THE FIRST PUNIC WAR was the greatest naval conflict of antiquity. The resources both sides lavished on their fleets were truly enormous, and their losses in men and material were staggeringly huge. If our sources are correct, then the battle of Ecnomus in 256 may have involved more people than any other sea battle in history. Sea battles were more common than major land actions during the war and ultimately proved decisive. Polybius marvelled at the scale of the naval war, but even more at the speed with which the Romans, who, he claims, had never before built a warship, adapted to the sea and created a navy able to defeat Carthage with its long maritime tradition. The early years of the naval conflict witnessed a spectacular, and almost unbroken, string of Roman successes over an enemy whose ships were better constructed and crews far more skilful. When the war ended in 241, Rome had replaced Carthage as the unchallenged seapower in the western Mediterranean. The navies created during the war made possible the later victories over Carthage and the Hellenistic kingdoms.1
It was not true that the Romans had no experience at all of constructing and manning warships before 260, but Polybius' exaggeration was pardonable. There had been little need for warships of any size during Rome's steady conquest of Italy, for even those enemies who possessed a navy could be reached and defeated on land by the legions. In 311 the Romans created a board of two officials, the duoviri navales classis ornandae reficiendaeque causa with responsibility for construction and maintenance of warships. Each duumvir seems to have commanded a squadron of ten ships, which were probably triremes. Little is recorded of their activities, although one squadron was defeated with dismissive ease by Tarentine ships in 282. Although certainty is impossible in this poorly documented period, it seems that the fledgling Roman navy was disbanded after the defeat of Tarentum. Instead the Romans chose to rely on ships supplied and crewed by those of her allies with a maritime tradition, notably the Greek cities of southern Italy. As we have seen it was in ships provided by the allies, notably Tarentum, Naples and Locri, that Appius Claudius crossed to Sicily in 264. This was essentially an extension of the traditional Roman reliance on allied military support, save that these cities, known as thesocii navalesy provided ships rather than soldiers. It is worth recalling that the 278 treaty with Carthage had provided for the possibility of Punic ships providing support for the legions. In 267 the number of quaestors was doubled from four to eight, the new magistrates being known as the quaestores classici. It is possible that one of the responsibilities of these men was the regulation of the naval allies, and each may have been responsible for the communities in a particular region of Italy. It is impossible to know just how many ships Rome's allies were capable of providing, but it is unlikely that either in overall numbers or in ship size they would have been capable of challenging Punic mastery, of the sea. Allied ships had difficulty both in transporting the Roman armies to Sicily and in supplying them there during the early years of the war in the face of Carthaginian naval activity. The role of the naval allies remained strictly subordinate to the land armies, which still formed the main effort in any Roman campaign.2
Therefore, the Senate's decision to construct and man a fleet of 100 quinqueremes and twenty triremes, with the intention of directly confronting the Carthaginian fleet, marked a major change in Roman practice. Polybius claims that the decision was made after the fall of Agrigentum encouraged the Romans to extend their war aims beyond the protection of the Mamertines and attempt to expel the Carthaginians entirely from Sicily. It is possible that there had been some advocates of constructing a fleet before 261. A very late source credits Valerius Messala, the consul of 263, with first realizing that a fleet was essential for ultimate victory in the war, but it is uncertain whether this tradition is accurate, or simply a later invention by a family eager to glorify its ancestors. A strong naval capability was clearly essential for the total subjugation of Sicily. Despite the acquisition of Syracuse as a base, it was still difficult for the Romans to supply, maintain, and reinforce their armies in Sicily when the sea routes were dominated by the Punic fleet. Ships were also essential if the Romans were to blockade cities with their own ports into submission, since otherwise the garrisons would be easily re-supplied by sea. Finally, it must have been clear that the Carthaginians' main strength was, and always had been, their fleet. The defeat of this fleet would inevitably be a major blow to Carthage, more so than the destruction of its mercenary armies, and would therefore be a major contribution towards forcing her to submit. This appears to be another example of a Roman decision to escalate the conflict in an effort to achieve a decisive result.3
In general the ancient sources are far less informative about operations at sea than on land. The problem is made worse by the essentially alien nature of oared warships to us. Maritime archaeology has started to provide some information, although wrecks of warships rather than merchantmen are exceptionally uncommon, and much has been learnt by reconstruction. Nevertheless there remain numerous gaps in our understanding of the construction and maintenance of classical galleys, and the strategic and tactical uses of fleets. An indication of this is our uncertainty as to the precise design of the quinquereme, the standard warship of the Punic Wars.
The great naval battles of the fifth century, when the Greeks had defeated the Persian invaders, and Athens and Sparta had vied for dominance, were fought and won by fleets of triremes. The evidence for this type of ship is relatively good, much of it coming from the literature and epigraphy of classical Athens and the excavations of the shipyards in the Piraeus harbour. The full-scale reconstruction of an Athenian trireme in the 1980s AD and its extensive sea trials vastly increased our knowledge. The trireme, or 'three', derived its name from the basic rowing group of three men. Each man sat at a different level and operated an oar 14 feet (c. 4 m) or so in length, those of the upper row projecting from an outrigger. Considerable skill was required in each oarsman for the successful functioning of the ship. Long and sleek, the Athenian trireme was about 120 feet (36.5 m) in length and just under 20 feet (6 m) across at its widest. It carried a crew of about 200, around 30 of whom were deck crew, officers and marines, and the remainder rowers. In trials the reconstructed version reached speeds of 8 knots and could maintain a steady 4 knots for hours on end, with half the rowers resting at any one time. Turns through 180 degrees were completed in a distance equivalent to two and a half ship lengths. These speeds were achieved despite the comparative inexperience of the modern crew and their use of oars which were probably heavier than the originals. Under sail the trireme was able to achieve 8 knots in a favourable breeze. All in all the performance of the reconstructed trireme was remarkably good and challenged many past assumptions about ancient naval warfare.4
In the fourth century several states began to construct larger warships than the trireme. The Carthaginians were the first to build 'fours' or quadriremes, whilst Dionysius I, the tyrant of Syracuse in the early fourth century, was responsible for the design of the 'five' - pentereis in Greek and quinquereme in Latin. The kingdoms which emerged in the Hellenistic world in the late fourth century were able to lavish huge resources on the construction of their fleets. Some of the largest ships were built by the Ptolemaic kingdom of Egypt, including such monsters as Ptolemy II’s 'thirties' and 'forty', but there is no record of anything larger than a 'ten' seeing actual combat. The realization that, although the trireme had three banks of oars, its name derived in fact from the number of each team of rowers goes some way towards understanding the design of these ships. Clearly, galleys with four or five banks of oars would have been absurdly impractical and ones with ten or more utterly impossible. In fact, there is no evidence for any warship in the classical world ever having more than three banks of oars. Therefore in 'fours' and larger ships at least some of the oars must have been operated by more than one rower.5
The quinquereme had a basic team of five rowers, but how were they arranged? Did it have one level of oars rowed by five men each, two levels, one rowed by three and the other by two, or three levels with two oars operated by a pair of rowers and one by a single man? The navies of the Mediterranean powers in the late Middle Ages included many galleys, all of which had a single bank of oars, regardless of how many rowers operated each one. Two men can sit side by side and operate an oar effectively, but if there are more than two rowers per oar, then it is necessary for them to rise to their feet to dip the blade and then hurl themselves back onto the bench when they deliver the stroke. This was the method employed in the Middle Ages and must also have been used in the larger galleys in the classical world. Of necessity, this design required a somewhat broader and heavier hull to accommodate the rowers, which probably made them slower and less manoeuvrable than the sleeker types. It has been suggested by Casson that this had the advantage of reducing the requirement for skilled rowers, since it was only essential for one man per oar to be highly trained. This might seem an attractive prospect for the Romans who were undertaking the creation and manning of a fleet on an unprecedented scale. On this basis Casson argued that the Romans used quinqueremes with a single bank of oars, each rowed by five men, unlike the Carthaginians who used more slender, three-banked 'fives'. This, he felt, explained why our sources emphasized that the Punic ships were individually faster and more manoeuvrable than their Roman counterparts. However, Polybius tells us explicitly that the Roman ships were copied from a captured Carthaginian 'five' and there seems no good reason to reject this evidence.6 The superior performance of Carthaginian ships for most of the war was a reflection of their more highly trained crews, and, at the very beginning, better construction, not a result of a fundamentally different design. On the whole, it is more likely that the quinqueremes of this period had more than one bank of oars. Two levels of oars, with three and two rowers respectively, would have meant an uneasy combination of the two different designs, and it is more probable that the quinquereme had three levels, the lowest with a single rower and the others with a pair. This would make the quinquereme a more logical development from the trireme. The upper level of oars in a trireme were mounted in an outrigger and this has sometimes been perceived as a weakness. Even if this was so and would have remained a failing in similarly designed 'fives', it is distinctly possible that the quinqueremes of the Punic Wars were constructed differently. Morrison and Coates have recently argued from the iconographic evidence that the Carthaginian 'five' differed markedly in its layout from Greek ships of the same size, suggesting that this distinction had its origins in Phoenician building methods. The Punic 'five' had all three levels of oars emerging from a single, deep oar-panel, the oarports being arranged in a chequer-board pattern. This they interpreted as an oarbox containing all the rowers and constructed separately to, and projecting from, the main hull. This would have produced a somewhat wider ship, but may have allowed the hull to be strengthened against ramming and possibly increased storage space in the main hull. This pattern of ship was, they argued, copied by the Romans, offering confirmation of Polybius' account, and continued in use with the Roman navy until well into the Principate. This system offered limited possibilities for development of higher ranked ships, since only a 'six', putting two crew on each oar, was really feasible within the confined space. As they point out, the Romans are not recorded as having floated anything larger than a six during the war, although whether these were of Punic or Greek pattern is unknown.7
Morrison and Coates' interpretation of the evidence is an attractive one, especially as it appears to confirm the literary tradition, but the evidence is too poor to reach a final conclusion. Ultimately there must remain some doubt about the precise nature of the quinquereme. However, certain reasonably confident assertions can be made about its capabilities and general characteristics. The crew of a Roman quinquereme consisted of 300 men, of whom about twenty were deck crew and the remainder rowers. At Ecnomus, Roman ships carried 120 marines, but this was because a major encounter was anticipated, and the normal complement was probably fewer, perhaps around forty. Athenian quadriremes were accommodated on the same slipways originally constructed for triremes and cannot have been much larger than these. 'Fives' were markedly higher than these earlier ships and were probably longer and a little broader as well. They were certainly slower and less manoeuvrable than threes and fours, although their greater mass allowed them to make better headway in rougher seas, and increased the effect when they rammed another ship.
There were two main tactical options open to ancient galleys, ramming and boarding. The amount of missile fire which could be delivered by a ship's marines, and the artillery mounted on the larger ships, was insufficient to inflict serious or incapacitating damage on an enemy vessel. At best such fire served to suppress an enemy crew preparatory to boarding. Shooting remained an adjunct to the main methods of attack and for this reason the wind was too uncertain a means of propulsion to be relied upon during a battle. Therefore all decisive combats necessitated physical contact between the opposing ships.
The earliest rams mounted on oared warships had been pointed in shape, but by the fifth century these had been replaced with much blunter devices. As ships became more powerful, there was a real danger that a narrow-headed ram would become deeply fixed into the enemy hull that it could not easily be extricated, immobilizing the ramming vessel as surely as the victim. For the same reason, it was normally inadvisable to ram from an angle higher than 60 degrees, since this also ran the risk of too deep a penetration. A third-century ram found off the coast of Athlit in Israel and now in the National Maritime Museum at Haifa is blunt-headed, broadens towards its tip and has wider projections on either side. It is 7 feet 6 inches (2.2 m) long, 30 inches (76 cm) at its widest point and 37.75 inches (96 cm) at its highest and weighs 1,023 lb (464 kg) The ram is probably from a Ptolemaic warship; Casson suggested that it may have come from a 'four' or 'five'. The ram found on the wreck of a small Punic warship discovered near Lilybaeum (Marsala in Sicily) was formed of timber encased on either side with a metal tusk, the whole ram curving upwards, presumably intended to puncture the enemy hull beneath the waterline. Rams were fixed to the ship's keel, but never formed a part of it, since this would have transferred too much of the force of a successful ram to the ramming ship's own hull. The other advantage of this design was that if the ram did become fixed in an enemy vessel, then it would probably break off and allow the ramming ship to withdraw.8
Ramming the bow of an enemy ship was dangerous and usually avoided, since this was the strongest part of a vessel and the resultant collision was likely to inflict serious damage to both ships. Instead, captains would manoeuvre their ships to ram the enemy's side. The ideal position was to attack from astern at a narrow angle, the ram not breaking through at one single point but rupturing a wide section of the enemy's hull, causing its seams to split and take on water. Sea battles therefore consisted of a series of individual duels as ships carefully tried to out-turn the opposition and strike from the flank, whilst trying to avoid making themselves vulnerable to another enemy, a type of combat sometimes compared to the aerial dogfights of the First World War. A highly skilled crew might choose to strike an opponent at such an angle that the ram ran along the enemy's side slicing off the ship's oarbanks and rendering them helpless, but this was difficult to achieve without damage to the attacker's own oars. Manoeuvres such as the periplus,which involved outflanking the enemy line, and the diekplus, which involved penetrating the enemy line to deliver rams from astern, cannot now be reconstructed precisely, but it is probable that they were tactics for squadrons rather than individual ships.9
The alternative to ramming was boarding, grappling the enemy vessel and overrunning it with a swarm of attackers. Success in the resultant hand-to-hand combat depended on the numbers, enthusiasm and fighting skill of the boarders compared to those of the defending marines and deck crew. As a result this method favoured the largest ships, which were able to carry more marines and also had a height advantage. Boarding placed far less demand upon the seamanship of a ship's crew whose main task was simply to bring their vessel into contact with an enemy ship and grapple it securely. Ramming required a far more highly skilled crew to perform successfully, since it relied upon speed and manoeuvrability. In the fifth century the Athenian navy had been brilliant exponents of ramming tactics, making use of their light, un-decked or aphract triremes, crewed by highly skilled rowers drawn from their poorest citizens. Few states other than the radical democracy of classical Athens were willing to pay huge numbers of rowers the regular wage needed to keep them in constant training. The Hellenistic kingdoms which emerged after Alexander were in general shorter of available manpower to provide crews than they were of the funds to construct fleets of increasingly large ships. The new emphasis on larger and larger warships diminished the importance of the ram, since such vessels were slower and less manoeuvrable and their main advantage was that they could carry greater numbers of marines. In addition to this, the hulls of the bigger ships were more strongly constructed and so perhaps less vulnerable to enemy rams, although a ram delivered by anotiier large and heavy ship was likely to cause great damage. By the third century the ram had become in effect a secondary weapon, although the well-trained Carthaginian navy were still to prove highly skilled in its use.
The crew of a quinquereme, like other galleys in the ancient world, was exceptionally large in proportion to its size, especially in comparison to the sailing ships of more recent history. The rowers who formed the majority of the crew were confined for most of a journey to their benches, since their bodily weight made up a significant part of the ship's ballast, making it undesirable for them to be allowed to move about. Galleys had very little space for the storage of food and, most important of all for rowers labouring in the heat of the Mediterranean summer, fresh water. This imposed a severe limitation on their strategic range, making journeys of more than a few days impossible for a properly crewed fleet. Ideally, ships would be drawn up on land at the end of each day to allow the rowers to rest, but beached squadrons were intensely vulnerable to attack by land or sea and this practice was unwise unless the landing could be protected by land forces. Fleets were therefore very dependent on secure bases where they could be re-supplied. Sicily and its offshore islands, and to a lesser extent Sardinia and Corsica, were ideally placed between North Africa and Italy to provide suitable staging points for each side's navies. The range of fleets was subject to further significant reduction if a major encounter with the enemy navy was anticipated, especially for fleets who relied primarily on boarding tactics. When a battle was expected it was normal to increase the number of marines carried on each ship, perhaps doubling or trebling the complement. This resulted in a much more rapid consumption of whatever supplies of food and water were carried. Even more importantly, it represented a great increase in the weight carried by a ship, drastically reducing both its speed and handling capability, problems only exacerbated if the marines were not evenly distributed and kept stationary as far as possible. Therefore, it was normal practice only to take on board the majority of marines immediately before a battle. This was not always possible and on several occasions fleets were placed at a severe disadvantage because they had failed to make contact with friendly land forces and draw marines from their ranks.