Roman naval warfare and the ships were frequently equipped with towers to give slingers, archers and artillery greater range and power. Incendiary missiles, particularly fearful weapons at sea, formed part of the arsenal. A missile barrage was fired before ships closed for close combat, and missiles continued to fire throughout the engagement, though not incendiary devices once the ships were at close quarters (App. B Civ. 5.119). Tactics varied depending on the size and manoeuvrability of the ships. The imperial navy, which was unlikely to face a large-scale naval engagement, consisted mostly of smaller ships appropriate to their duties – triremes and two-banked liburnians. The civil wars at the end of the Republic provided the last encounters that involved the larger quadriremes and quinqueremes that had been developed in the arms race of the Hellenistic era (see vol. i, pp. 357-61, 434-43); in the naval battles of the 40s bc size and design proved significant.
At Mylae, Sextus Pompey had smaller, more easily manoeuvrable ships manned by more experienced sailors, so he avoided ramming the enemy head on and instead concentrated on disabling Agrippa’s ships by breaking off the oars and rudders (which required considerable skill and timing), or isolating them and attacking them from all sides. With his sturdier, taller ships which were probably designed with his intended tactics in mind, Agrippa aimed to ram Sextus Pompeius’ ships anywhere and bring the battle to close quarters as soon as possible. Here he had the advantage of size, since his ships could hold more troops, and had the additional height to bring fire to bear on the Pompeian ships. His ships also used a grappling hook to haul the Pompeian ships in to the point where they could be boarded, a device that worked very well both at Mylae and Naulochus (App. B Civ. 5.106, 119). At Actium both sides were content to engage at close quarters, boarding ships and capturing them or destroying them, and this was probably not because of inexperienced or incompetent rowers. The preferred Roman tactics allowed them to play to their strengths in numbers and heavily armed infantry and were probably developed (along with the sturdier ships) for that reason, rather than because the Romans made poor sailors.
As with land battles, once the integrity of the line of battle was broken one side might turn to flight, at which point ships became isolated and more vulnerable to enemy attack. Because naval battles usually took place near to land, fleeing ships might be driven on shore, but pursuing ships had to curb their enthusiasm for the chase or they might end up on shore too (App. B Civ. 5.121). The majority of casualties drowned because they could not swim or because they could not get out of swamped ships, but at Mylae Sextus Pompeius’ smaller boats rowed round picking swimmers out of the water, and it is possible that such lifeboats were deployed in other naval battles (App. B Civ. 5.107).
Command and control in naval warfare was challenging because of the difficulties in seeing what was going on in the midst of battle from the deck of a ship, and also given the problems in communicating. Generals seem to have acted in much the same way as in land battles, commanding from the rear, often on land, or from a flagship in the middle of battle, as both Antony and Agrippa did at Actium. Agrippa had smaller auxiliary craft available at Actium to relay orders and information in the same way that cavalry did in engagements on land (Dio Cass. 50.31), and this was most probably a regular feature of naval battles. Sextus Pompeius controlled his fleet at Mylae from a hill and was able to signal them to disengage because he could see, probably more clearly than anyone commanding on the water, that they were being beaten (App. B Civ. 5.107).
In the Empire, naval operations tended to be on a much smaller scale and usually, with no other naval powers surviving, part of land-based operations such as supporting Trajan’s campaigns across the Danube and into Parthia. Even when fleets and marines were not available, soldiers still made use of the water when appropriate, and were able to operate effectively, mounting artillery on boats at Cyzicus in the civil war between Severus and Niger to fire at the flanks of the enemy armies that had deployed near the lake in an attempt to secure their wings (Dio Cass. 75.6). On Lake Gennesaret, in response to the Jewish waterborne attack, Roman soldiers ensured that their infantry skills could still be an advantage, building rafts which provided a relatively sturdy fighting platform from which soldiers fired on the Jewish boats and boarded them when they came too close (Joseph. BJ 3.505).
Caesar’s warships in the Channel played a key role in supporting the transports involved in his first landings, providing covering fire from slingers, archers and artillery, and ultimately driving the Britons back sufficiently for the infantry to start landing (B Gall. 4.25). The disadvantage with landing troops from warships was that their keels were too deep to beach properly, and the infantry were less than keen to jump into the deeper water; Caesar had transports with him that had a shallower draught, but was unable to use them under the threat from the Britons. For other waterborne operations armies usually had to construct small craft which were agile and had a shallow draught, able to transport infantry and cavalry and capable of acting as landing craft. They were used extensively in raids in northern Germany and in Suetonius Paulinus’ attack on Anglesey in AD 60 (Tac. Ann. 2.6, 14.29). These transports were less suitable for working at sea than on rivers, and nervousness on the part of soldiers in the vessels contributed to the huge losses sustained by Germanicus’ fleet when it was wrecked on the German coast in autumnal storms (Tac. Ann. 2.23-4).
Waterborne operations eased logistical difficulties and enabled troops to be moved swiftly into terrain that would have otherwise been difficult to penetrate, taking the enemy by surprise. Operating in that terrain once there, though, was a particular difficulty for legionary troops who, as we have seen, were not well equipped for operating in wetlands. Such amphibious operations regularly involved auxiliary units of Batavian infantry and cavalry. They, along with other tribes living in the Rhine delta such as the Cherusci and Canninefates, were skilled at fighting in flooded and marshy terrain, and caused major problems for successive Roman armies operating in northern Germany by meeting them on ground that they had chosen. As usual Rome recruited from the areas in which it was fighting and raised units of both Batavians and Canninefates, though it is the former who get all the glory. Batavians carried the river crossing in Kent that caught the Britons by surprise in AD 43, and were very probably the auxiliaries who crossed the Menai straits to capture Anglesey for Agricola. They could cross fast-flowing rivers under arms, providing a valuable element of surprise and fear. They provided both cavalry and infantry (who could also fight highly successfully in the front line of pitched battle) and were inordinately proud of their abilities. Their boastful behaviour and eagerness to show off their skills might be suggestive of the behaviour of ‘elite troops, but Rome had no `special forces’ and generals probably made the best use of the particular skills their units possessed.