Another meeting was clearly needed. Antony sailed for Italy in early spring 37: he was accompanied by 300 ships. The menace was unmistakable. Perhaps he claimed he was coming to help Octavian against Sextus;[60] if so, he was naturally disbelieved, and it seems that the townsfolk of Brundisium refused to admit the fleet.[61] Bewildered and nervous, they doubtless trusted that Octavian would applaud them. Antony sailed to Tarentum instead, and Octavian travelled there to meet him. Lepidus was again unrepresented. Negotiations were slow, and it was perhaps latejuly or August before agreement was reached.[62] The questions were indeed delicate: it was certainly not clear that it was in
Antony's interests to support Octavian at all emphatically against Sextus. The mediation of Octavia, it was said, was crucial113 - possibly romantic fiction once again, but she may indeed have played a part.
Finally Antony agreed to back Octavian against Sextus, who was stripped of his priesthood and his promised consulship. Octavian was to carry through the war, but it was agreed that he should delay his attack on Sextus to the following year: it was doubtless Antony who pressed for this, for it offered him the hope of synchronizing his invasion of Parthia with this further war in Italy. The propaganda possibilities were clear: while Sextus and Octavian were refusing to let the civil wars die, Antony would be doing what Roman generals should always have done, advancing the empire and spilling foreign blood. It was all to work out rather differently. They further agreed that Octavian would give Antony 20,000 men and 1,000 elite troops in return for 120 men-of-war and ten skiffs.114 The deal made sense, for Octavian vitally needed reinforcements for the fleet which Sextus had damaged so badly, while Antony had recently been unable to recruit Italian troops. But, from his viewpoint, there was one drawback. He left the ships there and then. Octavian merely promised the troops. They never came.
There was a further problem, of a constitutional sort. The triumvirate had formally expired at the end of 38, leaving the triumvirs' position uncomfortably vague. Probably nobody knew whether their power was now illegal. The triumvirate was an irregular magistracy: to which regular magistracy should it be regarded as analogous? To the consulship, which had a fixed term of one year, but formally ended when the consuls abdicated their office on the last day? On the one hand, the term had passed; on the other, the triumvirs had not abdicated.115 Or perhaps it was closer to a provincial governorship, normally assigned by senatus consultum, which continued until a successor was appointed and arrived? Here there were no successors. In some ways the vast task rei publicae constituendae left the triumvirs more closely analogous to a dictator, who was similarly appointed for a specified purpose and held his office until he abdicated on completion of the task: now the res publico was certainly not yet constituta. But the early, traditional dictatorship had also had a maximum duration of six months, and that had been scrupulously observed:116 what would have happened had a dictator outstayed that
App. ВСя>.9з. 390-1,96.397; Dio xlviii. 54. j;and especially Plut. Ant. 35. Wallmann 1989(0 243) 181-2 thus explains Octavia's prominence on coins of 37-36 celebrating the accord (CRR 12 56, 1262, 1266).
App. BCiv. V.9J.396-7; cf. (with slightly different, less credible numbers) Plut. Ant. 35.7; Brunt 1971 (a 9) J02.
The constitutional puzzle certainly exercised the minds of contemporaries: cf. the elaborate treatment of similar issues at Livy, ш.36.9, 38.1, И-5-6 (decemvirs not laying down their office when their term expired; the decemvirate was an irregular magistracy like the triumvirate); ix.33—4 (similar behaviour by a censor). 116 Cf. Mommsen 1887 (a 6;) n.i3 161.
period? No one knew. Admittedly, the more recent (and very uneasy) precedents of Sulla and Caesar furnished a dictatorship without any such legal maximum term.117 But those dictatorships had been voted in those terms, without any time-limitation. Now it was precisely the specification of a limit which differentiated the triumvirate: how crucial was this difference, and who was to say? Perhaps the closest analogy was to those few provincial commands assigned by lex rather than s.c., such as Caesar's command in Gaul. That had carried a fixed term - but the events of 51-50 had shown that the legal implications of its expiry were tangled and unclear. Were further confusion required, it was offered by the triumvirs' provincial commands. They had assigned these to themselves by virtue of their triumviral powers, but had also had them ratified by s.c.; it was not at all clear that their provincial imperium lapsed when their triumviral powers lapsed.118 The analogy with a regular proconsul, assigned a province by s.c., was close.
In short, the legal position was hopelessly confused. Perhaps it did not matter very much: the realities of power were clear enough. But the events of 51-50 had shown that legal issues could be important, at least in propaganda terms; and, anyway, the triumvirs had recently been making a show of their constitutionalism. It would certainly be comfortable to give their status more clarity. Reassuringly, the triumvirate was now formally renewed for another five years, very probably to expire on the last day of 33,119 and a little later this was ratified by the people of Rome.120 But the constitutional tangle was to return.
viii. the year 36 b.c.
While Antony and Octavian had been engaged at Tarentum, their lieutenants had been busy. Agrippa, consul in 37, had considerably strengthened Octavian's fleet; he had also recruited vast numbers of new seamen — 20,000 slaves were freed to allow them to serve.121 Most impressively of all, he had constructed the portus Iulius in Campania by linking the shallow Lucrine lake by a canal to the much deeper Avernus, then removing the dyke separating the Lucrine lake from the sea. The work was completed by two tunnels connecting the Avernus with
1,7 Cf. Mommsen 1887 (a 65) ii.i3 703-), 714-16. Caesar's dictatorship had originally been annual, then formally extended to ten years and then 'for life': MRR n »72, 285 n.i, 294-5, 30;, 317-18. 118 Cf. p. 20 and n.8o; Girardet 1990 (c 97). 119 See Endnote pp. 67-8.
130 Thus App. III. 28.80,... Ktu о Siĵ/ios (irtKtKvpwKci. There is no inconsistency here (as is often suggested) with BCiv. v.95.398, where the triumvirs agree the renewal oiStv tr 1 той Srjfiov Вет/ветте;. In BCiv. Appian is simply contrasting the procedure in 37 with that of November 43, when the triumvirs needed a lex to establish them in office (BCiv. iv.7.27). Their powers now authorized them (it could be claimed) to renew their own term: it still suited their current policy to obtain ratification for their acts from the Senate (cf. p. 20 and n. 80) or, as here, the people.
121 Suet. Aug. 16. i; cf. Brunt 1971 (a 9) 508.
Cumae and the beach.[63] Sextus had recently been concentrating his attacks on the Campanian coast:[64] now the tunnels would allow Agrippa to convey supplies safely, while the double lake would afford a protected expanse of water for training crews.