Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in the
English Historical Review
Vol. 13 (1898), pp417‑438 and 625‑642

The text is in the public domain:
Bernard W. Henderson died in 1929.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

 p417  The Campaign of the Metaurus

In England the battle of the Metaurus has attracted a much smaller amount of attention than have the other great battles of the second Punic war in Italy. This is doubtless owing to the scantiness of the topographical information which has reached us in our authorities for the battle, and in particular to the fragmentary state of the eleventh book of Polybius. Whether this be altogether a tribute to Polybius's power of description and topographical accuracy is not quite so evident. But on the Continent, in France, Germany, and especially in Italy, this battle is added to the number of the problems of the kind which harass the historian of the Carthaginian deeds in Italy. It is my purpose in this paper to point out what are the sites which contend for the honour of Hasdrubal's defeat, and what are the difficulties and assumptions involved by each. Inasmuch as the last decade of years has seen the controversy raging between German and Italian with a force unknown before, and as the storm has culminated in intensity in 1897, it may not be inappropriate for an English onlooker who has travelled up the river to describe the present position of the controversy.

I. Authorities — Ancient and Modern.

The ancient authors who describe or mention the battle are these: Polybius, XI.1‑3; Livy, XXVII.XLIII‑XLIX; Appian, 'Hannib.' 52; Frontinus, 'Strateg.' I.1.9, 2.9, II.3.8, 9.2, IV.7.15; Florus, I.22.50; Dio-Zonaras, IX.9; Ampelius, 'Lib. Mem.' 18.12, 36.3, 46.6; Eutropius, III.18.2; Valerius Maximus, VII.4.4.

The most important are the three first named. A recent theory would distinguish two traditions of the battle among these: the one, the Roman, cognisant of Nero's famous march, represented by  p418 Livy and his followers; the other, of Polybius and Appian, knowing nothing of the march. Now the reality of that march is, according as it is accepted or denied, one of the few important arguments used in the endeavour to determine one or other particular site. The would‑be champion of one or other must make up his mind on this question, the solution of which depends largely upon the acceptance or rejection of the theory just mentioned. Before a description of the sites proposed, it will be necessary, therefore, to consider this question and this theory. But, as will be seen, even a conclusive answer to this preliminary question is not conclusive for the main controversy.

The battle was fought either on the left or on the right bank of the Metaurus. To accept one or other has been enough for most modern historians. Others have particularised more nearly. From their researches it results that three sites are left seriously contending, two on the left bank, viz. the S. Silvestro and the La Lucrezia site, and one on the right bank, viz. the S. Angelo site. Their position will be clear from the maps which accompany this paper.​a

[image ALT: A map of the general area of the battle of the Metaurus.]

The modern controversy may be represented chronologically as follows:

1. 1618

Seb. Macci Durantino

De Bello Asdrubalis. Bk. III pp34‑56. Venice, 1613. Amusing and worthless. [Cf. Class. Rev., February 1898, pp11‑16.]

Many sites on both banks.

2. 1766

M. Joly de Maizeroy

Cours de Tactique. Vol. I pp403‑408. Paris, 1766.

No topographical details.

3. 1802

F. W. von Bernewitz

Leben des Hannibal. Vol. II pp197‑204. Pirna, 1802.

Right Bank.
No other topographical details.

4. 1812

F. G. de Vaudoncourt

Histoire des Campagnes d'Annibal en Italie. Vol. III pp77‑81. Pl. XXXIII. Milan, 1812.

Right Bank.
Site of La Lucrezia.

5. 1840‑41

T. Arnold

Cf. his Second Punic War, ed. W. T. A. Pp283‑290, and note O.

Right Bank.
No other topographical details.

6. 1855

H. G. Liddell

History of Rome. Vol. I pp393, 394.

Right Bank.
No other topographical details.

7. 1871

W. Ihne

History of Rome. Engl. ed. Vol. II pp388‑391.

Right Bank.
No other topographical details.

8. 1881

Th. Mommsen

Römische Geschichte. 7th edition. Vol. I pp647, 648.

Left Bank.
No other topographical details.

9. 1884

J. de la Chauvelays

L'Art Militaire chez les Romains. Pp217‑222. Paris, 1884.

No topographical details.

10. (p419: map)

F. Tarducci

Del Luogo dove fu sconfitto e morto Asdrubale. Pp22. Roma, 1888. Estratto dalla Rivista militare Italiana. Ser. III.XXXIII. Tom. II pp458‑477.

Left Bank.
Site of S. Silvestro.

11. 1889

L. Cantarelli

Rivista storica Italiana. VI fasc. 1, pp70‑72. A review of Tarducci's paper.

Left Bank.
Site of S. Silvestro.

12. 1891

Lieut.-Col. Hennébert

Histoire d'Annibal. Vol. III. pp303‑309. Paris, 1870‑1891.

Left Bank.
No other topographical details.

13. 1891

T. A. Dodge

Hannibal, pp546‑557. Plan, p554. Boston, 1891. Map vague. No names given.

All other topogr. details vague.

14. 1891

G. Bossi

La Guerra Annibalica in Italia, in the Studi e Documenti di Storia e Diritto. Vol. XII pp77‑106.

Left Bank.
Site of S. Silvestro.

15. 1896

W. W. How and H. D. Leigh

History of Rome, pp221‑223.

Right Bank.
No other topographical details.

16. 1897

R. Oehler

Der letzte Feldzug des Barkiden Hasdrubal und die Schlacht am Metaurus. Pp82 and maps. Berlin, 1897.

Right Bank.
Site of S. Angelo.

17. 1897

W. O'C. Morris

Hannibal, pp279‑284. No map. No names.

Right Bank.
All other topogr. details vague.

Besides these works, all of which I have consulted, I must mention two others which I have been unable to see:

7a 1866

C. Marcolini

Lettera al S. Conte Canonico Don Alessandro Billi. Fano, 1866.

15a? 1895

V. Pittaluga

A study of the question, discussed by Oehler in the Berl. philol. Wochenschrift. No. 9, Sp. 269, 1895.

The former suggests a second site on the right bank by Tombolina; the latter seems to be the originator of the S. Angelo site, and Oehler's treatise may well serve for both. Of the former Oehler says that die Darstellung . . . entspricht nicht den Angaben des Polybios und Livius. If this is the case or no it is impossible for me to say, as I was not able to obtain the letter. For the present paper, at least, Count Marcolini's site must be left out of account.

II. The Preliminary Data in our Ancient Authorities.

Before proceeding to describe and discuss the three contending sites of La Lucrezia and S. Silvestro on the left and S. Angelo on the right bank of the river Metaurus, it is essential to see exactly  p421 what are the data given in our ancient authorities, which may be used in such a discussion, and how the controversy has arisen. The data are not many; and for the course of events which preceded the actual battle, which is naturally all-important in determining the main question of right or left bank, Polybius fails us altogether. It is in the main to Livy that we owe the controversy and the arguments used on both sides. From his account in book XXVII I extract these topographical data:

(a) Quum in Umbria se occursurum Hasdrubal fratri scribat (XLIII.8).

(b) Ad Senam castra alterius consulis erant, et quingentos ferme inde passus Hasdrubal aberat (XLVI.4).

(c) Hasdrubal . . . suspicatus . . . id quod erat, receptui propere cecinit ac misit ad flumen unde aquabantur ubi et excipi aliqui possent et notari oculis (XLVII.2).

(d) His anxius curis, exstinctis ignibus, vigilia prima dato signo ut taciti vasa colligerent, signa ferri iussit. In trepidatione et nocturno tumultu, duces, parum intente asservati, alter in destinatis iam ante animo latebris subsedit, alter per vada nota Metaurum flumen tranavit. Ita desertum ab ducibus agmen primo per agros palatur, fessique aliquot somno ac vigiliis sternunt corpora passim atque infrequentia relinquunt signa (XLVII.8, 9).

(e) Hasdrubal, dum lux viam ostenderet, ripa fluminis signa ferri iubet, et per tortuosi amnis sinus flexusque quum errorem volvens haud multum processisset, ubi prima lux transitum opportunum ostendisset, transiturus erat, sed quum quantum a mari abscedebat tanto altioribus coercentibus amnem ripis, non inveniret vada, diem terendo spatium dedit ad insequendum sese hosti (XLVII.10, 11).

From this is appears, leaving (a) for the time on one side, that

1. The camps of Roman and Carthaginian lay at first about five hundred paces apart, near or at Sena (b).

2. A river (unnamed) was flowing close to both camps. Probably it separated them, for by advancing to this river the Carthaginians were able to see the Romans more closely (c).

3. When Hasdrubal decided to move his camp, his two guides escaped. One swam the river Metaurus by a ford and so got free (d).

4. Hasdrubal vainly searched for a ford over this river (the 'flumen' of § 10 must surely be that of § 9), advancing slowly up stream, but making little way, owing to the long windings of the stream whose course he followed (e).

Within this account it is urged there is a difficulty, an apparent inconsistency. From the Metaurus mouth to the city of Sena the distance measures some ten miles. Now presumably Hasdrubal's errant guide escaped over a river which was near at hand, and very probably this was the river which separated the camps. Whither should a deserter more naturally flee than to the enemy's camp which lay just across the river? The tone of the narrative and the  p422 probabilities of the case require that this river over which he escaped should be near at hand, and not several miles away to the rear of both armies. But this river over which he escaped was, it is certain, the Metaurus. Therefore the camps lay near the Metaurus. Almost certainly this was the river unde aquabantur, and thus the Metaurus separated the hostile camps, Hasdrubal being on the left, the Romans on the right, bank. The unnamed river of (2) must be identified as a result of this guide difficulty with the river of (3), which is the Metaurus.

Yet the camps lay near Sena, some ten miles away, far south of the Metaurus. This is the difficulty which has caused, or ought to have caused, at least some considerable part of the whole controversy. If the camps lay by Sena, Hasdrubal retreated to the Metaurus and was caught on the right bank, being unable to cross the stream. If it be asked, 'How is this probable, seeing that he had just recently crossed it by a ford on his march south, and surely he would be able even in the absence of his guide to recross it?' it is answered that the river must have been in flood at the time. If, again, to this it be objected that his runaway guide did cross it, and how is this consistent with the suppositional flood, it is answered that though the guide knew the ford, yet even he had to swim across it, as Livy expressly says. And in any case the darkness may have made Hasdrubal miss it at first. Thus when day dawned he had wandered away from it, and continued his vain wanderings up stream, instead of returning to look for the ford he had crossed once, and thus knew already.

If, on the other hand, the camps lay on the Metaurus, then Hasdrubal, always hoping to force his way into Umbria to join his brother, as we are expressly told (a), intended to cross the river southwards, marching not by the coast road beyond Fano — which would have led him to Apulia — but up the Metaurus valley, and so across it into Umbria. He therefore advanced up stream, looking for a ford over the river which would lead him south into Umbria. His guides, however, deserted him, one swimming the river which lay close at hand, and thus escaping successfully. Thus Hasdrubal was overtaken and killed on the left bank. And if it be objected that in this case it is impossible to say the camps lay first ad Senam, it is answered that this is a mere geographically loose expression, denoting adequately the country of the Senonian Gauls. It is, of course, undeniable that ad Senam is definitely stated to be the city of Sena in later authors, viz.:

Eutropius, III.18.2: is [i.e. Hasdrubal] apud Senam Piceni civitatem in insidias compositas incidit.

Appian, loc. cit.: Καὶ ἀντεστρατοπέδευσαν αὐτῷ περὶ πόλιν Σένας.

Dio-Zonaras, loc. cit.: ὃς [i.e. Livius] αὐτῷ [i.e. Hasdrubal] πρὸς τῇ Σένᾳ τῇ πόλει ἀπήντησεν.

 p423  These, however, it is argued, have but imposed their own view of the situation upon what they found in Livy or Livy's sources. And against them we may set another of the same class, viz.:

Valerius Maximus, VII.4.4: Salinator in Umbria apud Metaurum flumen proximo die dimicaturus, summa cum dissimulatione Neronem castris noctu recepit:

showing he believed the first camps to have been on the Metaurus. Yet it must be confessed that the strain put on the words of ad Senam by this interpretation is very great.

According, therefore, as we feel (1) the Sena difficulty, or (2) the guide and river difficulty, we select one or other bank of the river on which to search for some appropriate site suitable to the description of the actual battle given in our authorities. But first it is important (III) to see the bearing on the question of one great and preliminary problem, and then  (IV) to discover the requisites of the actual battle-field, as stated in our authorities, before proceeding finally (V) to describe and discuss the three sites chosen.

III. Nero's March.

Whether the Roman tale of Nero the consul's march of 'wondrous swiftness' (as Valerius Maximus says) from South Italy to join Livius the consul in North Italy be true or not, is itself a hotly disputed point. And it is nearly concerned with the Metaurus controversy. For if we allow some truth in the tale, it is evident that the more to the south Livius's camp lay, the less impossible would seem Nero's feat. Those who place the camps near Sena, and the battle therefore on the right bank of the river Metaurus, have less difficulty in accepting the tale of the march than those who place the camps on the Metaurus and the battle on the left bank. Similarly the higher up the river the battle is placed the greater grows the endurance with which we must needs endow, if we accept the tale, Nero's weary soldiery. The consequence of this is natural. Champions of the S. Silvestro site are wellnigh bound either to reject the whole tale as false, as a patriotic fairy tale, an invention (for obvious reasons) of the early principate,​1 and so forth, or at least to extend the limit of time and thereby destroy much of the marvel of the tale. In actual fact they are more inclined to choose the former alternative.

One order of the argument in this discussion it is surely most essential to avoid. We cannot be justified in arguing (as I fear both in this and other controversies those inquirers who love an a priori probable site in view of the appearances of the ground or local tradition do argue) from our battle to the possibility of the march. We have not enough independent material for the choice of the site to allow  p424 us to act thus. It is necessary to argue from the possibility of the march to the site.

This then is Livy's narrative of the march.2

The consuls of the year were C. Claudius Nero and M. Livius (XXXVI.10). The danger threatening Rome was twofold: on the north Hasdrubal was expected to come down on Italy over the Alps; on the south Hannibal must be kept busily occupied in fighting, lest he should be able to force his way out of Bruttium and march north to join his brother (XXXVIII.6, 7). The senate allowed the consuls full discretion as to preparations against both these dangers.

Senatus liberam potestatem consulibus fecit et supplendi unde vellent, et eligendi de omnibus exercitibus, quos vellent, permutandique, et ex provinciis, quout e re publica censerent esse, traducendi (xxxviii.9).

While the levy was in progress letters from L. Porcius, praetor in Gaul, increased the excitement. Hasdrubal had crossed the Alps. The Ligurians would join him unless attacked first. He would do his best to advance against the foe, but his army was but weak.

Hae litterae consules, raptim confecto dilectu, maturius quam constituerant exire in provincias coegerunt, ea mente ut uterque hostem in sua provincia contineret, neque coniungi aut conferre in unum vires pateretur (XXXIX.8).

Hence the consuls left the city, velut in duo pariter bella, Claudius Nero to face Hannibal; Livius, Hasdrubal (XLI.1.10, XLVI.4). Before Nero's arrival Hannibal had been harassed by two Roman armies in South Italy under C. Hostilius Tubulus and Q. Claudius, now propraetor (XL.10, 11; cf. XXI.6). When the consul Claudius Nero arrived to take command, he found Hannibal had emigrated from Bruttium and was encamped outside Grumentum. Here he inflicted a severe defeat on the Carthaginian. The latter retired northwards into Apulia, but was overtaken by Claudius Nero and again defeated near Venusia. He retired therefore to Metapontum; but, reinforced there by Hanno, returned to Venusia and thence marched yet further north to Canusium, Nero clinging close to him all the time (xlii).

Meanwhile Hasdrubal was wasting precise time in a vain attempt to take Placentia (XXXIX.10‑12). Desisting at last from this siege, he then sent six horsemen with letters to Hannibal. These made their way safely down through the heart of Italy, it seems, to Metapontum. Finding Hannibal had left that city, they attempted to follow him, but missed their way, and, captured close to Tarentum by a body of Roman scouts, were carried off to the propraetor Q. Claudius. He sent them and their letters on to the  p425 consul Claudius Nero under convoy of two Samnite troops of horse (XLIII.1‑5). Nero on reading the letters took instant decision. This was not the time for a general to observe strictly the limits of the province assigned him by the senate. He sent Hasdrubal's letters on to the senate at Rome, and explained at the same time his own plan. Hasdrubal had written to his brother he would meet him in Umbria: Quum in Umbria se occursurum Hasdrubal fratri scribat. The senate therefore must recall the legion then at Capua to Rome, hold a levy in Rome, and post this city army at Narnia to face the foe (XLIII.5‑9). This order they obeyed (cf. L.6).

Haec Senatu scripta. Praemissi item per agrum Larinatem, Marrucinum, Frentanum, Praetuttianum qua exercitum ducturus erat, ut omnes ex agris urbibusque commeatus paratos militi ad vescendum in viam deferrent, equos iumentaque alia producerent, ut vehiculorum fessis copia esset (xliii.10).

This done, Nero next selected from his whole army six thousand foot and a thousand horse, the flower of the whole, and left his camp at Canusium secretly by night with these, already warned to be in marching order. In charge of the camp he left the legate Q. Catius. His own troops were at first told it was intended to march on the nearest Punic city in Lucania. But Nero led them straight for Picenum, and as soon as he was at a safe distance from the enemy's camp revealed his real purpose to them (XLIII.11, 12, XLV.1‑9).

Then they hurried north quantis maximis itineribus to join Nero's colleague Livius (XLIII.12), carrying scarcely anything but their weapons (XLVI.2). The peasant hailed them as deliverers.

Invitare inde pro se quisque et offerre et fatigare precibus ut quae ipsis iumentisque usui essent ab se potissimum sumerent; benigne omnia cumulata dare. Modestia certare milites, ne quid ultra usum necessarium sumerent; nihil morari, nec abscedere ab signis, nec subsistere nisi cibum capientes; die ac noctu ire; vix quod satis ad naturale desiderium corporum esset quieti dare (xlv.10, 11).

Messengers were sent on before the main body to warn Livius of his colleague's approach. By their means it was decided it was better for Nero to enter his colleague's camp, then pitched ad Senam opposite Hasdrubal, secretly under cover of darkness (XLV.12). Itaque, quum iam appropinquaret, tectus montibus substitit Nero, ne ante noctem castra ingrederetur (XLVI.4). At night the troops entered Livius's camp unobserved by the foe. A council of war was held next day, at which the praetor L. Porcius Licinus was present. He had spent the time before Livius's arrival in harassing Hasdrubal's advance, and was now encamped beside the consular army. It was decided to give immediate battle (XLVI.5‑12). This Hasdrubal, discovering Nero's arrival, declined, and the following night retreated. The next day the Romans pursued after him and  p426 overtook him. He was defeated and slain (47‑49). Among the victors there was no delay: Nero ea nocte quae secuta est pugnam profectus citatiore quam inde venerat agmine, die secto ad stativa sua atque ad ostem pervenit (L. 1). Hannibal had never known of Nero's absence, and the head of his brother Hasdrubal hurled into the middle of the Punic outposts gave him the first news of the battle of the Metaurus, which was the Fate of Carthage (L.11, 12).

The tale of the march is given besides by Frontinus, Valerius Maximus, and Zonaras. Frontinus​3 gives no hint of the time taken in the march. Claudius takes with him 10,000 men, and meets his colleague in Umbria occultatis itineribus. Valerius Maximus (VII.4.4) speaks only of how Nero . . . ad opem collegae ferendam per longum iter celeritate mira tetendit. Salinator in Umbria apud Metaurum flumen proximo die dimicaturus summa cum dissimulatione Neronem castris noctu recepit. Zonaras (ix.9) gives a fair summary, but with no marks of time:

Καὶ ὁ Νέρων . . . δείσας μὴ τὸν Λιούιον ὁ Ἀσδρούβας τῷ πλήθει καταβιάσηται, μέγα πρᾶγμα ἐτόλμησε. Καὶ κατέλιπε μὲν μοῖραν ἐκεῖ ἀποχρῶσαν εἴργειν τὸν Ἀννίβαν, εἲ πῃ κινηθείη, ἐντειλάμενος πάντα ποιεῖν ἴνα καὶ αὐτὸς νομίζοιτο ἐνδημεῖν, τὸ δὲ καθαρώτατον τοῦ στρατοῦ ἀπολέξας ὤρμησεν ὡς πόλει τινὶ πλησιοχώρῳ προσμίξων, οὐδ’ ᾔδει τις τὴν διάνοιαν αὐτοῦ. Καὶ ἠπείχθη ἐπὶ τὸν Ἀσδρούβαν, καὶ ἀφίκετο νυκτὸς πρὸς τὸν συνάρχοντα, καὶ ἐν τῇ ταφρείᾳ τῇ αὐτοῦ κατεσκήνωσε.

The attitude to the question of Polybius and Appian is discussed later. Thus the dispute really rages round Livy's tale alone.

In Livy's account, then, we see that the only hint as to the time taken on the march is contained in the words die sexto. As Nero left to march back in the night, it is possible surely to count the day after the battle as the first of the six days. One day is spent in the battle. The day before in the camps ad Senam. As Nero's march back was even swifter than his march north, it is fair to allow seven days, but no more, for this last. He begins this on the night before the first of the seven days, and arrives ad Senam before nightfall on the seventh.

Taking then, as is only fair for the argument's sake, the greatest length of time justified by the words of the Livian narrative, we arrive at these results:

1. March from Canusium to join Livius's camp 7 nights and 7 days
2. Rest in camp 2 nights and 1 day 
3. Pursuit and battle 1 day 
4. March from battlefield to Canusium 6 nights and 6 days
Total from Canusium to Canusium 15 nights and 15 days

 p427  The distances are these:


Canusium to the River Cesano



English miles


Canusium to Sena Gallica



"      "


the Cesano to the Metaurus


"      "


Canusium to the Metaurus



"      "


Fano to La Lucrezia



"      "


Metaurus mouth to S. Angelo



"      "


Fano to S. Silvestro



"      "

The distances therefore of the three sites are:

From Canusium to

S. Angelo

site and back


English miles

 "     "     "

La Lucrezia

"        "


"     "

 "     "     "

S. Silvestro

"        "


"     "

From these tables it appears that on the easiest supposition (that of the camps on the Cesano and the S. Angelo site) Nero's troops marched over 230 miles in seven nights and days. Any site on the left bank involves an addition of about eight miles to be accomplished in the same time.

Also the return march


S. Angelo

site involves


miles in 6 days and nights


La Lucrezia

"        "


"         "         "


S. Silvestro

"        "


"         "         "

It may doubtless be argued that the tale of the return march in quicker time may easily be rejected, while we yet retain a belief in the week's march north. The addition is so picturesque a touch, and so easily devised. But even if we allow this, it is clear that a site on the left bank as far up as S. Silvestro involves practically either the rejection of the whole tale, or of the 'week's march' at least. If we may retain Livy's whole story, it by itself almost compels us to reject the S. Silvestro site, and implies a slight preference for the right bank over the left bank generally.

What then is urged against Livy's whole story? Bossi, Oehler, and Pittaluga combine to represent it as an annalistic invention — a mere patriotic flourish. The reasons for this are, in the main, the improbabilities they discover in Livy's account. To this Bossi adds the argument that there is another and a better tradition of the events of this year which knows nothing of the march, and is represented in our authorities by Polybius and Appian.

A. Improbabilities in Livy's Account.

These are the improbabilities which have been discovered in Livy's account of the march:

1. Livius would already be well informed of Hasdrubal's movements and intentions, and would inform both the senate and his colleague concerning them. The tale of the intercepted messengers is therefore unnecessary.

2. Nero's advice to the senate on reading Hasdrubal's letters  p428 is, if true, quite gratuitous. Therefore it is untrue. O sì! Il Senato aveva proprio bisogno di saperli da Nerone i provvedimenti ch' esso doveva adottare, esso che da Roma vegliava al generale andamento della guerra.

3. Nero need never have told the senate at all of his intention to join his colleague. For in anticipation of some such necessity the senate at the beginning of the consular year had already given him express permission to do so — as Livy (XXXVIII.9) says.

4. Hasdrubal left Placentia before sending the messengers. These wandered as far as Tarentum, were captured and sent to Canusium. Then Nero sends his message to the senate, and begins his march. He finds Hasdrubal arrived, perhaps as far as Sena, but certainly no further. There was no time sufficient, therefore, for all these intermediate events.

5. It is impossible that Hasdrubal's messengers could have reached safely as far as Tarentum. They must have been captured by the enemy long before.

These five arguments are in the main worthless, and do but cumber the ground. Thus it may be argued that the principles of criticism implied in nos. 1 and 5 would go far to destroy any historical narrative at all, no matter how good its authority. With regard to the others, no. 2 is similarly futile. Nero, according to the story, gave first two pieces of news to the senate, viz. that Hasdrubal meant to join Hannibal in Umbria, and that he intended himself to hurry north. He added a most important piece of advice, that the senate should therefore not rely wholly on the consular armies, but constitute a second line of defence at Narnia. How is any single item of all this gratuitous? Could the senate in view of its 'general supervision of the war' therefore read the mind and discover the intentions of both the leader of the foe in N. Italy and of their own consul in the south? To no. 3 there are two obvious answers. In the first place Livy's words, Senatus liberam potestatem consulibus fecit et supplendi unde vellent et eligendi de omnibus exercitibus quos vellent, permutandique, et ex provinciis, quo e republica censerent esse, traducendi, apply clearly to the original levy of the legions. And even if they may also be understood to apply to the transfer of troops at any time during the consular year from the one provincia to the other, this is a very different thing from the self-transferring of one consul from his provincia to that of his colleague. And in the second place, even granting what seems to me unlikely, viz. that Livy's words do imply this last permission, even so surely there was the greatest need, the most stringent military necessity, for Nero to inform the senate of his purpose at that particular time, when he had determined it was necessary to make use of the permission once granted him in general terms by  p429 that body. The time argument (4) is worthy of its predecessors. To it we may reply first that Nero's urgent haste was caused precisely by the knowledge of the exigencies of time; and secondly, that the time intervening between Hasdrubal's march from Placentia and Nero's arrival in his colleague's camp was sufficient for the events narrated by Livy as having taken place in the interim. For we know that Hasdrubal's march south was continually harassed, and therefore delayed, by skirmishes with the praetor L. Porcius Licinus, who

per loca alta ducendo exercitum, quum modo insideret angustos saltus, ut transitum clauderet, modo ab latere aut ab tergo carperet agmen, ludificatus hostem omnibus artibus belli fuerat.​4

Then, finally, the way was barred by the consul Livius, and we are expressly told that the two armies lay over against one another waiting many days before Nero's arrival. For Livius

αὐτῷ [i.e. Hasdrubal] πρὸς τῇ Σένᾳ τῇ πόλει ἀπήντησεν, οὐ μέντοι καὶ εἰς χεῖρας εὐθὺς ἠλθον. ἐπὶ πολλὰς δὲ ἡμέρας κατὰ χώραν ἔμεινεν. ὰλλ’ οὐδὲ ὁ Ἀσδρούβας τὴν μάχην κατήπειξεν, ἡσύχαζε δὲ τὸν ἀδελφὸν ἀναμένων.​5

Time, then, there was ample for the intervening events. These objections to Livy's narrative may lightly be disregarded. So also another objection is small and trifling indeed, viz.:

6. The districts mentioned in c. xliii.-10 are arranged out of order. Instead of per agrum Larinatem, Marrucinum, Frentanum, Praetuttianum, the due order from S. to N. is Larinum, Frentani, Marrucini, Praetuttii. Surely such an argument ἀπηλγηκυίας ἐστὶ ψυχῆς.

The ground is now cleared of the useless weeds of a luxuriant scepticism. There remains the one very real difficulty of Livy's account:

7. The march as described is of impossible rapidity. Choosing the site which makes the lowest estimate possible, we are told that an army of 6,000 foot and 1,000 horse marched 230 odd miles in 7 days and nights. It is perfectly true that no measure was left undevised to assist them. They carried no impedimenta save their arms only. The way was prepared before them. Food was brought down to the roadside. Horses and cattle in relays were ready to help the weary. The men were the very flower of the Roman army chosen to face a Hannibal. The country through which they marched was not only friendly but enthusiastic in their cause, and the people poured down to encourage and exhort. The men knew well the overwhelming importance of their mission and the need for haste. The road was a great main thoroughfare, level, for the most part of it lay by the seashore, and running straight without windings — no small encouragement to the temper and  p430 therefore speed of any hurried traveller, as every pedestrian knows. But when all is said and done, it is argued that no army under the most favourable conditions (which must be in this case conceded) could as a matter of physical possibility march 230 miles in 7 days and 7 nights, or an average of 33 miles in the twenty-four hours for a week.

Thus Bossi points out that the modern soldier under great pressure marches at most thirty kilometres (= 19 English miles) a day. Caesar in 58 B.C. for all his urgent need of haste took four weeks marching from Aquileia to Lyons, a distance of about 600 kilometres, making an average of 20 to 25 kilometres per day (about 14 English miles). Vegetius (I.9) says that the soldiers of the empire in extraordinary marches could accomplish 24 miles a day (= about 22 English). No theory of degeneration can explain the discrepancy of these statistics with the 33 miles a day of Nero's republican troops. If the march actually did take place, Nero must have taken a month's absence from S. Italy at least. Hannibal therefore must needs have discovered his absence and marched north to meet his brother. Tutte queste inverisimiglianze da me notate nel racconto liviano mi fanno ritenere impossibile la marcia di Nerone da Canusia al Metauro. Pittaluga and Oehler agree at least in this. Blüme​6 is invoked as witness to the position that no troops can march for consecutive days more than three to four German miles (= 14 to 18½ English) without undue fatigue. Night-marching too is wellnigh fatal to their discipline and temper. It is essential that every three marching days be followed by one day of rest. Thus Nero's force must have taken at least 15 days, though in the very best of health and spirits, to accomplish the 230 miles, viz. 12 marching days of 19 miles each with three intervening days of rest. We seem reduced to one of two alternatives. Either the whole tale of the march is a fiction, or the tale of 7 days and 7 nights at least is hopelessly inaccurate. And if we are inclined to choose the latter and believe that Nero made the best speed he could, then it is all the harder to explain Hannibal's long-continued inactivity and negligence. And of course in this case the whole question of the masonry can throw no light on the site controversy. This, however, is in this connexion of no importance.

Never, I suppose, will this question of the possibility of 33 miles a day for 7 days in 207 B.C. be settled beyond dispute. For myself I cannot feel the improbability of the tale so deeply, It is true that it is hard to discover in records ancient or modern a greater or even an equal feat of marching, and this march is indubitably the 'best on record' if we consider not only the number  p431 of miles accomplished per day, but also the number of days of continuous marching. But the discrepancy between this and other marches does not seem to me sufficiently great, bearing in mind all the noteworthy circumstances of the case, to justify us in pleading the a priori argument of physical impossibility. The following are the best marching feats of other armies. I owe them (with the exception of the last-named of the 'ancient marches') to Colonel Dodge's useful appendices to his Lives of Caesar and Gustavus Adolphus.

Ancient Marches

Event No. of Troops Miles Days Rate per Day
Spartans to Marathon 2,000 150


Macedonians for capture of Bessus 6,000 150


Macedonians: Jaxartes to Maracanda 15,000 170

Romans: Samarobriva to relief of Cicero (N.B.: winter roads) 8,000 110


Romans: Gergovia to Aeduan army and back 16,000 50

24 hours

Romans: Asparagium to Dyrrachium 21,000 45

26 hours

over 41    
Romans: Ebro to New Carthage (Polyb. X.9.7) 25,000 foot
2,500 horse

7 days

Romans: Canusium to ? Sena 6,000 foot
1,000 horse

7 days


Nero's feat is thus put into the shade by Scipio's Spanish march. But Polybius's Spanish geography is scarcely accurate enough in general to allow us accept as true this really incredible feat, nor does Polybius here seem to realise what marching powers he has ascribed to Scipio's soldiery. But putting this on one side, it is evident from the other data that Nero's march does not occupy a position of such overwhelming superiority over all other marches as some perhaps are apt to imagine.

The record of modern marches is less good.

Event No. of Troops Miles Days Rate per Day
A.D. 1657 Turenne: Scheldt to Lys 30,000 75 3 25
1644 Condé: Moselle to Freiburg 10,000 210 13 16
1632 Gustavus: Donauwörth to Naumberg
(N.B. bad roads)
20,000 270 18 15

The armies here, however, are larger, and probably more encumbered with artillery &c. and by bad roads, than the 7,000 picked troops of Nero's force.

On the whole it is fair to insist once more on the extraordinarily favorable conditions of everything concerned with Nero's march,  p432 on the urgent need for haste, the directness and ease of the road, the enthusiasm and excellent condition of the force. In view of these facts, which all will concede, and the evidence of other marches, it is surely justifiable to refuse to reject the great tale of 33 miles a day for 7 days, regarding it as a physical impossibility, and therefore the story as an annalistic fairy tale. Even to‑day place an invading army at Lancaster face to face with a British force, and it is hard to believe that 7,000 picked troops could not be induced to march from London to the rescue, covering the intervening 230 miles in seven days and nights, and then fight after a rest of 2 nights and one day.

B. The 'Other Tradition.'

The second and only other serious argument against Livy's tale is that suggested first, so far as I know, by Gaetano Bossi. Briefly it is this, that side by side with the annalistic tale of the events of 207 B.C. there existed quite another and a more trustworthy tradition, represented to us by Polybius and Appian. Thus the many tales in Livy of Hannibal's successive details in South Italy in the chapters xl-xlii of book XXVII are at once inconsistent not only with probability but also with Polybius's direct assertion that, up to Zama, Hannibal had always been ἀήττητος (XV.16.5). Livy's chapters are almost certainly derived from the annalists, those men of patriotic figment. Bossi ascribes them to Valerius Antias; Soltau​7 to Claudius Quadrigarius.

Similarly Bossi argues that the tale of Nero's march is peculiar to the annalistic tradition. The other tradition knew nothing whatever of it. In both Polybius and Appian it finds no place. Both consuls, according to this, originally proceed to the north to oppose Hasdrubal's progress. Hannibal is kept in check in South Italy, not by the consul Claudius Nero, but by the propraetor Q. Claudius Flaminius. It is true that both consuls fought Hasdrubal, but untrue that Nero accomplished his famous march from the south for this purpose, inasmuch as he was already in the north.

Appian's words, on which in the main this theory is based, are contained in his Ἀννιβαΐκη, c. lii:

Ἀσδρούβας . . . ἐσέβαλέ τε ἐς Τυρρηνίαν . . . καὶ γράμματα πρὸς τὸν ἀδελφὸν ἔπεμπε δηλῶν ὅτι παρείη. Τούτων τῶν γραμμάτων ὑπὸ Ῥωμαίων ἁλόντων, οἱ ὕπατοι Σαλινάτωρ καὶ Νέρων μαθόντες αὐτοῦ τὸ πλῆθος τῆς στρατιᾶς ἀπὸ τῶν γραμμάτων, συνῆλθον ἐς τὸ αὐτὸ πάσαις ταῖς δυνάμεσι καὶ ἀντεστρατοπέδευσαν αὐτῷ περὶ πόλιν Σήνας.

Secondly, Polybius, it is argued, also contained this view of events. True, his account of everything which preceded the actual battle is most unfortunately lost. But he does not mention Nero's  p433 return march. The inference, therefore, should be drawn that he knew nothing of the march to Sena; that his lost account of events preceding the battle placed both consuls from the first in North Italy as Appian is supposed to do; and that therefore we have all the weight of Polybius's authority as well as Appian's for rejecting the tale of the march. The annalists confounded the commander of the Roman retaining army in the south, Q. Claudius Flaminius, with the consul Claudius Nero. Yet it was well known that both consuls fought Hasdrubal. Hence the necessity for the invention of the march, which found no place, however, in the best tradition of the events of 207 B.C.

I cannot think this alleged confusion of names very happy as an argument or very likely as a fact. Livy — who surely according to the argument represents these much-misled annalists — distinguishes very carefully between the two Claudii (XXVII.XLIII.1‑5). It is therefore far more likely that his authorities were similarly clear. And if these were not the annalists, but Polybius, then the whole structure of this part of the argument of course breaks down at once. Let us also consider more nearly the arguments concerning the accounts of Polybius and Appian.

1. Polybius. — As it stands the argument is ingenious rather than convincing. It is based simply on Polybius's silence as to Nero's return march. Therefore it labours under this capital defect, viz. we have no right whatever to suppose that this was not mentioned in the part of Polybius subsequent to the battle, also unfortunately lost. The extant passage forming the beginning of his eleventh book begins with Hasdrubal's battle formation just before the actual engagement. Then follow the Roman formation, the battle, the death of Hasdrubal, and a long and eloquent panegyric on the dead chief as a valiant soldier and a wise general. Then the Roman collection of the plunder and a summary of the losses on both sides. Finally (c. 3 §§ 4‑6) comes a description of the reception of the news in Rome, and with that the account breaks off. There is not one word of any military movements after the victory. All this is lost to us. We surely have not the least right in the world to assume that Polybius did not mention one particular event in those subsequent movements because his entire account of all those subsequent movements is lost. And still more certain is the wrong of so doing when it is clear that Livy consulted Polybius (cf. infra) and Livy's account of Nero's return march is itself subsequent to his account of the reception of the news in Rome; cf. XXVII.l, li.1‑10 (reception of the news); LI.11‑13 (the return march and Hannibal's movements). This argument, which alone serves to bring Polybius into line with Appian, and makes him directly contradict Livy, seems to me not worth the ink with which it was first written by its over-ingenious deviser.

 p434  Further, it is undoubted that Livy did at least consult Polybius for his history of the events of this year. Bossi himself is perfectly ready to admit this. 'In some parts of the narrative,' he says, 'I find not only a great similarity, but an almost perfect identity between the two accounts.' As illustrations of this, we may set side by side the two accounts of the reception of the news in Rome:

Primo magis auribus quam animis id acceptum erat, ut maius laetiusque quam quod mente capere aut satis credere possent. . . . Ipsos deinde appropinquare legatos . . . L. Veturius . . . ipse planius omnia quae acta erant exposuit cum ingenti assensu, postremo etiam clamore universae contionis quum vix gaudium animis caperent . . . Omnia templa per totum triduum aequalem turbam habuere . . . Statum quoque civitatis ea victoria movit ut iam inde haud secus quam in pace res inter se contrahere . . . auderent.

Τῆς δὲ φήμης ἀφικομένης εἰς τὴν Ῥώμην τὴν μὲν ἀρχὴν ἠπίστουν τῷ λίαν βούλεσθαι τοῦτο γενόμενον ἰδεῖν· ἐπειδὴ δὲ καὶ πλείους ᾑκον οὐ μόνον τὸ γεγονὸς ἀλλὰ καὶ τὰ κατὰ μέρος διασαφοῦντες, τότε δὴ χαρᾶς ὑπερβαλλούσης ἦν ἡ πόλις πλήρης, καὶ πᾶν μὲν τέμενος ἐκοσμεῖτο, πᾶς δὲ ναὸς ἔγεμε πελάνων καὶ θυμάτων, καθόλου δ̓̓’ εἰς τοιαύτην εὐελπιστίαν παρεγένοντο καὶ θάρσος ὤστε πάντας τὸν Ἀννίβαν, ὃν μάλιστα πρότερον έφοήθησαν, τότε μήδ’ ἐν Ἰταλίᾳ νομίζειν παρεῖναι.

And the 'Laudatio Hasdrubalis:'

Hasdrubal . . . dux quum saepe alias memorabilis, tum illa praecipue pugna. Ille pugnantes hortando pariterque obeundo pericula sustinuit; ille fessos abnuentesque taedio et labore nunc precando nunc castigando accendit; ille fugientes revocavit omissamque pugnam aliquot locis restituit; postremo, quum haud dubie fortuna hostium esset, ne superstes tanto exercitui suum nomen secuto esset, concitato equo se in cohortem Romanam immisit. Ibi, ut patre Hamilcare et Hannibale fratre dignum erat, pugnans cecidit.

Ἐν πᾶσι τοῖς εἰρημένοις καιροῖς ἀξίως τοῦ πατρὸς Βάρκα καλῶς δὲ καὶ γενναίως τὰς περιπετείας καὶ τὰς ἐλαττώσεις διετέλει φέρων . . . Ἀσδρούβας δ’ ἔως μὲν ἧν ἐλπὶς ἐκ τῶν κατὰ λόγον τοῦ δύνασθαι πράττειν ἄξιόν τι τῶν προβεβιωμένων, οὐδενὸς μᾶλλον προενοεῖτο κατὰ τοὺς κινδύνους ὡς τῆς αὐτοῦ σωτηρίας· ἐπεὶ δὲ πάσας ἀφελομένη τὰς εἰς τὸ μέλλον ἐλπίδας ἡ τύχη συνέκλεισε πρὸς τὸν ἔσχατον καιρὸν οὐδὲν παραλιπὼν οὔτε περὶ τὴν παρασκευὴν οὔτε κατὰ τὸν κίνδυνον πρὸς τὸ νικᾶν, οὐχ ἧττον πρόνοιαν εἶχε καὶ τοῦ σφαλεὶς τοῖς ὅλοις ὀμόσε χωρῆσαι τοῖς παροῦσι καὶ μηδὲν ὑπομεῖναι τῶν προβεβιωμένων ἀνάξιον.

That Livy for his account of the Hasdrubalic invasion used patriotic and exaggerating sources as well as Polybius, is of course undeniable. Thus, while Polybius gives the losses on both sides as 'not less than 10,000 Carthaginians and Celts, and about 2,000 Romans' (XI.3.3), in Livy's narrative both numbers,  p435 as we are bound to expect, are greatly magnified. Of the enemy, 56,000 are slain and 5,400 taken, while the Roman dead with those of the allies number about 8,000 (XLIX.6, 7).

Again, a slighter discrepancy between the two is that, according to Polybius, Hasdrubal himself took command not of the right wing, but of the centre: μέσον αὐτὸν θεὶς τῆς παρατάξεως κατὰ τὴν τῶν θηρίων προστασίαν (XI.1.3); whereas Livy says of Hasdrubal: Ipse dextrum cornu adversus M. Livium sibi atque Hispanis . . . sumpsit; Ligures in medio post elephantos positi (XLVIII.6, 7). This is, however, but a small matter. For Polybius himself says that Hasdrubal attacked the left wing of the foe (XI.1.3), and also that Nero, having accomplished his flanking movement, προσέβαλε κατὰ κέρας τοῖς Καρχηδονίοις ἐπὶ τὰ θηρία (XI.1.7). That is, the Carthaginian right wing was quickly merged with the centre, and the elephants belonged to both. They form together one of the two halves of the Punic array, and here Hasdrubal took his station. Livy in like manner describing Nero's flank attack couples together Ligurians and Spaniards: Ita ex omnibus partibus, ab fronte, ab latere, ab tergo, trucidantur Hispani Liguresque (XLVIII.15). Livy's invention, in fact, of a Carthaginian centre distinguished from the right wing is an unnecessary refinement, and may not justly be quoted as a serious discrepancy from Polybius, nor used to demonstrate that Livy's tradition of the battle differed notably from that followed by Polybius.

Another discrepancy is that while Livy says of Hasdrubal's line of battle, Sed longior quam latior acies erat (XLVIII.7), Polybius expressly says of him, Καὶ τὸ βάθος αὐξήσας τῶν τάξεων, καὶ ποιήσας ἐν βραχεῖ χώρῳ τὴν ὅλην δύναμιν (XI.1.3). Vaudoncourt calmly proposes here to understand by longitudo aciei the depth of the line (vol. III p79). If, however, we feel some hesitation in following him in this, the discrepancy remains. But is it of such moment as to justify a theory of opposed Livian and Polybian traditions?

To sum up: the grounds upon which it is urged that Polybius and Livy represent two distinct traditions of the campaign of 207 B.C., the one inventing the Neronian march of which the other justly knows nothing, I must think poor, insufficient, and also opposed to probability. That Polybius, had all his narrative been preserved, would have mentioned, though doubtless without certain annalistic embellishments, the march of Claudius Nero from Canusium to Sena, seems to me a more likely supposition than Bossi's theory of the 'Other Tradition.'

2. Appian. — Only Appian, therefore, is left to represent this 'Other Tradition.' His words, quoted above, prove clearly, according to Bossi, that his authorities knew nothing of the Neronian march. It can't be said that Appian's account of the events of the  p436 year is of very great value. It is very brief and bears a strong family resemblance on the whole to the account of events in Livy and our other authorities, looking somewhat like a hasty and a careless summary of the whole. Who, for instance, but the incredible blunderer Appian would have confused the Tuscan Sena with the Sena in Picenum and so brought Hasdrubal into Etruria, a wildly impossible story, as Bunbury points out? Considering the hopeless confusion and ignorance which this author displays when he deals with Italian topography, an ignorance e.g. which has for ever wrecked the possibility of our attaining any clear conception of the movements of the armies in the Social War, this blunder as regards Sena cannot but be the result of Appian's own foolish imagination, and not proof of another and opposed tradition. It is the unfortunate explanation of a commentator, not the independent tale of a rival tradition.

Again, Appian states that the reason of Hasdrubal's silent retreat from before the consular armies was his desire to join his brother: ὁ δ’ οὔπω μάχεσθαι κεκρικὼς, ἀλλὰ τῷ ἀδελφῷ συνελθεῖν ἐπειγόμενος, ὑπεχώρει. Livy represents him on the other hand as retreating because tormented by his uncertainty as to what was the real significance of the presence of both consuls in the opposing army:

Quonam modo alter ab Hannibale abscessisset cura angebat. . . . Magno opere vereri, ne perditis rebus serum ipse auxilium venisset . . . Interdum litteras suas ad eum non pervenisse credere, interceptisque iis consulem ad sese opprimendum accelerasse. His anxius curis . . . signa ferri iussit (XLVII.5‑8).

If now we put Zonaras's account of Hasdrubal's motives for retreat side by side with these two, we may win, I know, a fair insight into the probabilities of the matter. Zonaras says of Hasdrubal:

ὑποπτεύσας οὖν ἡττᾶσθαι τὸν Ἀννίβαν καὶ ἀπολέσθαι, περιόντος γὰρ ἐκείνου οὐκ ἂν ἐπ’ αὐτὸν ὁρμῆσαι τὸν Νέρωνα ἐλογίζετο, ἔγνω πρὸς τοὺς Γαλάτας ἀπαναχωρῆσαι καὶ ἐκεῖ τὰ περὶ τὸν ἀδελφὸν ἀκριβώσασθαι καὶ οὔτω κατὰ σχολὴν πολεμῆσαι.

Thus Livy says Hasdrubal declined battle and retreated because he was uncertain whether Hannibal had been defeated, or whether simply the Roman consul had stolen a march upon him. Now, his desire throughout was of course to effect a junction with his brother. But had Hannibal been worsted, he himself must needs abandon the whole attempt. Only a retreat to Gaul was left him. Had Hannibal on the other hand simply been deceived, he himself must avoid battle and seek to find some other route whereby he could still join his brother. Zonaras or his authority chooses the former alternative: Appian, it seems, so far as he had any clear conception  p437 of the problem at all, the latter. But either is a fair inference to draw from the account of Hasdrubal's perplexities found in Livy or his authorities. Again this difference of view is hardly proof of a separate Appian-represented tradition.

Indeed how closely the three accounts, Livy's, Appian's, and Zonaras's, are allied is shown by the vaunting comparison of the Metaurus battle with that of Cannae, i.e. as an adequate quid pro quo, which is drawn by each of these three authors.

Nunquam eo bello una acie tantum hostium interfectum est, redditaque aequa Cannensi clades vel ducis vel exercitus interitu videbatur.

Θεὸς δέ μοι δοκεῖ τόδε Ῥωμαίοις ἀντιδοῦναι τῆς ἐπὶ Κάνναις ἀτυχίας, οὐ πόρρω τε ἐπ’ ἐκείνῃ καὶ ἰσοστάσιόν πως ἐκείνῃ γενόμενον. στρατηγοί τε γὰρ οἱ ἑκατέρων ἀπώλοντο, καὶ στρατοῦ πλῆθος ἐγγυτάτω μάλιστα ἐπ’ ἴσης, κ.τ.λ.

Φθείραντές τε ἄλλους πολλοὺς καὶ τὸν Ἀσδρούβαν καὶ λάφυρα πλεῖστα λαβόντες . . . ἱκανῶς τὴν Καννηίδα συμφορὰν ἀνειληφέναι ἐνόμισαν.

Appian's narrative, then, seems to me little more than a summarised and a badly summarised account based, if not on Livy himself, yet at least on the authorities whom Livy consulted. Thus in his sentence already quoted —

Ἀσδρούβας . . . ἐσέβαλέ τε ἐς Τυρρηνίαν . . . καὶ γράμματα πρὸς τὸν ἀδελφὸν ἔπεμπε δηλῶν ὅτι παρείη. Τούτων τῶν γραμμάτων ὑπὸ Ῥωμαίων ἀλόντων, οἱ ὕπατοι Σαλινάτωρ καὶ Νέρων μαθόντες αὐτοῦ τὸ πλῆθος τῆς στρατιᾶς ἀπὸ τῶν γραμμάτων, συνῆλθον ἐς τὸ αὐτὸ πάσαις ταῖς δυνάμεσι καὶ ἀντεστρατομέδευσαν αὐτῷ περὶ πόλιν Σήνας —

He narrates in the most concise form possible the coming of Hasdrubal, the sending of the letters, their capture by the Romans, the junction of consul with consul as a result of this (συνῆλθον ἐς τὸ αὐτό), and their combined camp pitched against the foe near Sena. In view of this rapid sweep through events, and the vital omissions in it (e.g. we ask Appian in vain, 'How were the letters captured and where?' 'Where were the consuls before their capture?' 'What was Hannibal doing all this time?' &c. &c.), the fact that Nero marched north to join his colleague may easily lurk half concealed in the words συνῆλθον ἐς τὸ αὐτό. In my view, in this same one single sentence of summary Appian succeeds in making two desperate blunders. Hasdrubal did not enter Etruria, nor did both consuls face him πάσαις ταῖς δυνάμεσι, for more than half Nero's army was acting as retaining force to Hannibal in South Italy. The sentence is in every way so typical of the careless Alexandrian that surely it is more reasonable to ascribe these mistakes to his own devising than to take them as evidence of the existence of that 'Other Tradition.' The conclusion of the whole  p438 matter seems to me this: Bossi argues that Polybius and Appian represent another and a better tradition which knows nothing of Nero's march. His arguments in the case of Polybius are far from convincing and less probable than the opposite theory. Appian's account is so summarised that it may easily give scope for many theories of the kind, which are based, as this is, on 'omissions in the narrative.' Even if Appian were to represent another tradition, he would surely be rash who asserted that this was better than its opponent, in view of the circumstances of the case, the possibility of the march, the incapacity and brevity of Appian, and so forth. But it appears more probable that Appian relied on Livy's authorities (at least) for his account, and his attempt to summarise these, combined with an entire absence of precision and geographical knowledge ended, and that not unnaturally, in confusion and disaster. Then his web of error is happily used in after centuries by an ingenious but mistaken theorist to prove Nero's march a fiction.

We conclude then on the whole that on every ground it is historically wiser to accept the tale of that march as genuine. And now, but not till now, we are justified in deducing this conclusion as a result of this discussion, viz. that the nearer a proposed site for the Metaurus battle is to Canusium, the more probable, ceteris paribus, this site will be. If a site be proposed so far removed from Canusium that of the reach it Nero's troops must have performed feats which really are physically impossible, grave doubt is at once thrown on the correctness of the choice of that particular site. As I have already said, we argue in this question from the possibility of the march to the site; we are not justified in arguing from any one site of our particular predilection to the possibility of the march.º

 p625  IV. Data of the Battle Site.

Before coming finally to the actual sites proposed for the battle, it is necessary to recall to mind the data to help us in the judgment of them all. We have seen that the camps were five hundred paces distant with a river (unnamed) in the immediate neighbourhood, probably flowing between, and that the camps were 'by Sena,' whatever that may mean. Hasdrubal departs silently from his camp, and endeavours vainly to cross the Metaurus river. He follows up this river, but finds it winds in and out very considerably; and the higher up from the sea he proceeds, the steeper grow its banks. When the Roman cavalry and light-armed troops overtook him, continues Livy, he abandoned the idea of further flight and began to fortify a camp on a hill above the bank of the river. Quum . . . castra metari Poenus in tumulo super fluminis ripam vellet, advenit Livius (XXVII.48.2). The Roman mainguard, however, came up, and the Romans drew up their battle array, Claudius on the right wing, Livius on the left, the praetor in the centre. Giving up therefore the idea of completing the camp, Hasdrubal drew up his forces. On the left wing he placed the Gauls; he himself took command of the Spaniards who formed his right wing; the Ligurians formed his centre; and in front of his line were stationed the elephants (XXVII.48.5‑7).

Other data are added by our other authorities. Hasdrubal drew up his forces on the hill in question. The ground was rough and broken, and his line was also protected by vineyards. Cum  p626 Hasdrubal bello Punico secundo decernendi necessitatem evitans in colle confragoso post vineas aciem direxisset. . . .8

His men, as Livy and Appian tell us, were weary from their flight, and perhaps the Carthaginian general was anxious to avoid a decisive engagement. True, as Livy says, he saw some battle that day was inevitable (postquam pugnandum vidit, 48.5); but he clearly trusted to the natural strength of his position to render the engagement indecisive, or at least to protect his left wing, on which he relied little, from attack. And Livy and Polybius agree in stating that the Roman efforts remained fruitless until Nero's manœuvre, unexpected alike by friends and foes, turned the fortunes of the day.​9 Hasdrubal began the battle by attacking the Roman left wing, and here the struggle was stubborn, neither side having the advantage.​10 While the battle was raging here, Nero vainly endeavoured to storm the hill on which the Gauls were stationed. Here he could make no progress.

ὁ δὲ Κλαύδιος ἐπὶ τοῦ δεξιοῦ κέρατος τεταγμένος προάγειν μὲν εἲς τοὔμπροσθην καὶ περικερᾶν τοὺς ὑπεκαιτίους οὐκ ἐδύνατο διὰ τὰς προκειμένας δυσχωρίας, αἷς πεπιστρευκὼς Ἀσδρούβας ἐποιήσατο τὴν ἐπὶ τὰ λαιὰ τῶν πολεμίων ἔφοδον.​11

So also Livy:

Gallos prominens collis tegebat; ea frons quam Hispani tenebant, cum sinistro Romanorum cornu concurrit; dextra (?) omnis acies extra proelium eminens cessabat; collis oppositus arcebat, ne aut a fronte aut ab latere aggrederentur (48.8)

Then Nero ὑπ’ αὐτοῦ τοῦ συμβαίνοντος ἔμαθεν ὃ δέον ἦν τράττειν.​12 Taking some of the cohorts from his own wing, the right, he marched with these round behind the rest of the Roman army and suddenly and unexpectedly made his appearance on its extreme left wing. By so doing he outflanked the Carthaginian right wing. His troops overlapped, and closing in they fell upon the foe both flank and rear. The Iberians found themselves assailed now no longer only in front. There thronged on every side save one a press of foemen.​13 This decided the day: ἄνισος ἧν ἡ μάχη. Hasdrubal, seeing all was lost, charged into the midst of the Romans and met his death bravely. The Spanish and Ligurians were cut down, and, from the ground these had occupied, the hill on which the Gauls were posted was scaled with ease, and its drunken garrison put mercilessly to the sword.14

This is the account of the battle given us by Polybius and Livy.  p627 Of our other authorities only Frontinus gives us any precise information concerning the details of the fighting:

Livius Salinator et Claudius Nero, cum Hasdrubal bello Punico secundo decernendi necessitatem evitans in colle confragoso post vineas aciem direxisset, ipsi diductis in latera viribus vacua fronte ex utraque parte circumvenerunt eum atque ita adgressi superarunt.​15

This account is not very clear. Frontinus realises that there was (1) a hill, (2) a flanking movement. But he seems to have conceived all the Carthaginians as posted on this hill, which in front was impregnable; wherefore the Romans, by diverging to the left and right, stormed it on both sides, and so won the day. It is perfectly true that the hill in question was stormed from one side, but not true that this was of any importance in deciding the battle. The battle was decided by the flanking movement, and this was directed not against the troops on the hill, but against those who had moved to the attack in the open, and therefore more level, ground on the Carthaginian right wing. These, then, are all the data we have given us for the actual battle site: a rough hill on one bank of the Metaurus river, at some distance from its mouth, between which hill and the sea the river exhibits considerable windings. The banks of the river hard by must be steep. As the left wing of the Carthaginians was posted on the hill and the Romans came up in pursuit from the sea, if the battle took place on the left bank some considerable space must separate the hill and the river channel (unless the Romans fought with their rear ranks hard on the river, which is unlikely, and leaves no room for the flanking movement); for between hill and river must be left space for the Carthaginian centre and right wing, and also for the outflanking movements of Nero and his troops. If, on the other hand, the battle was fought on the right bank of the river, the hill may be sought in the immediate vicinity of the river channel, and some more level, though not necessarily lower, ground be found on its southern side.

Again, if we take ad Senam of the city, as is more natural, and accept the tale of Nero's march, which I feel bound to do, any site further up the river than (say) Fossombrone, fifteen miles from the mouth, becomes an impossibility. Hasdrubal cannot have retreated from (say) the Cesano to beyond Fossombrone, some twenty-three miles in all, in the one night and part of a day before he was overtaken. And even were the camps on the Metaurus, the difficulty of the site increases every step we go beyond that city.

There are yet two other possible data to aid the search: (1) the present nature of the river bed and banks, and (2) tradition and finds. Neither, however, is conclusive.

1. To‑day the Metaurus valley, as far as Fossombrone, is wide,  p628 fertile, and well cultivated. The river winds considerably, and its bed is broad. The volume of water in it, however, when I saw it at the end of March, was not so great as to preclude an apparently easy possibility of fording it by day in more places than one. The banks, however, were high and steep for the most part, and Dr. Arnold urged with force that in the course of some two thousand years the river bed may very possibly have been considerably raised.​16 Certainly to‑day above Fossombrone, where the character of the valley changes, the bare red water-scarred Lanark-like mountain sides, and the impetuous torrent beds falling into the main river, are proof evident how great must be in winter the torrent of water, earth, débris, stones &c. poured into the Metaurus. Now Hasdrubal appeared in north Italy in early spring, while his brother in the south lay still in his winter quarters. He wasted time, however, in the siege of Placentia, and the battle can hardly have been fought before April. In this Oehler concurs. The traditional date is as late as 24 June, and Pittaluga places it in the summer. In any case, April seems the earliest date possible, and Matzat's 16 Feb. can hardly be accepted. If this be the case, it was probably rather the steepness of the then banks than the then volume of water (though doubtless this too was greater then than now) which made the passage of the river a matter of such difficulty. This, indeed, seems implied in Livy, 47.11. Thus the hill by the river is a just object of search. The windings of the river bed must also be duly found — for hollowed out as this is in the otherwise very level surface of the valley, we cannot easily suppose that the river has taken to itself one or more new courses since 207 B.C. But the steepness of the banks and the volume of water are more doubtful elements of discovery in the search.

2. Very hazardous, too, is the argument alike from traditional place-names and from 'finds.' How extremely dangerous and inconclusive is the argument from place-names as applied to this very Metaurus controversy, I have endeavoured to show recently elsewhere.​17 There is one well-known instance of the ridiculous in this kind of argument. Two hours' walk or more to the south of the plain of S. Silvestro rises a lofty mountain, called Monte Nerone. Che codesto nome abbia una qualche relazione con quello di Claudio Nerone? asks Tarducci, with deservedly tentative temerity.​18 Was it so called because Claudius concealed himself there the last day before joining Livius? This proves too much even for Tarducci's loyal supporter Cantarelli. The mountain is so far away, nor is it at all in the direction of Nero's supposed march. Quella denominazione pertanto, says Cantarelli,​19 o non ha relazione  p629 alcuna col vincitore di Asdrubale, ovvero è un altro indizio che i fatti non avennero quali la tradizione annalistica ce li ha tramandati. That the first is the right hypothesis admits of small doubt. The Monte d'Asdrubale gives us trouble enough in this connection, especially as at the beginning of the seventeenth century there seem to have been two hills in the Metaurus valley bearing the name.​20 To worry further with this distant Monte Nerone would show simply a passion for superfluous toil.​b And in general it would certainly be unwise to attach as much value to this evidence of traditional place-names as Tarducci requires of us in championing his S. Silvestro site.

'Finds' again are chiefly of bones, elephant and human. In any case but sorry evidence, they are the more doubtful in this particular question as they are triumphantly produced on two of the contending sites, the S. Angelo and the S. Silvestro. Oehler rightly is disposed to attach far less weight to this kind of evidence than his Italian contemporaries, Bossi, Tarducci, and Pittaluga. Other more definite finds, in the shape of armour, an inscribed patera, &c., are used to champion the S. Silvestro site. This class of finds is doubtless more valuable as evidence than the former. Its precise importance in this question must be discussed in the consideration of that site.

V. The Three Sites Proposed.

1. La Lucrezia (see map on p419, ante, Plan III). — The great road from Ariminum to Rome, the Via Flaminia, leaves the Adriatic coast at Fanum, the modern city of Fano, and strikes inland up the valley of the Metaurus towards Fossombrone and the Furlo Pass. First from Fano at a distance of about two miles it passes through the village of Rosciano, and rather over three miles further crossed a brook, and passes the hamlet of La Lucrezia beyond. So far the river valley has been from about two miles broad. But beyond La Lucrezia this valley begins to contract, and as far as the next small village of Tavernelle is less than a mile in width. Perched up on a hill to the north of the road lies Saltara, and from this point the hill ridge slopes gradually down in an easterly direction to the village of La Lucrezia. The brook already mentioned, flowing to the north of the village, here for some considerable distance flows roughly parallel to the Metaurus, from which it is over a mile distant. Here in this space between the brook and the main river, to the west of La Lucrezia, and on the left or northern bank of the Metaurus, lies the 'La Lucrezia' site of the battle. The hill slope from Saltara is the collis where Hasdrubal stationed  p630 his Gauls. His right wing was posted on the more level ground by the river.

[image ALT: A map of the battle of the Metaurus, based on a supposed site at La Lucrezia.]

This is the site selected by General de Vaudoncourt in 1812.​21 In his view both the Carthaginian and the Roman camps lay at first by Fano, and both thus north of the Metaurus, the Carthaginian camp being to the west of the Roman. Nero joined Livius, and crossed the river in so doing. Hasdrubal, always intending to penetrate Umbria by the Via Flaminia, declined battle and retreated up the valley some six miles.

L'armée carthaginoise partie de son camp à la première veille de la nuit (on était alors en été) ne pouvait certainement pas avoir fait plus de six milles avant le jour, surtout errant à l'aventure et sans guide et suivant tous les contours de la rivière.

He was therefore overtaken near La Lucrezia. The battle was fought on the site described. The nature of the ground thus selected was such that Hasdrubal must needs have drawn up his men in double, and the Romans in triple, line of battle.​22 Nero's flanking movement is best seen marked by the dotted line in the plan given above on p419.

The advantages of this site are:

(a) By placing the camps north of the Metaurus it is possible to make Hasdrubal's intended route south to have been from the first not that by the Adriatic coast, but in the direction followed by the Via Flaminia, i.e. the main road south via Umbria. This explains why he wrote to his brother that he would meet him in Umbria, and why Claudius wrote urging the senate to block the way by placing the city levies at Narnia.​23 If we suppose, on the other hand, that Hasdrubal had crossed the Metaurus and was following the coast route south till stopped by Livius, then these references to Umbria in general, and Narnia in particular, are unintelligible. This is a strong argument in favour of the left bank over the right. And it explains, too, why Hasdrubal was eager to cross the river, viz. not to retreat but to make his way south to join his brother, as Appian says. He trusted doubtless to his night's start to enable him to outstrip the foe in his advance.

(b) Yet, though the site is found on the left bank, Livius's camp by Fano is not so far removed from Canusium as to make the tale of Nero's march a physical impossibility. It involves an addition of at most only some eight miles to that march, when compared with the alternative theory of the camps near Sena on the Cesano. This addition is a mere trifle.

(c) The 'guide difficulty' as already described is solved. For the unnamed river unde aquabantur is the Metaurus. The errant  p631 guide could thus easily escape over this, as it was flowing just beside the camp.

The difficulties of this site are:

(a) Ad Senam must either be rejected altogether or taken to mean 'in the district of Sena' — surely a strained interpretation.

(b) The river cannot be said to wind greatly between the mouth and La Lucrezia, neither can the left bank be justly said to increase in height between Rosciano or Papirio S. Michele and that village.

(c) If Hasdrubal had simply to follow the broad road from Rosciano to La Lucrezia, he could not have missed his way, nor have spent the whole night in traversing less than six miles, nor have followed the windings of the river.24

2. S. Silvestro (map, Plan IV). — The road continuing up the river beyond Tavernelle reaches the very picturesque little town of Fossombrone, the ancient Forum Sempronii, fifteen and a half miles from Fano. Some two miles beyond the town the way divides. The Via Flaminia proper strikes off to the left and crosses the Metaurus in a wide sweep round by a stone bridge, into which is built an inscription of Trajan's reign. This road now continues up the left bank of the river Candigliano, a tributary of the Metaurus, flowing into it below Trajan's bridge. The road, a masterpiece of engineering, climbs through the rockiest of ravines, and pierces the rock finally only by means of the great tunnel which gives the name of the Furlo to this pass, and is adorned with a Vespasian inscription. After the tunnel the ravine broadens again, and eight miles from Fossombrone reaches the village of Acqualagna. Here again the road divides. One way continues up the Candigliano valley, which runs roughly parallel with the upper Metaurus valley, and leads to the towns of Montiego and Piobbico. The Via Flaminia, however, strikes off southwards up the valley of the Burano, a tributary of the Candigliano joining it at Acqualagna. This leads in five miles to the rough grey little Apennine city of Cagli, and so, now climbing steeply up, now dropping steeply down, now crossing rocky passes, now passing through lanes worthy of Devon, leads into Umbria​c and eventually by Fossato and Nocera Umbra to the great central Umbrian valley at Foligno, whence runs the way via Bevagna (ancient Mevania) to Narni and Rome.

[image ALT: A map of the battle of the Metaurus, based on a supposed site at S. Silvestro.]

From the first parting of the ways by Fossombrone another road continues to follow the upper Metaurus valley, crossing the river and leading to Fermignano. Here a mule track still called the 'Strada Romana' leads over the hills southwards by the Passo di San Gregorio to enter the Via Flaminia again at Acqualagna, thus avoiding the 'Jaws of the Furlo.'​d And it seems from Macci's account in 1613​25 that this was thought an easier way for travellers  p632 bound south than the passage of the Furlo. The interest of this will shortly appear. The main road, however, continues beyond Fermignano up the Metaurus, which here winds very greatly, and rather less than halfway between Fermignano and Urbania, and some twenty-eight miles from Fano and four from Fermignano, reaches on the left bank of the river the small plain of S. Silvestro. This small plain is formed chiefly by a great bend which the river Metaurus here makes. It is closely bounded by hills on the north, among which a few miles distant lies the city of Urbino. From the plain, while the main road continues west to Urbania, another road leads over the mountains to the south to the upper Candigliano valley at Piobbico. This road is described by Tarducci as being molto praticabile. 'To make his way hence to Città di Castello in the upper Tiber valley,' he continues, 'would prove but an easy task to a general who in two months had crossed the Alps' (compare throughout map, Plan I.)

These are the roads, rivers, and cities of the district which concern this site of S. Silvestro, selected by Tarducci as the scene of the battle. This selection is accepted by his fellow countrymen Cantarelli and Bossi. It must be admitted that Tarducci's method of procedure in the investigation was, though simple and a favourite one with military topographers of all nations, yet none the less vicious. One September, he tells us, he found himself on this plain of S. Silvestro. Knowing that traditions of the battle clung about the district, he made inquiries of the peasants. They pointed out to him a mound they called the 'Tomba d'Asdrubale.' They told him that many bones were dug up in this place. He himself noticed on the plain a quadrilateral-shaped raised plateau of earth measuring some 76 yards by 70½. It was uncultivated and retained traces of an oak plantation, but it seemed to him clearly of human handiwork. He had no money to spend in excavations. Hence he was reduced to argue, as he says, 'inductively.' This was the site of the battle. The Carthaginian dead were buried where they fell, but the Roman dead were laid to rest in this earthen mound. So Nero buried his dead at Grumentum.​26 Then, not till then, with this prepossession based on 'place-names,' local tradition, and 'finds,' Tarducci proceeded to consult Polybius, Livy, and Appian. It is scarcely surprising that he found their data in entire harmony with his selected site. And in this manner he represents the course of events:

Hasdrubal always intended to march by the Via Flaminia to Narnia, there to join his brother. Hence, on reaching Fano, he struck inland up the Metaurus, meaning to find some pass over the Apennines into the central Umbrian plain. The Furlo was unsuitable. It was too narrow for an army. The Passo di San Gregorio  p633 at Fermignano offered a far better opportunity, as, indeed, we must, on Macci's testimony, allow. Arrived here, however, he found the way South barred by the consular armies. He therefore followed at nightfall the windings of the river some four miles further up, intending to cross by the mountain road leading to Piobbico; but the guides on whom he relied to show him the ford escaped, the one over the Metaurus, i.e. to find refuge in the hostile camp. Thus he lost time, was involved in the twists and turns of the river, and thus, as Livy says, quum haud multum processisset (47.10), was overtaken on the plain of S. Silvestro, where he fought, his right wing resting on the river, his left on the hill to the north of the plain. In favour of this site may be urged:

(a) The 'Umbrian' argument.

(a) The 'guide difficulty' argument (for on this site it is easily solved), common to both sites on the left bank. Further, as advantages belonging to this site alone —

(c) The agreement with local tradition.

(d) The 'finds,' viz. bones (according to the peasants); a helm and horse's armour (according to Macci); and thirdly a striking find. Last century a grave was opened near Montiego, which lies four miles south of Urbania, in which among other objects was found a silver patera engraved with three inscriptions, which are unintelligible, but whose type is similar to that found on coins of New Carthage and that district, of the epoch immediately subsequent to the second Punic war. The patera, it seems, no longer exists. The three inscriptions, containing the first thirteen, the second nine, the third four, complete characters, were preserved in Lanzi's papers at Florence. The patera is engraved therefore in Iberian characters, and is believed by Lenormant​27 and Hübner​28 to have belonged to one of Hasdrubal's Spanish troops who fell in this Metaurus battle.​29

(e) Livy expressly says that Hasdrubal did not advance far up the river from the camps. For the four miles from Fermignano to S. Silvestro, the river may truly be described as tortuous, whereas for eight miles up from the mouth such a description would be ridiculous.

Thus, then, Tarducci sums up in favour of his site:

O io prendo un solennissimo abbaglio, o le circostanze di quella battaglia, accennate dagli storici, trovano tutte nel piano di S. Silvestro il loro riscontro pieno, naturalissimo. Qui abbiamo la necessaria distanza dal passo di S. Gregorio per rispondere al 'Haud multum processit (sic)' di Tito Livio; qui le ripe del fiume sono esattissimamente o alte, come  p634 dice lo stesso storico, o giacenti in terreni acquastrinosi, come aggiunge Appiano: qui una via naturalissima, rispondente in tutto ai bisogni di Asdrubale, ci dà ragione del 'transiturus:' qui, sulla distesa delle colline, abbiamo naturalissimo il 'collis prominens' che copriva i Galli, e l'altura dove la destra dei Romani se ne stava inoperosa a guardare la battaglia (dextra extra proelium eminens cessabat); qui la pianura si trova seminata di ossa; qui povera gente ignorante, senza sapersi rendere alcuna ragione del nome e degli avvenimenti, vi dice e ripete che in questo luogo fu combattuta la battaglia di Asdrubale, e v' indica il luogo dove il vinto duce giacque sepolto. Mi pare che le prove sovrabbondino a dare piena conferma alla tradizione.​30

And thus Bossi follows suit:

Quella pianura, oltre che risponde esattamente a tutti i dati topografici tramessici dagli antichi, trovasi anche oggidì seminata di ossa, e gli abitanti di que' luoghi conservano ancor viva la tradizione, che in quella Pianura sia stato sconfitto e morto Asdrubale, di cui indicano per fino il luogo della sepoltura.​31

There are, however, at least three serious difficulties which militate against the possibility of this site:—

(a) As was the case with the La Lucrezia site, the words ad Senam present an almost insuperable difficulty, short of rejecting them altogether. Is it really possible to understand the words, as Tarducci expounds them, to mean Tutta quella parte dell' Umbria che restava ad est degli Appennini e che aveva per suoi estremi confini a nord il Rubicone, ad est l'Esino?

Peculiar to this site are the other two:

(b) It is impossible to accept both this site and the tale of Nero's march. The latter must go. From Canusium to S. Silvestro measures 270 miles. To march this in seven days and nights becomes a real physical impossibility. With La Lucrezia the march is just possible; with S. Silvestro it surely is not. From this Cantarelli and Bossi argue that the tale of the march is an annalistic invention. This seems to me an unjustifiable order in argument. As on other grounds we have seen reason to accept the tale of the march, the conclusion to be drawn is surely from the march to the site, not vice versa. This is a grave obstacle to the S. Silvestro site.

(c) The space for the battle on this site is such that the Carthaginian right wing, it seems, must have rested on the river, towards S. Lorenzo in Farnetella. Nero, therefore, in his flanking movement must have crossed the Metaurus behind the Roman left wing, marched up the right bank, and crossed it again to fall on the Carthaginian right wing. Not only is this extremely improbable both from the view of tactics and of possibility, but the  p635 feat could scarcely have been left unmentioned in our accounts of the battle.32

The difficulties in the way of the S. Silvestro site seem very grave. Neither can it be urged that the arguments advanced on its particular behalf, i.e. not common to any site on the left bank, are of a very convincing nature. We may admit the local tradition and the windings of the river. It is true that even with regard to the first, the use of the name 'Tomba d'Asdrubale' cannot be traced further back than to Macci. It is, however, not quite justifiable to argue with Oehler that the name is the product of Macci's inventive genius. A closer acquaintance with the old Venetian book would serve to show that probably its author found the name applied in his own day,​33 and it does not seem very clear if Oehler knows the book at first hand at all.

But to build this theory on the 'finds' mentioned is to construct on but an unstable foundation. Bones go for very little, both here and at S. Angelo (cf. infra). The helm and horse's armour rest on Macci's authority, and even if these be adequate, what proof have we that they were Carthaginian? Finally the patera, however interesting from the point of view of ancient dialects, can hardly be said to be of any use whatever in this topographical controversy. Oehler's answer, indeed, to its evidence, is that when the officers of the Austrian Militär-geographisches Institut devised their Carta dell' Italia centrale in 1851 they attached no importance to the find or the tradition. This is scarcely to the point. Surely it is more satisfactory to remember that the patera was found in a grave near Montiego, four miles away from the Metaurus in the Candigliano valley; that there is a S. Silvestro in the immediate vicinity, but that this is totally different from the S. Silvestro on the northern shore of the upper Metaurus. Shall we therefore, on the strength of this supposed Iberian patera, place the battle site in the upper Candigliano valley?

In actual fact, we must balance a vague local tradition against the impossibility of Nero's march and the difficulty of the flanking movement as arguments peculiar to the S. Silvestro site alone. I cannot but think the balance inclines to its rejection. Any site on the left bank of the Metaurus has two great advantages — the explanation of the Umbria and Narnia allusions, and the solution of the guide difficulty. When it comes to a choice between the two challenging sites on this bank, I cannot but think the La Lucrezia site the more possible of the two.

3. S. Angelo (map, Plan II). — This site lies on the right bank of the Metaurus, about four miles from the mouth. In February 1896 the German Oehler and the Italian Pittaluga spent two days  p636 together in investigating it, and the former has recently published the result of his investigations at length, quoting considerable extracts from his colleague's notes. From these extracts in the main, I give the following description of this site:

[image ALT: A map of the battle of the Metaurus, based on a supposed site at S. Angelo.]

Some four miles from the sea on the right bank of the river, a hill descends to the stream which may be called the hill of S. Angelo from the chapel of that name built upon it. This hill is bounded on the east by a long ravine, down which flows the Rio di Caminate, on the other side of which rises the hill of S. Costanzo. The western limit of the hill of S. Angelo is formed by the Fosso dell' Acqua Salata di Ferriano. To the north flows the Metaurus. On the south are the northern spurs of the mountain ridge along which runs the road from S. Costanzo to Cerasa. The hill of S. Angelo is thus bounded. But it itself is pierced in the middle by a small ravine, at the bottom of which a small nameless brook trickles into the Metaurus. At the mouth, where it enters the Metaurus, this ravine is broad, about 490 feet in all measured from one side to the other as the crow flies. Its sides here, especially the right-hand ones, are very steep, so that it would be impossible for an army on the one to descend and scale the other. The ravine is rather over 2,000 yards in length, and towards its upper end but a few yards separate one side from the other, while, though the height of the ground in general here is about 325 feet above the Metaurus level, there is so little difference of elevation between the sides and bottom of the ravine that movements of troops from one side to the other are perfectly feasible.

If then Hasdrubal stationed his left wing, the Gauls, on that part of the hill to the west of the ravine abutting on the Metaurus, that position was secure from all attack by the Roman right wing posted on the opposite side of the ravine. The position was, in fact, impregnable, save on its southern side from the upper end of the ravine. Here were stationed the Carthaginian centre and right wing, and the nature of the ground here at this upper end allowed battle between these troops and the Roman left wing. As the hill by the river could hardly be stormed, however numerically superior the assailants, Hasdrubal was enabled, as Polybius says, to 'deepen his line at the expense of his front.' And if Polybius adds τούτων μὲν ἤρεσκεν οὐδέν, we must understand Hasdrubal's dissatisfaction to have been caused, not by the weakness of his position, but by the condition and temper of his troops.

Lastly, so far as the hill is concerned, some rising ground to the east of the Carthaginian position would secure quiet cover for Nero's flanking movement. His troops could unobserved climb from the Metaurus road from Fiorenzuola, and so crossing above the ravine descend into the Acqua Salata valley, climb the other side, and so take up a position west, and at the back of the hill of  p637 S. Angelo, thus taking the Carthaginian right wing and centre in flank and rear.

This, then, is the Oehler-Pittaluga picture of events both preceding and during the course of the battle:

Hasdrubal intending to join his brother as quickly as possible chose the most direct route south, viz. the Adriatic coast road. He crossed the Metaurus by a ford near the mouth, and pushed on down the coast till he found his way barred by the two consuls who lay encamped on the south bank of the river Cesano, the unnamed river, therefore, unde aquabantur. [Oehler rejects the tale of Nero's march: this does not now concern us. In any case the tale is most easily accepted if the camps are on the Cesano.] He therefore pitched his camp on the northern shore of the Cesano, and, finding it useless to think of forcing his way through the opposing army, decided to retreat, cross the Metaurus again, and (apparently) make his way south by the Via Flaminia into Umbria instead of by the coast route, now impracticable.

As soon as it was dark, i.e. about 8 P.M., he broke up his camp and retreated in the three columns from the Cesano. Two columns crossed the mountains which separated the Cesano valley from the Metaurus. The third and strongest, including the elephants and the cavalry, marched along the shore. All were to cross by the ford with which Hasdrubal was already acquainted. It is not probable, therefore, that either the way thither or the ford itself was missed, as the army was simply retracing its steps.

To march some seven and a half miles over the hills, estimating the march of the columns at from one and a half to two miles an hour, requires four to five hours. The troops then would debouch on the south bank of the Metaurus near the mouth, somewhere between midnight and 1 A.M. But here the army found itself unable to cross, not because the ford could not be found (an unlikely supposition), but because (as we must therefore assume) the river was found to be in flood. This is suggested by the fact that the guide who now escaped over the Metaurus, though we are expressly we are told he crossed by the ford, yet had to swim, not wade, across.34

At this point the Carthaginian cavalry disappear from the narrative. Pittaluga supposes they were overtaken in the retreat and cut to pieces. He produces the evidence of bones (again!) found on the coast by Marotta. Oehler, however, replies it would yet have to be proved these were the bones of Carthaginian cavalry, for it cannot be doubted that this strip of the coast has been the scene of many petty battles in the last twenty centuries. Neither is it likely that the cavalry should have been overtaken while the foot escaped. He suggests that either Hasdrubal had no cavalry, or that these managed to swim their horses over the flooded river.  p638 In this latter case, however, it is strange we hear nothing more of them. Certainly some cavalry were in the Punic camp on the Cesano.

Hasdrubal's sole resource now, therefore, was to follow up the right bank of the river seeking for some ford whereby to cross to the other side. In the dark, and following the windings of the river, as he was bound to do, since on the southern bank there was no road, he cannot have reached the hill of S. Angelo long before dawn. Here he therefore encamped and waited for the daybreak. If dawn were at 5 A.M., the space of three hours seems to leave barely time for the wandering up the river. However, he doubtless had time enough left him to encourage his tired troops, and also the Gauls found time enough to drink heavily of wine they may have found on the spot.​35 Hesselbarth's scepticism on this point is unjustified.36

The position thus occupied on the hill of S. Angelo was a strong one. Its proximity to the river secured a supply of drinking water, and also the means of crossing when the flood sank. Meanwhile the triangular-shaped plateau of S. Angelo was safe from attack on all its sides save the south, and this too was secure so long as the defence of the whole line of the ridge remained unbroken. Its area, somewhat over a quarter of a million square yards, was more than sufficient for an army of about 40,000 men. Here, therefore, Hasdrubal began to fortify his camp, though elaborate fortifications were impossible owing to both to the rapid pursuit of the Romans and Hasdrubal's need of recruiting the energy of his men by rest after the march.

The Romans pursued after the fugitives also in three columns. It was above all thing important to overtake Hasdrubal before he succeeded in reaching the left bank of the Metaurus. The cavalry under Nero pushed on by the coast up the river from the mouth. The mid-column of light-armed hastened over the hills to support the cavalry. The third and strongest column of legionaries and other foot under Livius followed necessarily at a slower pace over the hills.

Meanwhile at daybreak Hasdrubal had remained quietly in his position. No further march up the right Metaurus bank was attempted. The foe was too near; the country beyond unknown and probably unfavourable for a stand. To attempt to cross the river in daylight with the enemy at hand would be madness. Their sole hope had been a passage in the night just passed away. That hope had proved fallacious. A battle was inevitable, and all that Hasdrubal  p639 could do was to choose a strong position with some hope of escape if he could repel the foe. And this he had surely found on the hill of S. Angelo.

The Romans arrived and made ready for battle. Hasdrubal, ready some time before, opened the engagement by charging with the right wing on the main bulk of the Roman army, the heavy-armed on the Roman left wing with Livius. Here at upper end of the ravine on the south side of S. Angelo the struggle raged indecisively. Nero on the Roman right, finding that the steepness of the ravine made the position of the Gauls an impregnable one, led the 6,000 picked troops of the famous march​37 behind the rest of the Roman army under shelter of the covering hill already mentioned, and so down into the upper Acqua Salata ravine; after which he brought them up the other side, and thus fell unexpectedly on the extreme right wing of the foe, the Spaniards, both flank and rear. The Romans maintained their position on the crest. The Carthaginian army gave way or was cut down. The camp on S. Angelo was then easily stormed from the south and the foreign invader utterly destroyed.

This is Oehler's account of the battle. Livy's text says that Nero after his flanking movement in sinistrum hostium latus incurrit (XXVII.48.14). This is usually taken as a simple misreading for dextrum, as by Madvig and Oehler. Pittaluga, however, to keep the manuscripts' reading, suggests that the Carthaginian right wing advanced so to speak far as to leave a gap between it and the right wing. Of this Nero took advantage, and circling round at the back, not of the whole Roman army, but of his own wing — the Roman right — plunged into the gap and fell on the left flank of the Carthaginian right wing. Polybius, too, says he attacked first the elephant riders, and these were posted in the van of the centre (XI.1.7). However ingenious this suggestion is, it is still hard to see why under these circumstances any flanking movement via the rear of the line of battle, as so carefully described by both Polybius and Livy, was necessary at all. Hence in Plan II on the map given above on p419, I have represented the flanking movement in accordance with Oehler's view rather than with Pittaluga's.

This, then, is the S. Angelo site for the battle and the Oehler-Pittaluga view of events. There is no one but must confess it is a most admirable site simply from the view of probabilities of general­ship and the details of the fighting, even though Hasdrubal's time table appears so precise as to be rather absurd. Further, this site has the great advantage of allowing us to take the words ad Senam in their natural and obvious sense. And lastly, 'finds' also are  p640 on its behalf. Traces of a Roman encampment are found on the slope of the hill to the north-west of Caminate, and Pittaluga suggests this is where Livius's troops encamped after the battle. A caveful of bones of ancient date, by S. Paterniano, just to the south of S. Angelo, was reinforced as evidence as late as 1896 by the discovery of more human bones and complete skeletons, about five hundred yards distant from S. Angelo, the remains, it is urged, of a great battle. The peasants say that the bones in their generation have always been in the cave, but preserve the legend that at some distant date they were all brought together for interment there. Forse quelle ossa, says Pittaluga, sono tutti resti mortali della battaglia del Metauro. Sparse dapprima su tutto il terreno circostante al burrone furono raccolte in tempi diversi da pietosi contadini e riunite in quella grotta. Lastly, the district is full of elephant bones. Oehler remarks gravely he is unable to be sure whether these are bones of the particular species Elephas africanus. In any case, he continues, evidence of the kind is very inconclusive unless a great mass of such bones were found all together, and this is not the case. Indeed, Oehler rightly regards the skeleton evidence as of very little value, and the more contentedly doubtless because on the hill of S. Angelo itself no human remains are known to have been found.

Such evidence is not wanted perhaps, for the S. Angelo ground is undoubtedly a strong claimant among disputing sides. And yet I cannot think that it distances at once all possible rivals. Gratitude is due to the German investigator for his painstaking inquiry and able monograph. Yet this must not blind us to the difficulties which his choice presents, as well as its most dangerous rival, the La Lucrezia ground. Two of these are peculiar to the S. Angelo site by itself; two others are applicably to any site proposed on the right bank of the river Metaurus.

(a) The supposed flood in the river is undoubtedly a device to escape a real difficulty. If Hasdrubal had but just crossed the Metaurus by a ford on his march south, is it probable that even without a guide he would have been unable to discover it on his return a few days later? And, argues Tarducci with some justice, had Hasdrubal been unable to ford the river because this was in flood, is it conceivable that neither Livy nor Appian would have mentioned the fact?​38 If then, the theory of the flood be viewed with some suspicion, in spite of the swimming powers of the guide, how came it that the Punic army was baffled in its search for the ford? To‑day there are but few places where the Metaurus, unless in flood, cannot be crossed. Two fords are especially easy. One lies on a road leading from Cerasa to Fano to the north of S. Angelo, and is probably of very old use. The second is at Cerbara, though replaced now by a bridge. This ford, however, is less easy to cross than the former,  p641 owing to the rapidity of the stream. The inference, then, is that Hasdrubal could not find the ford because he was searching for a new one, not for one by which he had lately crossed; because he was striving to cross from the left bank to the right, to make his way south, not retreating from the right bank back to the left. Livy at least does not assert Hasdrubal's night march was a retreat, though it is so represented by Appian and Zonaras.

(b) The height of the banks and windings of the stream between the sea and S. Angelo are not, it is urged, of such a nature as to satisfy the requirements of Livy's narrative. As regards the banks to‑day, the left bank may be described as steep between Papirio S. Michele and the Madonna del Ponte at the river mouth; the right bank at the entrance of the Rio di Caminate, but not to any great extent until Cerbara is reached, when it continues steep as far as S. Oliviero. The left bank opposite this last section of the river is low. Perhaps, however, even here we have steepness enough to satisfy Livy's requirements. And Pittaluga urges an argument already suggested in this paper: Nulla toglie però che all' epoca della battaglia l'alveo del fiume potess' essere maggiormente intagliato, e che il potere erosivo a monte (for la valle del Metauro è una valle d'erosione come tutte altre dell' Adriatico) possa poi attraverso al tempo con successive dejezioni avere rialzato il letto del fiume a valle. As regards the meanderings of the river, it is possible to describe its course as tortuous between the Rio di Caminate and S. Angelo with enough correctness to save the situation. None, however, would deny that simply from this point of view it would be more satisfactory to find a site west of Cerbara, as indeed is La Lucrezia.

And in general to any site on the right bank are attached two further objections:

(c) The 'guide difficulty' already explained.

(d) The 'Umbria' and 'Narnia' difficulty, a very real one, and I cannot see how with any site on this bank it may be solved. For, if Hasdrubal wrote from Placentia to his brother that he would meet him in Umbria, can it possibly be supposed he would cross the Metaurus and try to according to his way south by the coast road? Can it possibly be supposed that Nero would have recommended the senate to place an army at Narnia unless he had good reason to suppose Hasdrubal might direct his march thither? Can it possibly be supposed that Hasdrubal, after having definitely sent word to his brother of his intended route, all but at once changed his mind — that he set out gaily for Apulia while believing Hannibal was expecting him in Umbria?


This, then, is the present position of the Metaurus controversy, and the conclusion, so far as any is possible, of the whole matter.  p642 The battle must have been fought on the right bank or the left. To any site on the former there apply two main objections — the lesser, viz. the guide difficulty; and the greater, viz. the Umbrian difficulty. To neither can I find a satisfactory solution. To any site on the left bank similarly there exists at least one great objection, which may be called the ad Senam difficulty.

In particular, three sites are the most important rivals: two on the left bank, the S. Silvestro and the La Lucrezia, and one on the right bank, the S. Angelo. Of the two on the left bank the La Lucrezia site is to be preferred for this one great reason, that a choice of the other involves almost certain rejection of the story of Claudius Nero's great march, while for quite independent reasons I have endeavoured to show how very great should be our reluctance to accept such a rejection. Thus are left the La Lucrezia and the S. Angelo sites as rivals. Each is attended by the general difficulties of its position north or south of the river sketched above. Each labours also under particular though minor encumbrances of its own. It may seem but a weak conclusion to refuse to decide definitely between the two. Yet I cannot believe that a certain conclusion is possible. It was, indeed, for a very direct reason that I proposed at the beginning of this paper merely 'to point out what are the sites which contend for the honour of Hasdrubal's defeat, and what are the difficulties and assumptions involved by each.' For where no one certain and positive result may, it seems, be attained, it is surely but idle pleasantry to promise it. And the history of the controversy from the writings of Macci in 1613 to its late vigorous revival in Italy and Germany seems to point to the fact that such a result may hardly in this case be won.

Bernard W. Henderson.

The Author's Notes:

1 Cf. Macaulay, Life and Letters, pp692, 693.

[decorative delimiter]

2 The Livy references are throughout to chapter and section; the book is always XXVII.

[decorative delimiter]

3 Strateg. I.1.9.

[decorative delimiter]

4 Livy, XLVI.6.

[decorative delimiter]

5 Zonaras, IX.9.

[decorative delimiter]

6 Strategie (2. Aufl., Berlin, 1886), pp81 sq.

[decorative delimiter]

7 Livius' Geschichtswerk, Leipzig, 1897.

[decorative delimiter]

8 Frontinus, Strat. II.3.8.

[decorative delimiter]

9 Polybius, XI.1.8‑10. Livy, XXVII.48.9 sqq.

[decorative delimiter]

10 Polybius, xi.1.3, 4, 8. Livy, C. 48.9‑11.

[decorative delimiter]

11 Polybius, XI.1.5.

[decorative delimiter]

12 Ibid. 6.

[decorative delimiter]

13 Ibid. 7‑11. Livy, XXVII.48.12‑15.

[decorative delimiter]

14 Polybius, XI.2.3. Livy, XXVII.48.15‑49.4.

[decorative delimiter]

15 Strateg. II.3.8.

[decorative delimiter]

16 Cf.  Nissen, Rhein. Mus. XXII.570.

[decorative delimiter]

17 See the Classical Review, February 1898, pp11‑16.

[decorative delimiter]

18 Del luogo &c. p22, note 2.

[decorative delimiter]

19 Riv. Stor. Ital. VI.72.

[decorative delimiter]

20 See the Classical Review, February 1898, pp11‑16.

[decorative delimiter]

21 Hist. des Camp. d'Annibal en Italie, III.77‑81.

[decorative delimiter]

22 Ibid. plate XXXV.

[decorative delimiter]

23 Livy, XXVII.43.8, 9.

[decorative delimiter]

24 Oehler.

[decorative delimiter]

25 See the Classical Review, February 1898, pp11‑16.

[decorative delimiter]

26 Livy, XXVII.42.8.

[decorative delimiter]

27 Revue Archéologique, 1882, XLIV.31.

[decorative delimiter]

28 La Arqueologia de España, Barcelona, 1888; Adiciones, p280, quoted by Oehler.

[decorative delimiter]

29 Cf. too W. T. Arnold, Second Punic War, note O.

[decorative delimiter]

30 Op. cit. p21.

[decorative delimiter]

31 La Guerra Annibalica, XII.78.

[decorative delimiter]

32 Oehler.

[decorative delimiter]

33 See the Classical Review, February 1898, pp11 sqq.

[decorative delimiter]

34 Livy, 47.9.

[decorative delimiter]

35 Cf. Frontinus, post vineas.

[decorative delimiter]

36 Historisch-kritische Untersuchungen zur III. Dekade des Livius (Halle, 1889), p549.

[decorative delimiter]

37 So Polybius, XI.1.7: τοὺς αὐτοῦ στρατιώτας. Livy simply: cohortes aliquot subductas e dextro cornu, 48.13.

[decorative delimiter]

38 Op. cit. p11.

Thayer's Notes:

a In the text as printed, here followed a large plate assembling four maps. Not being constrained in the same way as an early‑20c print journal, I've broken it apart into its components, which lets you see each map rather larger than I could have managed otherwise. I've also colorized the maps for readability. Purists may still consult the original maps; and the combined key to the author's maps and my colorizing, which opens in another window, is here.

[decorative delimiter]

b Monte Nerone is a large dark hulk of a mountain; common sense suggests the obvious derivation from nero, "black", and the suffix -one, "big", very commonly met with in Italian placenames.

[decorative delimiter]

c The author has inadvertently slipped into the modern age here. The campaign of the Metaurus, from a Roman standpoint, is already in Umbria, since the Roman region by that name included most of what is now the northern Marche. Here, however, by "Umbria" is meant the modern region, which starts a few miles south of Cagli; in fact its western areas include part of what the Romans thought of as Etruria.

[decorative delimiter]

d The rather dramatic appellation, 'Jaws of the Furlo', is an artefact of translation into English. The standard Italian name for the gorge is Gola del Furlo, where gola is more properly, and appropriately, the throat — which of course is the meaning of the French word gorge.

[image ALT: Valid HTML 4.01.]

Page updated: 9 Apr 20