previous next


CANNAE (Κάνναι, Strab. et al. Κάννα, Polyb.: Eth. Cannensis: Canne), a small town of Apulia on the S. bank of the Aufidus, about 6 miles from its mouth, celebrated for the memorable defeat of the Romans by Hannibal, B.C. 216. Although no doubt exists as to the site of Cannae itself, the ruins of which are still visible on a small hill about 8 miles from Canosa (Canusium), and the battle was certainly [p. 1.500]fought on the banks of the Aufidus in its immediate neighbourhood, much question has been raised as to the precise locality of the action, which some have placed on the N., some on the S. of the river: and the previous operations of the Roman and Carthaginian armies have been interpreted so as to suit either view. But if the narrative of Polybius (who is much the most clear and definite upon this question) be carefully examined, it is difficult to see how any doubt can remain, and that of Livy, though less distinct, is in no respect contradictory to it. The other accounts of the battle in Appian, Zonaras, and Plutarch afford no additional information on the topographical question.

Hannibal had wintered at Gerunium, and it was not till early in the summer that he abandoned his quarters there, and by a sudden movement seized on Cannae. The town of that name had been destroyed the year before, but the citadel was preserved, and the Romans had collected there great magazines of corn and other provisions, which fell into the power of the Carthaginians. Hannibal occupied the citadel, and established his camp in its immediate neighbourhood. (Pol. 3.107; Liv. 22.43.) The Roman generals, having received orders to risk a general engagement, followed Hannibal after some interval, and encamped at first about 50 stadia distant from the enemy: but the next day Varro insisted upon advancing still nearer, and the Romans now established two camps, the one on the same side of the Aufidus, where they previously were, (that is evidently the S. side), and the other, containing a smaller division of the forces, on the opposite bank, a little lower down the river, about 10 stadia from the larger Roman camp, and the same distance from that of Hannibal. (Pol. 3.110.) The Aufidus at this season of the year1 is readily fordable at almost any point, and would therefore offer no obstacle to their free communication.

On the day of the battle we are distinctly told that Varro crossed the river with the main body of his forces from the larger camp, and joining them to those from the smaller, drew up his whole army in a line facing the south. Hannibal thereupon also crossed the river to meet him, and drew up his forces in a line, having its left wing resting on the river, where they were opposed to the Roman cavalry, forming the right wing of the consular army. (Ib. 113; Liv. 22.45, 46.) From this account it seems perfectly clear that the battle was fought on the north bank of the Aufidus, and this is the result arrived at by the most intelligent travellers who have visited the locality (Swinburne's Travels, vol. i. pp. 167--172; Chaupy, Découv. de la Maison d'Horace, vol. iii. p. 500), as well as by General Vaudoncourt, who has examined the question from a military point of view. (Hist. des Campagnes d'Annibal, vol. ii. p. 9--34, 48--57.) The same conclusion appears clearly to result from the statement of Livy, that after the battle a body of 600 men forced their way from the lesser camp to the greater, and from thence, in conjunction with a larger force, to Canusium (22.50).

The only difficulty that remains arises from the circumstance that Polybius tells us distinctly that the Roman army faced the S., and the Carthaginian the N. (3.114): and this is confirmed by Livy, who adds that Hannibal thereby gained the advantage of having the wind, called the Vulturnus, behind him, which drove clouds of dust into the face of the enemy (22.47). There seems little doubt that the Vulturnus is the same with the Eurus, or SE. wind, called in Italy the Scirocco, which often sweeps over the plains of Apulia with the greatest violence: hence this circumstance (to which some Roman writers have attached very exaggerated importance) tends to confirm the statement of Polybius. Now, as the general course of the Aufidus is nearly from SW. to NE., it seems impossible that the Roman army, resting its right wing on that river, could have faced the S., if it had been drawn up on the N. bank, and Chaupy, in consequence, boldly rejects the statement of Polybius and Livy. But Swinburne tells us that “exactly in that part of the plain where we know, with moral certainty, that the main effort of the battle lay, the Aufidus, after running due E. for some time, makes a sudden turn to the S., and describes a very large semicircle.” He supposes the Romans to have forded the river at the angle or elbow, and placing their right wing on its bank at that point, to have thence extended their line in the plain to the E., so that the battle was actually fought within this semicircle. This bend of the river is imperfectly expressed on Zannoni's map (the only tolerable one) of the locality; and the space comprised within it would seem too confined for a battle of such magnitude: but there is no reason to doubt the accuracy of Swinburne, who took his notes, and made drawings of the country upon the spot. “The scene of action (he adds) is marked by the name of Pezzo di Sangue, the ‘Field of Blood;’ ” but other writers assign a more recent origin to this appellation.

Notwithstanding the above arguments, the scene of the battle has been transferred by local antiquarians and topographers to the S. side of the river, between Cannae and Canusium, and their authority has been followed by most modern historians, including Arnold. Niebuhr, on the contrary, has adopted Swinburne's view, and represents the battle as taking place within the bend or sweep of the river above described. (Vorträge über Röm. Geschichte, vol. ii. pp. 99, 100.) It may be added that the objection arising from the somewhat confined space thus assigned as the scene of the battle, applies with at least equal force to the opposite view, for the plain on the right bank of the Aufidus is very limited in extent, the hills on which Canusium and Cannae both stand flanking the river at no great distance, so that the interval between them does not exceed half a mile in breadth. (Chaupy, l.c.; Swinburne, l.c.) These hills are very slight eminences, with gently sloping sides, which would afford little obstacle to the movements of an army, but still the testimony of all writers is clear, that the battle was fought in the plain.

The annexed plan has no pretensions to topographical accuracy, there being no good map of the locality in sufficient detail: it is only designed to assist the reader in comprehending the above narrative.

We have little other information concerning Cannae, which appears to have been, up to this time, as it is termed by Florus, “Apuliae ignobilis vicus,” and probably a mere dependency of Canusium. [p. 1.501]


A. First camp of the Romans.

B. Second camp of the greater part of the forces; called the larger camp.

C. The smaller do.

D. Camp of Hannibal.

E. Scene of the actual battle.

F. Town or citadel of Cannae.

G. Canusium.

H. Bridge of Canusium.

K K. The Aufidus.

But its name occurs again during the Social War, B.C. 89, when it was the scene of an action between the Roman general Cosconius and the Samnite Trebatius. (Appian, App. BC 1.52.) It appears to have been at this time still a fortress; and Pliny enumerates the Cannenses “nobiles clade Romana” among the municipal towns of Apulia (3.11. s. 16). It became the see of a bishop in the later period of the Roman Empire, and seems to have continued in existence during the middle ages, till towards the close of the 13th century. The period of its complete abandonment is unknown, but the site, which is still known by the name of Canne, is marked only by the ruins of the Roman town. These are described by Swinburne, as consisting of fragments of altars, cornices, gates, walls, and vaults, in themselves of little interest. Little or no value can be attached to the name of Pozzo di Emilio, said to be still given to an ancient well, immediately below the hill occupied by the town, and supposed to mark the spot where the Roman consul perished. (Romanelli, vol. ii. p. 273; Vaudoncourt, l.c. p. 49.)


1 The battle of Cannae was fought, as we learn from Gellius (5.17; Macr. 1.16), on the 2nd of August; but it is probable that the Roman calendar was at this time much in advance of the truth, and that the action really took place early in the summer. (Fischer, Röm. Zeittafeln, p. 89.)

hide References (5 total)
  • Cross-references from this page (4):
    • Appian, Civil Wars, 1.6.52
    • Livy, The History of Rome, Book 22, 43
    • Livy, The History of Rome, Book 22, 45
    • Livy, The History of Rome, Book 22, 46
  • Cross-references in notes from this page (1):
    • Gellius, Noctes Atticae, 5.17
hide Display Preferences
Greek Display:
Arabic Display:
View by Default:
Browse Bar: