that there were cabriolets in Barletta, and, if the road was passable, I determined to adopt that mode of travelling, as I should get over the ground more quickly. I left him my passport to get the signature of the police magistrate for Canosa, and I had no doubt that he would expedite that business in order to get rid of me. This annoyance of my passport has entirely arisen from Cito, and every time I am put to the inconvenience, I am apt to bless that worthy in no very Christian spirit.

Barletta is a large city, with a population of twenty thousand, and, to a stranger, it looks to be in a prosperous state. Its streets are wide, well paved, with many handsome houses. Its cathedral is Gothic, with a lofty spire. Churches abound, and there is a theatre of considerable size. I went down to its harbour, which is formed by a pier running a good way into the sea, on which a lighthouse is erected; the port admits only small vessels, and there were very few at this time. Its chief trade is in corn, which it exports in large quantities from the plains of Apulia. There is a colossal statue of bronze, which is generally considered to be of the Emperor Heraclius (A.D. 610 to 641), or, as others think, of the Emperor Theodosius (A.D. 401 to 450). The upper part of the statue -head, arms, and breast-is of finer workmanship than the lower, and has led to the belief that some parts of it must be restored. There is an inscription at Canosa supposed to refer to this statue, which states that the inhabitants of Apulia and Calabria had erected an equestrian statue in honour of Theodosius, and that is believed to be the same statue.

On returning to the locanda, my landlord found a cabman who assured me that he knew the road to Canosa, twelve miles distant, and though the road was not good after we left the great post-road leading to Naples, he had no doubt that he could convey me safely to Canosa. My way lay across the plains of Cannæ, a spot which I could by no means pass without an examination of its appearance. As soon as my passport was procured, I mounted the cabriolet, and proceeded along an excellent road, till we reached a bridge which is thrown over the Ofanto, the ancient Aufidus, and here we turned up a by-road along the south bank of the river. The banks were without trees, and the river contained a scanty supply of water, so that I was rather disappointed to find that the poet had drawn on his imagination in his description of it. In the winter season, however, it evidently flows with greater vehemence, being swollen by the winter's torrents. Horace speaks repeatedly of its vehement character (Carm. iv. 14, 25):

Sic tauriformis volvitur Aufidus,
Qui regna Dauni præfluit Appuli,
Cum sævit, horrendamque cultis
Diluviem meditatur agris.

So branching Aufidus, who laves

The Daunian realms, fierce rolls his waves,
When to the golden labours of the swain

He meditates his wrath, and deluges the plain.

To the south of the river lay the wide plains of Apulia, as far as the eye could reach, already stripped of the grain, so early is the harvest in this part of Italy. They leave the greater portion of the stubble on the field, cutting off little more than the ear, and they afterwards set fire to

the straw, which is thus burnt on the field, and serves for manure. The practice of burning the stubble upon the lands has been handed down from the earliest times, and is followed in many other parts of Italy. It begins here in the month of July, and it is surprising, I am told, in how short a time the fire runs over a whole field of corn. They never commence except when a brisk wind is blowing, and they set fire of course to windward. The following is the description of Virgil (Georg. i. 84):

Sæpe etiam steriles incendere profuit agros,
Atque levem stipulam crepitantibus urere flammis;
Sive inde occultas vires et pabula terræ
Pinguia concipiunt; sive illis omne per ignem
Excoquitur vitium, atque exudat inutilis humor;
Seu plures calor ille vias et cæca relaxat
Spiramenta, novas veniat qua succus in herbas;
Seu durat magis, ct venas adstringit hiantes,
Ne tenues pluviæ, rapidive potentia solis
Acrior, aut Borex penetrabile frigus adurat.

Long practice has a sure improvement found,
With kindled fires to burn the barren ground.
When the light stubble, to the flames resign'd,
Is driv'n along, and crackles in the wind.

Whether from hence the hollow womb of earth
Is warm'd with secret strength for better birth;
Or, when the latent vice is cur'd by fire,

Redundant humours through the pores expire;
Or that the warmth distends the chinks, and makes
New breathings, whence new nourishment she takes;
Or that the heat the gaping ground constrains,
New-knits the surface, and new strings the veins;
Lest soaking show'rs should pierce her secret seat,
Or freezing Boreas chill her genial heat,

Or scorching suns too violently beat.

Three miles from the bridge over the Aufidus we reached the memorable field of Cannæ, where the Romans were defeated by Hannibal, and here I alighted. I was fortunate enough to meet a gentleman who addressed me in French, and who turned out to be the proprietor of the ground. He had resided three years in France in the time of Napoleon, and was kind enough to act as my guide, giving the traditions of the place. The ruins of the ancient village, which was occupied by Hannibal before the battle, are distinctly visible on a small hill about four hundred yards from the southern bank of the river, and you can trace the foundations of what seems to have been a fortress. My guide told me that excavations had been made, and that Roman coins and small images of terra-cotta had been discovered. There is a tradition that Æmilius Paulus, one of the Roman generals, died near a spring, and of course the inhabitants have fixed on the very spot where that melancholy event took place, and, stooping down, I took a refreshing draught from the "Pozzo d'Emilio" "Well of Emilius"- -as they still call it. Immediately at the foot of this hill, in an angle formed by the curvature of the Aufidus, there is a piece of ground called "Pezzo di Sangue"-" the field of blood”—and here they suppose the crisis of the battle took place. This angle of

ground of which I speak is united to the land on the north, yet has all the appearance of being traversed-as all low-lying lands on the side of rivers are in various directions, according as the water excavates its course. It is, therefore, impossible to say how the river flowed in the year B.C. 216, when the battle was fought, nor do I think that with the data before us we can decide authoritatively the point. The battle is said to have been fought on a plain, and this is the chief reason why that spot on the river is fixed on. Yet, though the character of the ground a mile down the river cannot be called a plain such as this is, yet neither is it hilly; there are merely slight eminences, sloping gently down, and they could have proved no obstacle to the movements of an army. The first question that arises in respect to the battle of Cannæ is, in what direction the Romans advanced towards the Carthaginians. Was it from the direction of Canusium, which lies about six miles from Cannæ on the same side of the river-that is, on the south side-or did they approach from the north, and reach the neighbourhood of Canna with the river Aufidus lying between them and Cannæ ? The Romans and Carthaginians, according to Polybius (iii. 107), during the winter and early spring of B.C. 216, lay, the Romans at Larinum, and the Carthaginians at Gerunium. This was between forty and fifty miles north of Canna, at a spot where the Apennines are beginning to slope somewhat down towards the plains of Apulia. The Romans were acting on the defensive, knowing that time was in their favour, and Hannibal was aware that every day he put off bringing matters to a point was lessening his chances of success. The harvest drew to an end in Apulia. I find that it is over now in a great measure, and this is towards the second week of June. Hannibal broke up his camp at Gerunium, and knowing that the Romans had collected at Cannæ large stores from the district of Canusium, which was particularly friendly, he pounced suddenly upon Cannæ, and secured the citadel of Cannæ, which was an important point, as it commanded the plains of Apulia. The city, or rather village of Cannæ, had been, we are told by Polybius, destroyed some time before. The Romans lying at Larinum did not immediately follow, as the generals sent several despatches to Rome to state what happened, and requested to know whether they were to pursue Hannibal to what they knew was the comparatively level ground of Apulia, which enabled him to bring his cavalry into full play. The armies in the field were under the command of the consuls of the former year, Cn. Servilius and M. Regulus, while the Consuls Æmilius Paulus and Terentius Varro remained at Rome to deliberate on the measures to be pursued, and to raise new levies. Servilius continued to act cautiously, and there is no reason to suppose that the army descended into the plains till the arrival of the consuls. I follow the account given by Polybius, though Livy appears to state that the consuls followed Hannibal as soon as he started for Cannæ.

Though Lucera is not mentioned in immediate connexion with these events, except as firmly attached to Roman interests, I should expect that the Roman army leaving Larinum would be encamped on these heights, the last slopes of the Apennines, before descending into the treeless flat of the Tavoliere, which they had to cross in pursuit of Hannibal.

What period of time it required to communicate with Rome and receive an answer we cannot say; but pretty nearly six weeks seem to have elapsed before the Roman troops-eighty thousand infantry and six thousand cavalry-came up with the Carthaginians. From the time the Romans began their march under the command of the consuls, they took two full days before they reached the vicinity of Hannibal, at Cannæ, and this is about the time the army might take in marching across the Tavoliere-fifteen miles to the neighbourhood of where Foggia now stands, and about the same number of miles to the vicinity of the lower part of the river Aufidus, towards the spot where the bridge spans the river, which I left on my right as I approached Cannæ.

Another point to be considered is, whether Hannibal had his troops occupying the ground round the citadel of Cannæ, which he had taken in the beginning of June, or whether he was on the opposite side of the river. Livy says that some of the fugitive Romans took refuge in the ruined city of Cannæ, and were obliged to surrender. If Hannibal's troops were in occupation of the citadel, it seems strange that the fugitives should have thought of taking refuge in the village in its immediate vicinity. This slight fact shows, in my opinion, that the battle must have been fought lower down the river than Cannæ, else the fugitives could not have come in contact with Cannæ at all, as their natural place of refuge was Canusium, six miles up the river. In none of the accounts is there any allusion made to Canusium till after the battle, nor of the army crossing the Aufidus, which they must have done if they advanced from the side of Canusium.

Besides this, an army of ninety thousand men and upwards would be sadly cramped in the narrow ground between Canusium and Cannæ, and were cut off in a great measure from its natural granary, the fertile plains of Apulia, and the towns along the coast of the Adriatic, which were still friendly to the Roman cause.

The natural and direct course for the Romans advancing from Larinum, or the neighbourhood of Lucera, would be what is now the great post-road which leads from Foggia to the bridge over the Aufidus, where I left the post-road. In those days there would be nothing more than a mere tract, or mule-path, such as we still find in every part of this country. There are no roads such as we understand, but mere paths, along which a mule may jog, but no wheel-carriage can pass along with safety. The Romans approached with caution, taking care to reconnoitre as they came near to Hannibal. They did not require to cross the river, but kept on the northern or left side. The ground on both sides of the river for a couple of miles up is comparatively level, and would be no great obstacle to an army. As you approach to the spot opposite Canna the ground rises about fifty feet above the river, but in some places slopes gently down. From the level and soft nature of the ground the river has a meandering course, having many curves, and, in some places during the winter, evidently overflows the level land on its sides. None of the curves are large, and the ground therefore enclosed is small. The largest, called Pezzo del Sangue, opposite to Canuæ, does not appear to my inexperienced eye at all capable of containing upwards of a hundred thousand men in order of battle, and yet this is the spot fixed upon as the site of the battle. I inquired of my intelligent guide,

who had been a soldier in his younger days, whether he thought that a hundred thousand men could be deployed on the small plain before us, or whether sensible men would place an army in such a position; and he confessed that it was quite out of the question. I suppose the Roman army to advance from the north, and to encamp at first at some distance from Hannibal, fifty stadia, as Polybius says. The country is described by Polybius to be plain and open, very fit for cavalry; and this description I found to be such as exactly suits its present appearance. Hannibal is lying with his army at or near the citadel of Canna. The Roman consuls are Æmilius Paulus and Terentius Varro, who command the army alternately. Varro is rash and headstrong; Emilius cautious and wary. Æmilius wishes to wait, and, by his flank position, will be able to keep Hannibal in check from getting provisions from the plains of Apulia. This is the true Fabian policy; whereas Varro is anxious for immediate action, and on his day of command advances nearer to the Carthaginians-so near, that Hannibal sends a body of cavalry to attack them. The Carthaginians are repulsed, but Æmilius, though still earnest in refusing battle, saw that it was now impossible to retreat with safety, and therefore encamped next day with two-thirds of all his forces along the Aufidus. This is the first time that the river is mentioned in connexion with these transactions; and if the Roman army had been advancing from the side of Canusium, we can scarcely imagine that the river would not have been alluded to. It must have been passed to reach Canusium, and they must have marched along its right bank to reach the neighbourhood of Cannæ. Where the Romans struck the Aufidus would be about two miles down the north side, where I found the ground to rise somewhat above the river. There I place the larger camp of the Romans. The other third he ordered to pass the river, and (observe what Polybius, iii. 110, says) to advance up the stream— ἀπὸ διαθέσεως πρὸς ἀνατολὰς—and then to entrench themselves about ten stadia, a little more than a mile, from his own camp, and about the same from Hannibal. If the Roman army had been advancing from Canusium, this body of men must have been going down the river, and not up the stream, as Polybius says.

Here, then, we have the position of the two armies lying in wait for each other, two-thirds of the Romans across the river on the north, and the main body of Hannibal at Cannæ. Hannibal harangues his troops, and says the gods had delivered the Romans into their hands by inducing them to fight on the level ground, where the Carthaginians had such an advantage. Hannibal then passes the Aufidus from Cannæ to the side where the larger camp of the Romans is placed, but it is not said how far he went down the river. The next day he allows for the refreshment of his army, and to prepare for the struggle. On the third day he offers battle, which Æmilius refuses to accept, and makes such dispositions as may secure his camp from insult. Hannibal then returns to his entrenchment, and sends a body of cavalry to fall upon the Romans of the lesser camp while fetching water from the Aufidus. Then comes the fatal 2nd of August, B.C. 216, as Gellius (v. 17, Macrob. Sat. i. 16) tells us, when the rash Varro had the command. He orders the soldiers of the larger camp to cross the river, and those of the lesser camp to join them. The ground is sufficiently level towards the great plains of Apulia to enable the largest of armies to deploy. No doubt the ground