¹ ‘Bedriacensia’, RCCM 7 (1965), 467Google Scholar.

² Fabia, P., Les Sources de Tacite … (Paris, 1893)Google Scholar; Fabia, P., ‘La concentration des Othoniens sur le Pô,’ REA 43 (1941), 192–215CrossRefGoogle Scholar; Gerstenecker, J., Der Krieg des Otho und Vitellius in Italien im Jahre 69 (Progr. München 1882)Google Scholar; Hanslik, R., ‘Die Auseinandersetzung zwischen Otho und Vitellius bis zur Schlacht von Bedriacum nach Tacitus’, WS 74 (1961), 113–25Google Scholar; E. Hardy, G., ed. Plutarch's Lives of Galba and Otho (London, 1890)Google Scholar; Hardy, E. G., Studies in Roman History, Second Series (London, 1909), 158–202Google Scholar; Hardy, E. G, ‘Tacitus as a Military Historian in the “Histories”,’ JP 31 (1910), 123–52Google Scholar; Henderson, B. W., Civil War and Rebellion in the Roman Empire, A.D. 69–70 (London 1908)Google Scholar; Heubner, H., P. Cornelius Tacitus: Die Historien: Kommentar … Band II: Zweites Buch (Heidelberg, 1968)Google Scholar; Koestermann, E., ‘Die erste Schlacht bei Betriacum’, 69 n. Chr.,' RCCM 3 (1961), 16–29Google Scholar; Momigliano, A., ‘Vitellio,’ SIFC 9 (1931), 117–61Google Scholar; Mommsen, Th., ‘Die zwei Schlachten von Betriacum,’ H 5 (1871), 161–73Google Scholar = GS IV, 354–65; Niccolini, G., ‘La prima battaglia di Bedriaco e la foce dell'Adda,’ Rend. Accad. Lincei 15 (1906), 278 ff.Google Scholar; Passerini, A., ‘Le due battaglie presso Betriacum’ in Studi … offerti a E. Ciaceri (Genova, etc., 1940), 178–249Google Scholar (= 1–71 of offprint); Paul, P., ‘Kaiser Marcus Salvius Otho,’ RhM 57 (1902), 76–136Google Scholar; Puhl, M., De Othone et Vitellio imperatoribus quaestiones (Diss. Halle, 1883)Google Scholar; Wellesley, K., ‘Suggestio Falsi in Tacitus,’ RhM 103 (1960), 272–88Google Scholar.

³ Tacitus 163 (and cf. 164).

⁴ 149: ‘Der Bericht des Tacitus über die Schlacht von Bedriacum gehört, obwohl der Parallelbericht Plutarchs (Otho II, 1 ff.) vorliegt, zu den schwierigsten und anscheinend unlösbaren Problemen der Tacitusphilologie.’

⁵ H 11, 42, 2; 43, 1; III, 21, 2; 24, 1; 27, 2.

⁶ Alberini, C. P., ‘Municipium Cremona,’ Bollettino Storico Cremonese 19 (1954) 47 ff.Google Scholar; and plan opp. 32.

^6a Much of this grid is most helpfully (and for the first time) overprinted in red on a reproduction of the 1:100,000 military map of the Cremona area, forming Appendix no. 3 to Pontiroli, G., ‘Cremona e il suo territorio in età romana’, Atti del Centra Studi e Documentazione sull' Italia Romana I (1967–1968), 165–211Google Scholar.

⁷ There is no bridge nowadays.

⁸ It may also be noted that the straight stretch of Roman road running southwards from Bozzolo and passing near Rivarolo would, if projected northwards from Bozzolo, meet the Via Postumia at its crossing of the Oglio.

⁹ Hanslik (115) was misled by Tacitus' vagueness into suggesting that Caecina's camp was moved upto the scene of ad Castores.

¹⁰ Barnabei, F., ‘Frammenti di una cassa militare della Legione IV Macedonica scoperti in Cremona’, Not. d. Scavi 1887, 209–21Google Scholar with Tav. IV (cf. Fig. 1 in Pontiroli's article cited in n. 6a). They are now in the museum at Cremona.

¹¹ H III, 21, 1.

¹² This estimate, which I put forward in RhM 103 (1960), 279Google Scholar, has been questioned by Heubner 131 f.: ‘Diese Rechnung scheint mir daran zu kranken, dass … Wellesley die Strecke, die der an Caecina gesandte Bote zurückzulegen hatte, nicht berück-sichtigt hat.’ Not so: there were several exploratores (II, 41, 1), and it is incredible that on hearing of the Othonian advance, which they in any case expected, it should not have occurred to two of these, on arriving at the division of ways leading to the camp and the bridge, to separate and so announce the tidings to Caecina and Valens at approximately the same time. It therefore remains plausible to argue that Caecina will have begun his ride at the same moment as the arms began to be issued.

¹³ II, 43, 1; cf. 70, 3 ‘deflectere uia, spatium certaminum recognoscere’.

¹⁴ II, 40.

¹⁵ Otho II.

¹⁶ AAT 31 (1896), 920 ffGoogle Scholar.

¹⁷ Passerini cites Dragoni, , Sulla Storia ecclesiastica cremonese, etc., Cremona, 1858, 23Google Scholar.

¹⁸ Hardy himself preferred Adrae.

¹⁹ D. Olivieri, Dizionario toponomastico di Lombardia, s.v.

²⁰ So Heraeus, Wolff-Andresen, La Magna. W. C. Summers (ed. 1904) more egregiously suggested the Adige! Costa (ed. 1938) speaks of ‘Il Po, l'Adda, l'Oglio ed altri minori e canali e rivi’.

²¹ My pupil Mr. C. Hamilton pointed out to me in 1970 that in M adue is preceded by the adue of ad uellandum (bellandum). The prior appearance of this group of letters (standing in M immediately above the river-name) may have helped to generate corruption.

²² See below, p. 49 f.

²³ Against the theory of a change in the course of the Adda see Passerini 43, n. 104. The relevant 1:25,ooo maps are 60 I SE. (Grumello Cremonese), 60 II NE. (Monticelli d'Ongina) and 61 III NW. (Cremona). Note particularly: (i) the ‘high’ ground (50 m) at Spinadesco between the lower course of the Adda (banks 47–39 m) and Cremona (ca. 46 m); and (ii) the course of the Cavo Morbasco.

²⁴ Tacitus 679. This view was attacked by Koestermann.

²⁵ I have tried to balance the good and the bad qualities of our author as a military historian in ch. IV of Tacitus, ed. Dorey, T. A. (London, 1969Google Scholar).

²⁶ There is an Arda Morta NE. of the village of Soarza.

²⁷ Koestermann's defence (27) of Adra is noteworthy: ‘Da es ostwärts Cremona heute in diesem Raum keine Zuflüsse von nennenswerter Bedeutung zum Po gibt, … musste Tacitus notgedrungen einen der zahlreichen Flussläufe, die vom Süden her dem Po zustreben, zur Ortsbestimmung wählen.’ But if the sources of Tacitus and Plutarch were ignorant of any northern confluence near Cremona, why not say that the Othonians were aiming at the area of the bridge? But of course a historian sufficiently well informed to know of the insignificant Arda will also have been sufficiently well informed to know of what is now called the Naviglio Civico; and he will have referred to the latter, not the former.

²⁸ Dizionario … 46.

²⁹ There are few Latin river-names relating to Ialy and ending in -ula or -ella. I have found only Albula, Angitula, Bersula, Entella and Himella.

³⁰ NH III, 131.

³¹ A I 79, 3 ‘Tiberim … accolis fluuiis orbatum’. In this adjectival sense the word is rare.

³² Padi et Aduae could have been prompted by olybius II, 32 or Strabo IV, 192, etc.

³³ In 1970 my pupil Mr. C. J. Cressey suggested aculae, even nearer to agde; but fluminum is incompatible with it; and accola often has the form accula with the glossaries.

³⁴ Cicero, , De Inuentione I, 28Google Scholar (among others).

³⁵ Koestermann's view that Tacitus did not explain Proculus' plan ‘wegen seiner offenbaren Torheit’ is scarcely convincing. The variety of reasons offered by Plutarch (O 8–9 ) shows that no clear account was available of Otho's motives for speed.

³⁶ For instance, the mysterious visit of the two Othonian praetorian tribunes—very probably, as Passerini suggests, from Flavius Sabinus' force on the south bank opposite—paid to Caecina on the morning of 14 April (41, 1). The aliquod honestum consilium could have been an armistice and the remission to the senate of the choice of an emperor, as opposed to the inhonestum consilium of going over to Vitellius.

³⁷ Hardy, JP 145: ‘It was the unaccountable slowness of the Danube army which deranged the sound defence strategy of the Senatorial generals.’

³⁸ Henderson's reconstruction of the Othonian plan made much of an alleged pincers-movement whereby the Vitellians were to be caught between a western and an eastern Othonian force. It is a weakness of this and similar theories that the exact whereabouts of the approaching Balkan legions is not stated (and perhaps was not clearly known) by Tacitus. It was therefore easy for Henderson's critics to dismiss his doctrine by pointing to the certain absence of the Moesian legions from North Italy and the element of doubt surrounding the position of the others.

³⁹ I adopt this date for ad Castores from Koester, F., Der Marsch der Invasionsarmee des Fabius Valens …, (Diss. Münster 1927), 18Google Scholar; and cf. Holzapfel, L., Klio 13 (1913), 289 ffCrossRefGoogle Scholar.

⁴⁰ If the behaviour of XIII and its vexillations is typical, we can claim that the main parties cannot have arrived later than 14 less 5 = 9 days after their vexillations; and on the other hand the dispatch of the legions in two parties would hardly have any point if less than 2 or 3 days intervened. On this reckoning the range within which the interval must fall is 3–9 days. Bearing in mind the adjective modicis (interuallis), I find 4 days or thereabouts a probable estimate.

⁴¹ H II, 85, 1; Suet., , Vesp. 6, 2Google Scholar.

⁴² No date is given by Tacitus for the Othonian council-of-war, but it must have taken place after the arrival of Otho, Titianus and Proculus, and before 13 April. If Otho left Rome on 15 March and travelled at an average speed of 15 mp daily, he would have reached Bedriacum (390 mp away) on 9 April. Possible dates are therefore 10, 11 and 12 April.

⁴³ For the resentment of the Othonians cf. II, 44; 66, 1; 86, 1; III, 24, 1 (reading cur irati).

⁴⁴ The difficulties inherent in this sentence are discussed by Heubner 52 f. I have dealt with some aspects in RhM 103 (1960), 272–4Google Scholar.

⁴⁵ II, 86, 2.

⁴⁶ On the governors of Dalmatia at the time see Wilkes, J. J., Dalmatia (London 1969), 84 and 444Google Scholar. The chronological context of II, 66, I angebat Vitellium is after the emperor's departure from Lyon on or about 25 April, and some time before his arrival at Cremona on 23 May. One of his dispositions was to order XIII to build amphitheatres at Cremona and Bononia, an operation that will surely have taken some weeks, but which, so far as Cremona was concerned, was finished before 23 May. Unless Tacitus has transposed the sequence of events, Vitellius' anxiety, based on the arrival of the message from Italy, must have arisen soon after his leaving Lyon. It seems therefore probable that all the legions specifically mentioned by Tacitus, I, XIII, XIIII, VII (Galbiana, presumably) and XI, were in Italy, and at different places in Italy, by about 20 April. It will be noted from our table that the one legion among those with which we are closely concerned which had not arrived in Italy by this date was VII Claudia, which seems not to figure in Tacitus' list either. For the location of these legions on 15 April see below, p. 50. (The remnants of I were at Bedriacum.) After the capitulation no formation is likely to have moved until Vitellius' order dispersed them. There are slight hints of the surrender of the scattered Balkan legions at II, 49, 4 (‘aliisque in castris’ and 52, 1 ‘posito ubique bello’. At 67, 2 ‘undecima ac septima suis hibernis redditae’, Tacitus has unfortunately failed to avoid serious ambiguity because he has not qualified septima by Galbiana. It is clear from the table that this is indeed the legion meant, and the only excuse one can offer for Tacitus' lapse is that the legio Claudia is first mentioned specifically at II, 85. Thereafter the two legions are nearly always carefully distinguished. But since Tacitus is so shortly to tell us that VII Claudia advanced to Aquileia (admittedly a few days later), the omission here is careless.

⁴⁷ Stevenson, (CAH X, 820Google Scholar) suggested that the Moesian legions were detained by the invasion of Moesia by 9,000 cavalry of the Rhoxolani. But this invasion had been repulsed at latest by early February: H I, 79, esp. §5; in, 24, 2; AFA, under 1 March (‘ob laurum positam’).

⁴⁸ Five days before, on 15 April, when the praemissi spoke to Otho, it was in the neighbourhood of Emona, so that the remark attributed to them by Plutarch (O 15, ἀπαγγέλλουσι τὴν ἐκ Μυσίας ἡμῶν δύναμιν…ἤδη καταβαίνουσαν ἐκ Μυσίας ἡμῶν δύναμιν…ἤδη καταβαίνουσαν ἐπὶ τὴν Άδρίαν) is slightly optimistic, even if taken to apply to the vexillation of VII Claudia.

⁴⁹ Heubner underestimates Otho's difficulties when he concludes (53): ‘Es kann doch keine Frage sein, dass Otho und seine militärischen Berater sich über die Entfernungen der Standorte und die für den Marsch von dort nach Oberitalien, wo man Widerstand zu leisten beabsichtigte, benötigten Zeiträume im klaren waren und der Marschbefehl an die vier Legionen schon längst vor der Schlacht ergangen war.’ This was Heubner's answer to my introductory remark (272) that the legions were (in a general sense) summoned ‘shortly before the First Battle’.

⁵⁰ I, 67, 1.

⁵¹ Judging retrospectively, Hardy was probably not far wrong, though he made no detailed calculation, in stating (127) that ‘Otho began to move a month and a half too late’. True: but a move at the end of January was not a political possibility. It is noteworthy that Tacitus explicitly denies that Otho was slow to move: I, 85, 1; 87; 89, 3; II, 11, 2–3.

⁵² 1, 90, 1; AFA ‘pr. id. Mart, uota nu <n>cupata pro s[al]ute et reditu [Vitellii] Germanici imp.’ The inscription was carved after Otho's death, and Vitellius' name gauchely inserted.

⁵³ Some few units (cohortes urbanae?) were left behind: 55, 1 ‘quod erat in urbe militum’.

⁵⁴ I, 76, 1–2.

⁵⁵ II, 60, 1; 74, 1; Suet., Vesp. 6.

⁵⁶ At 1, 87 (‘Poeninae Cottiaeque Alpes et ceteri Galliarum aditus Vitellianis exercitibus claudebantur’), exercitibus is usually taken as an instrumental ablative. But a dative is possible and indeed yields better sense since (a) not all the Gallic passes were shut by Caecina and Valens; (b) a motive for an attack on Gaul is provided not by a Vitellian occupation of the passes but by a belief that North Italy was still safe from invasion because the passes were blocked by snow; thus troops could be spared to prevent a Vitellian advance along the coast of Liguria and/or to stage a possible attack on Gallia Narbonensis. To divert these troops when North Italy was imminently threatened by invasion would have been a very eccentric plan of campaign.

⁵⁷ F. Koester, o.c. (cf. n. 39), brings Valens to Briançon on 20 March, to Turin on 30 March.

⁵⁸ I, 87, 2 ‘destinati’ (date imprecise). There is no mention of them, as we feel there should be, at II, 11, 2. Heubner (55) holds that they left Rome with Otho. It is more likely that they preceded him without troops, since only 20 days separate 15 March nd 3 April (allowing a short interval before ad Castores), and this would imply a march of 18 mp daily over 20 consecutive days—too high a rate. Furthermore the troops murmured against them at he time of Macer's successful crossing of the Po (II, 23, 2) apparently at the end of March; and there must be time for Otho, informed of the troops' grievances while marching north, to summon Titianus from Rome. The latter arrives at Bedriacum in time for the council-of-war. At II, 23, 5 I take accitum as happening in late March, and praeposuit as operative about 10 April. It seems to follow that Paulinus left Rome about 10 March without troops hence the omission of his name at II, 11) and made a quick journey of 7–8 days on horseback, arriving in the north in the latter half of March.

⁵⁹ Tacitus introduces his mention of Otho's decision to replace the high command after describing Macer's reluctance to allow his troops to advance far from the Po opposite Cremona. It is a reasonable guess that it was the loss of Cremona to Caecina (not mentioned by Tacitus, and surely played down by his source Paulinus) that decided Otho that a more vigorous policy was required than the cautious veteran was inclined to pursue. This would explain the awkward transition at ch. 23, 3–5 from the restraint of Macer to the replacement of Paulinus and Celsus as joint commanders-in-chief. Macer himself remained in command of his own force until after the bridge-building began (36), i.e. until very shortly before the First Battle of Cremona. The loss of this important town following immediately upon the successful defence of Placentia will have been attributed to Paulinus and/or Gallus, the latter of whom stopped at Bedriacum while on his way to relieve Placentia, and manifestly failed to occupy Cremona before Caecina did so. There is a strange suppression of this unfortunate loss at 32, 2 (‘obiacere flumen Padum, tutas uiris murisque urbes, e quibus nullam hosti cessuram Placentiae defensione exploratum’). One can imagine Otho's retort.

⁶⁰ Plutarch's account (O 8–9) appears to be mere speculation: cf. esp. 8 καὶ Πρόκλῳ μὲν ἐδόκει καὶ Τιτιανῷ τῶν στρατευμάτων ὄντων προθύμων καὶ προσφάτου τῆς νίκης διαγωνίσασθαι…

⁶¹ Hardy (CP 130) holds that ‘the bridge-building was a reply to Otho's plan for advance, not vice versa’. It is true that at a first glance both Tacitus and Plutarch seem to put the bridge-construction after the Othonian council-of-war. But Tacitus' imperfects at 34 (opperiebantur, dirigebantur, claudbat) and his inchoato are temporally imprecise, as is Plutarch's συνέβη δὲ ταῖς ημέραις ἐκείναις καὶ περὶ τὸν Ἠριδανὸν ἀγῶνα γενέσθαι, τοῦ μὲν Κεκίνα ζευγνύντος τὴν διάβασιν κτλ. It is more probable that Caecina embarked on his bridge-building immediately after the check ad Castores (i.e. about 7 or 8 April) in order to neutralize discontent. In any case the council-of-war cannot be pushed back to a date sufficiently early to allow time for the building of the bridge and the associated fighting after it. The gladiatorial force opposite the bridge will have reported its construction to Otho, who immediately gave orders for the Placentia reinforcement to move eastwards to their assistance (36, 2). Caecina for his part brought in the Batavian contingent of Valens' army to display its amphibious skills and protect the engineers. The structure was still incomplete by 14 April, no doubt owing to the fire it had suffered—described by Plutarch but passed over by Tacitus.

⁶² For possible fears of treachery on the Othonian side, cf. 37, 1; 41, 1; 42, 1; 44, 1.

⁶³ Koestermann (28 n.) rightly points to the importance of the fact that the testimony of Spurinna, present with his forces opposite the bridge on 14 April (Flavius Sabinus commanding the combined army of praetorians and gladiators) may well have been available to Tacitus. One may add that the failure of the Othonian attack over the river on that day, described more fully in Plutarch than in Tacitus, may have been played down by the latter out of the same respect for Spurinna as prompted the somewhat inflated account of his successful defence of Placentia. The praetorians denounced by Plutarch (O 12) can only be the troops commanded by Spurinna and it may have been their tribuni who were attempting negotiations with Caecina on the morning of 14 April (cf. Passerini 53 and Krauss, L., De Vitarum Imperatoris Othonis fide quaestiones (Progr. Zweibrücken 1880), 48Google Scholar).

⁶⁴ Not of course Vespasian's elder brother, as Hanslik (121) states: at 36, 2 Tacitus inserts consulem designatum to prevent this misconception.

⁶⁵ Nagl's view (PW 1 A, 2, 2051) that ‘Die Schwierigkeit des Geländes bedingt die Unterbrechung des Marsches and Errichtung eines Lagers am 4. Meilenstein’ is pure phantasy, for the ground is flat and without obstacles except for the R. Delmona crossed by a bridge (III, 17, 1) near the marching-camp. Hanslik (121), without providing a reason, points out rightly that the form of the sentence at II, 39, 2 ‘promoueri ad quartum a Bedriaco placuit’ deutet darauf hin, dass von vornherein gar nicht geplant war, weiter vorzurücken’.

⁶⁶ II, 44, 1 ‘multo adhuc die’.

⁶⁷ Syme (Tacitus 678) thinks that in Tacitus there were two debates, one at the marching-camp and one en route from it. I prefer to believe, with Heubner, that Tacitus has merely re-arranged his material, as we find it in Plutarch, so as to bring the epigram about Otho (‘aeger mora et spei impatiens’) into a more prominent position. There was, no doubt, a debate that spread over the two days. Heubner (151) suggests that the ablative absolute Celso et Paulino abnuentibus … hints at a prolongation of the discussion during the march, in which these generals continued to press for the avoidance of undue proximity to the enemy. In Plutarch the ἀντιλοϒία precedes the arrival of the Numidian, and thereafter ἄραντες ἐΧώρουν. The ‘tableau’ sentence including he imperfect petebant, whose temporal context is not lear, should probably be taken closely with the preceding quod factum est, as in the quotation at the head of this paper. Heubner is also probably justified in saying that ‘Die Formulierung ist … absichtich … unbestimmt’, i.e. that Tacitus himself was not clear about the duration of the deliberations.

⁶⁸ Passerini 40: ‘L'avanzata di 4 miglia era stata evidentamente u n mezzo dilatorio’.

⁶⁹ Virg., , Cat. 10, 12Google Scholar ‘lutos a Gallia’; 15 f. ‘tua… in uoragine/tua in palude’; 17 ‘per orbitosa milia’ (of the Mantua-Cremona-Brixi a area). Cf. the desire of Antonius Primus to get away from the north Italian plain in October-November of this same year (III, 50, 1 ‘propinqua hieme et umentibus Pado campis’).

⁷⁰ Syme (Tacitus 678) acutely remarks of this phrase: ‘That indication of strategic plan is not false where it stands—but ought it not to have been placed where it properly belongs, namely as the reason for the march out from Bedriacum the day before?’ In Tacitus, statements are made where they are most effective, not where they are most desirable logically. The question—impediti or expediti?—becomes crucial only on 14 April.

⁷¹ Dessau, H., Geschichte der römischen Kaiserzeit II 331–2Google Scholar: ‘Der Zuzug aus dem Osten, auf den er [Otho] wohl rechnen konnte, war zwar auf dem Marsche, aber in grossen Abständen und noch weit entfernt.’ He does not attempt any calculations.

⁷² As Sherwin-White, (JRS, XLIX, 1959, 145Google Scholar) aptly says, ‘Otho gave up before his Waterloo.’