¹ Degrassi, '55 and '56 (for bibliographical abbreviations, see following page). Accepted by e.g. Panciera, pp. 50–51; Fraccaro, '59, 10–11; Marcello, J., La via Annia alle porte di Altino (Venezia 1956), p. 10Google Scholar; Berti, L. and Boccazzi, C., Scoperte Paletnologiche e Archeologiche nella provincia di Treviso (Firenze, 1956), p. xGoogle Scholar. Broughton and Gray are non-committal: MRR Supplement (1960), 5–6, Greenidge and Clay, Sources ² (1960), 4–5.

² ILS 5807. Mansuelli, 39, 56 for Ariminum and not Forum Popilli as the Popillia's starting point.

³ NS 1953, 343, ILLR 454a: ‘T. Annius T.f. pr.’

⁴ Mommsen, , CIL i, pp. 154–155Google Scholar; v, p. 935; Degrassi '55, 263; Brusin '49–50, 126.

⁵ As Ferrua observes (p. 241), if the Rhegium road was an Annia and all three were built by consuls, we should in any case have to assume that two of them were built by one of the two consular Annii. However, (i) if an Annius built the Rhegium road, it was as praetor or propraetor (see below), and (ii) the Etruscan Annia was a comparatively minor road and could conceivably have been built in the same year as either of the others; the simultaneous construction of the two important highways to Aquileia and to Rhegium is a different matter, See also Addendum, p. 37.

⁶ CIL i², 638, ILLR 454: ‘viam fecei ab Regio ad Capuam et in ea via ponteis omneis miliarios tabelariosque poseivei.’ Degrassi (’55, 261; '56, 39) avoids this objection by suggesting that the author of the elogium did not claim to have set up all the milestones, but just all the bridges. Whether or not this interpretation is accepted, the language of the elogium does not imply that there was enough still uncompleted to merit a new appointment to finish the job.

⁷ For the Etruscan Annia, between Sutrium and Falerii Veteres, see ILS 1038, 1059, 1109; CIL ii, 1532; iii, 1458; AE 1926, 77; Frederiksen, M. W. and Ward-Perkins, J. B., PBSR xxv (1957), 190–193Google Scholar.

⁸ CIL v, 7992 (ILS 5860), Brusin '49–50, 125–127, cf. CIL v, 7992a. CIL v, 1008a (ILS 5375); Panciera (49 n. 9) suggests a provenance from Altinum or Concordia, and T. Terentii occur comparatively frequently in Patavium.

⁹ See below, p. 28.

¹⁰ See Panciera, pp. 49, 51–52 and notes. (With reference to the viae Annia and Postumia, the name ‘Concordia’ is used throughout as shorthand for ‘the site of Concordia’; the colony itself was founded only under the Triumvirate.)

¹¹ E.g. G. F. Chilver, Cisalpine Gaul (1941) 34, n.8; A. Calerini, Aquileia Romana (1930) 251; Brusin '49–50, 116. Livy (xliii, 17.1) puts him first of the three, before M. Cornelius Cethegus, cos. 160 but this need not mean that he was Cethegus' senior: compare Livy xxxiv, 45.1–5, 53.2, on the colonisations of 197 and 194, with MRR i, 334, 345.

¹² Valeria: Livy ix 43.25, cf. Diod. xx, 90.3, 101.5, Livy ix, 41.5, x, 1.1. Aemilia and Flaminia: Livy xxxix, 2.6, 10. Note the road built by a L. Caecilius Q.f. Metellus through Amiternum to Hatria (ILS 5810, 5749; NS 1896, 97); usually dated to 117, but a L. Metellus of uncertain affiliation was cos. 283; Castrum Novum and Hatria were founded between 290 and 286, Amiternum subjected in 293 (Livy, per. xi, x, 39.3Google Scholar).

¹³ It is odd that the road was not built at the time of the colony's foundation, but there must have been a pre-Roman route into Venetia from the great Etruscan entrepôt at Bologna (see below, n.38), and before the Roman road was built traffic to Aquileia presumably used that, or went by sea. When it was recognised that this route was insufficient, perhaps ai a result of a renewed Istrian war, the task of replacing it may well have fallen to ‘Luscus who as an ex-IIIvir must have been one of the patrons of the colony.

¹⁴ '59, 8.

¹⁵ ‘ILS 5806 (ILLR 452). Fraccaro '52 passim.

¹⁶ Fraccaro '52, 270–275.

¹⁷ Fraccaro '52, 272, who identifies Campo mollo with ‘la vasta umida R. dei Camoi a sud della strada Sacile-Fontanafredda.’ The IGM map sheet 39 IV NW (Sacile) shows a ‘contrada dei Camoi’ some 2 km. WSW of Fontanafredda (875090).

¹⁸ Fraccaro '52, 273–274 for Codroipo as Quadruvium, i.e. where the stradalta crossed Augustus' road from Concordia to Noricum. See sheets 40 IV NW and NE and I SW (Codroipo, Mortegliano, Palmanova) for the stradalta.

¹⁹ Cf. for instance Panciera, pp. 52–3; Brusin, , Atti Ist. Ven. cxiv (1956), 286Google Scholar; Tabula Imperii Romani, foglio L33, Trieste (Roma, 1961Google Scholar)—but see next note.

²⁰ The map in Tabula Imp. Rom. softens the angle by indicating a line from Oderzo to Pordenone; but this is hardly justifiable without abandoning the evidence of ‘Postoyma.’

²¹ Tac., Hist. iii, 21.2Google Scholar; note the contrast between the agger and the ‘patenti campo.’

²² Admittedly the ponte della Delizia near Codroipo crosses the least wide point on this stretch of the Tagliamento (see fig. 2, a, and sheet 39 I NE Casarsa della Delizia), which is no doubt why the Roman stradalta, whether or not it was the Postumia, was directed to this point: see below, n. 27. But the river is still easier to cross further south.

²³ Cf. Olivieri, D., Saggio di una illustrazione generale della toponomastica Veneta (Città di Castello, 1914), 27–89Google Scholar. Notice the Jugoslavian town of Postumia (Postojna) in the Julian Alps, some 50 km. ENE of Trieste.

²⁴ Fraccaro '52, 274. Only a chapel ‘della maestra’ (indicating a via maestra ?) 2 km. NW of Pordenone—sheet 39 IV NE (Pordenone) 942097.

²⁵ See fig. 2, a. Sheet 24 II SE (S. Giorgio della Richinvelda) 010141 etc. for the Arzene stretch; the road NW of Vivaro is best seen on sheet 24 II NW (Arba).

²⁶ See fig. 1. The identity of the Piave valley road as the Via Claudia Augusta ab Altino was claimed by the authors of La via Claudia Augusta Altinate (1st. Veneto, 1938Google Scholar), challenged in a review by Fraccaro, (Opuscula vol. iii (1957), 229–232Google Scholar) and elsewhere, and has recently been restated, with a full bibliography of the discussion, by Anti, C. in Studi in onore di Aristide Calderini e Roberto Paribeni vol. iii (1956), pp. 495–511Google Scholar. The decisive argument is surely the commemorative stone (ILS 208) found at Cesio, 10 km. NE of Feltre, which is difficult to reconcile with a road heading towards Tridentum.

²⁷ The bend in the line from Aquileia to the Cellina valley will have been caused by the necessity of crossing the Tagliamento near Codroipo. See n. 22.

²⁸ The narrow and spectacular Valcellina and the corresponding gorge on the western side of the pass have been greatly altered in the last thirty years by hydro-electric schemes and the construction of an improved motor road.

²⁹ Cf. CIL v, 8000 (Maxentius) from the Opit ergium-Concordia-Aquileia route.

³⁰ '56, 33.

³¹ Gloria '81, 100. Neither is on the IGM map.

³² Gloria '79, vol. 2, no. 1377; '81, 97, referring Abbanum to the district of S. Croce on the SSW side of Padua. The present Piazzale S. Croce is at 50 II SE (Padova) 334469.

³³ Gloria '77, no. 206; '79, vol. 1, no. 425; '81, 101. Favrese is not marked on the IGM map, but the map in Gloria '81 puts it about 5 km. south of Padua.

³⁴ Gloria '77, no. 42 (p. 62): ‘de parte de contra Caput silvi et de contra Tribano per via antiqua que vadit per palude maiore. da puzo que dicitur Visignale. justa prato de Grauneto. veniente ipsa via usque in levata maiore que venit de caput silvis. que dicitur Agna.’ Gloria ('81, 102) claims this as evidence that the present-day village of Agna was named after the via antiqua, but it is clear that the latter went through Tribano. It is uncertain where the palude maiore was; it could have been anywhere along the northern bank of the Adige.

³⁵ Cf. Gloria '79, vol. 1, p. lv.

³⁶ Gloria '81, 88–94, based on the place-names Montagnana, Montagnone and Legnago. As for the Padua-Agna-Adria route, it is conceivable that Annius built a branch road from the port on Atria to Patavium to deal with traffic coming from Ariminum by sea in order to avoid the Po marshes.

³⁷ V, 217.

³⁸ Certainly there was at some date a Roman road between Bononia and Padua, though no traces of it remain. See Mansuelli pp. 41, 44; the route may be followed by the evidence of Roman settlement at S. Giorgio di Piano, Ghergenzano (sheet 75 II SE 927690), Gavasete (961712), and Maccaretolo (965729); from there it went presumably through Poggio Renatico, Ferrara and Rovigo.

The route was an old one; there had been traffic between Bologna and Venetia in pre-Roman times, as is proved by the Etruscan influence on Venetic art and language. ‘If the native Atestine work of the third period shows some Ionic and some Oriental influences … the route by which such influences came there was not along the Adriatic but from western Italy through Etruria and Bologna’ (Whatmough, J., The Foundations of Roman Italy, Methuen, 1937, p. 181Google Scholar). Since Whatmough wrote, the discovery of the large Etruscan city at Spina and Etruscan and Villanovan settlements at Ravenna have proved the existence of a coast route, but the density of Villanovan and Etruscan sites along the passes from Florence and Pistoia to Bologna show clearly that these formed the main route. (The location of Venetic, Villanovan and Etruscan sites is shown on the map in the Catalogue of the Mostra dell' Etruria Padana e del la città di Spina (Bologna, 1960). No inferences can be made from their location about pre-Roman routes within Venetia, except that few of the Venetic settlements are situated in the plains.)

³⁹ Cassii were consuls in 171, 164, 127, 124, 107, 96 and 73, censors in 154 and 125.

⁴⁰ According to a Ravennate document of 1299, there was a place Agnavia in the territory of Cornelii., Forum Regesto della chiesa di Ravenna (Regesta Chartarum Italiae, 1st. Storico Italiano, 1931), vol. 2, p. 262Google Scholar: ‘territorium Corneliense et plebatu S. Apolinaris q.v. Agnavia.’ However, this is probably a red herring, since the parish of S. Apolenaris in Imola (Forum Cornelii) included a fundus known as Aquaviva, which is probably what the Ravennate scribe had in mind. I owe this information to Mr. P. A. B. Llewellyn, who quotes a document of January 1033 from the archives of S. Lorenzo in Imola: ‘… possessiones quas habet in terr. Corn, videlicet fundos Aquavivae cum ecclesia S. Apolenaris.’

⁴¹ Gloria '77, no. 28 (p. 42): ‘usque ad Caput argere et ad … usque ad Pupilia.’ Fraccaro ('59, 9) thought that this could be a reference to the island of Poveglia in the lagoon NW of Malamocco, but this was an unnecessary subtlety forced on F. by the need to preserve Degrassi's theory.

⁴² Sheet 51 III NE (Mirano) 541492. Gloria '55 103.

⁴³ 43'55, 265; '56, 40.

⁴⁴ CIL i², 638 (ILS 23, ILLR 454).

⁴⁵ NS 1953, 343. Bracco '54, repeated more succinctly in '60.

⁴⁶ CIL vi, 31338a, 31370.

⁴⁷ '54, 18–29

⁴⁸ Cf. Ferrua. 238–239.

⁴⁹ Mr. Fredenksen points out to me that Forum Claudii (Ventaroli) and Forum Popilli (Carinola) in the ager Falernus south-east of Suessa were a mere two miles apart.

⁵⁰ Italische Landeskunde ii, 901, approved by Degrassi '55, 262 and Ferrua 241–242, but based on an imaginary Via Annia running to Potentia.

⁵¹ It should be noticed that if the Rhegium road was not built by a Popillius there is nothing to date h foundation

⁵² 52 Ferrua, 243–244. See above, n. 7, for the Etruxcan road.

⁵³ Cf. CIL xiv, 3610, Platner and Ashby, Topographical Dictionary of Ancient Rome, 561; ILS 1038, 1052, 1059, 1093, 1109; CIL ii, 1532; AE 1926, 77; NS 1913, 342; 1925, 36; Frederiksen, and Ward-Perkins, , PBSR xxv (1957), 192–193Google Scholar.

⁵⁴ CIL ix, 5959. But see Addendum, p. 37.

⁵⁵ CIL iii, 1456; probably constructed by Marcus Aurelius.

⁵⁶ MRR index and Degrassi '55, 261 for the possibilities; Bracco ('54, 23–25 and '60, passim) favours Annius Luscus. But if the Venetian Annia was built by Luscus in 153, as argued in section 1, prothe most likely candidate is T. Annius Rufus, praetor in or before 131.

⁵⁷ Apart from the historical difficulty of a praetor in Sicilia operating in Italy, there is also the argument reproduced by Mommsen (CIL i², p. 509), that ‘et eidem praetor’ in line 9 of the elogium implies that the record of the road's construction that comes before applies to a time when the author was not praetor. This is a serious objection: all the parallel examples in CIL i² (24, 610, 800) express an antithesis with a different magistracy. It is difficult to argue from analogy since the Polla stone is unlike all orthodox elogia in being in the first person; it is possible that the missing line of the inscription did not mention any magistracy, but just the author's name, presumably at the foot of his staute. The language certainly implies that the author built the road in some year other than that of his praetorship, but it could have been either before or after, since the order of events need not be chronological—and indeed cannot be be the forum was constructed at the same time as the roagued for for a further year and he built the road as (pro)praetor with proconsular imperium (cf. McDonald, JRS xliii (1953), 143–144). He could still have called himself praetor on the milestone; for a contemporary parallel, compare M'. Aquillius' roads in Asia (CIL i², 646–651), too extensive to have been made all in one year, but with milestones all bearing the simple title cos. This solution of the difficulty was suggested to me by Mr. Frederiksen.

⁵⁸ Summarised in CIL i², p. 510.

⁵⁹ Lauria-Castellucio, 965 m.; Mormanno-Morano, 1022 m.; Rogliano-Nicastro, 936 m. On the latter pass there were still several inches of snow on 1 April 1963. However, the Roman road probably avoided it by descending the Savuto valley to the coast (cf. It. Ant. ‘ad flumen Sabaturn’). The fiumare or torrents from the nearby Appenines would complicate the building of a coast road, but similar conditions were overcome in Liguria by Aemilius Scaurus; cf. Strabo iv, 187, on the hazards of the coast road into Narbonensis before the building of the Via Julia Augusta.

⁶⁰ Cic., Verr. ii, 99Google Scholar, iii, 63, Planc. 65, Att. ii, 1.5, Suet. Titus 5, Acta apost. 28.13–15. Lucilius 107 ff. M for the road journey: ‘omne iter est hoc labosum atque lutosum.’ Cicero used the road in 58 to go to Vibo, (Att. iii, 2Google Scholar, ‘iter esse molestum scio’), but in 44 he went there by sea—Att. xvi, 6.1. According to Sallust, however, an army bound for Sicily and Africa inarched overland to Rhegium in 111 (BJ 28.6 “legiones per Italiam Regium”); could this have been a show of force connected with the allies' agitation at this time over their rights to ager publicus (Lex agraria 1. 29; cf. Cic., de Or. ii, 284Google Scholar)?

⁶¹ Livv xxiv, 2.3, xxxii, 29.4, xxxix, 23.3.

⁶² Livy xxiv, 20.15, Varro, , RR ii, 9.6Google Scholar.

⁶³ Seneca, tranq. an. 9.13Google Scholar, cf. Festus 392L ‘saltus est ubi silvae et Pastiones.’

⁶⁴ Hor., Sat. ii, 3.234Google Scholar, Ep. i, 15.22, Virg., Georg. iii, 219Google Scholar, Aen. xii, 715. Cf. Pliny, NH iii, 74Google Scholar, Sall., Hist. iv, 33MGoogle Scholar on Sila, Virg. Georg. iii, 146Google Scholar, on wooded Lucanian hills; Varro, , RR ii, 1.2, 10.11Google Scholar, Hor. epod. 1-27-8. Strabo vi, 255, on Bruttian herdsmen; Lucilius iv, 247–8 M on Lucanian bulls.

⁶⁵ RR ii, 10.6–7. Cf. Tac., Ann. xii, 65Google Scholar, on Domitia Lepida's ill-controlled slaves.

⁶⁶ Sall., Hist. iii, 98MGoogle Scholar, Orosius v, 24.2, Flor. ii, 8.5; Sall., Cat. 42.1Google Scholar; Caes., BC iii, 21.2Google Scholar, Dio xlii, 24–5; App., BC iv, 43Google Scholar.

⁶⁷ Diod. xxxvii, 2.13, Plut., Crass. 10.4 ff.Google Scholar, App., BC i, 92Google Scholar περί τὰ στενὰ, 118 f. etc.

⁶⁸ Livy xxviii, 12.8; Cic. Brut. 85 for a caedes in the forest of Sila.

⁶⁹ Plut. Crass. 10.6–7, Cic., Verr. ii, 99Google Scholar, iii, 63, App., BC iv, 100Google Scholar, cf. v, 19.

⁷⁰ Cic., Verr. v, 39–41Google Scholar; Suet., DA 3.1Google Scholar ‘negotio sibi a senatu extra ordinem dato'—Octavius was going to his province.

⁷¹ Suet., DJ 19.2Google Scholar. Compare Cic. Sest. 12 for the danger that Catiline might have seized ‘Italiae calles, pastorum stabula’; and Tac., Ann. iv, 27.1–2Google Scholar for a quaestor ‘cui vetere ex more calles evenerant,’ sent to deal with a slave-rising in Calabria., CIL ix, 2438Google Scholar (Saepinum, 168 A.D.) for the violence of shepherds and runaway slaves 'per itinera callium. Cf. also Cic. Tull. 18–19 (Thurii).

⁷² Cic., Verr. v, 161–4Google Scholar, cf. Caes., BC iii, 22Google Scholar. Cosa was probably the modem Cassano—E. Aletti, Sibari (1960), 29.

⁷³ Practically all the evidence comes from the century, but the paucity of sources for the second century alter Livy fails us means that this need not be significant. Livy (xxxix, 29.8, 41.6) records a praetor of 185 being sent to Apulia and Tarentum to deal with a ‘pastorum coniuratio’; although the trouble was severe enough to necessitate a prolonged command, it is mentioned neither in the periochae nor in the Oxyrhyncus epitome. That the danger existed before the first century in the south-west as well may be inferred from Diod. xvi, 15, on runaway slaves terrorising Lucania and besieging Thurii in 354. See also the passage of Obsequens referred to below (n. 75).

⁷⁴ Diod. xxxiv-v, 2, App., BC i, 9Google Scholar, Flor. ii, 7.3, for the large numbers of slaves in Sicily; fewer herdsmen than chain-gangs working cornfields, according to Scramuzza, , Roman Sicily (ESAR vol. iii, 1937), 240–244Google Scholar. But cf. Cic., Verr. ii, 5 f.Google Scholar, 149 ff., 188 for grazing and pecuarii, and Strabo (vi, 272) remarks that much of the island was occupied by shepherds.

⁷⁵ Obs. 27 (134 B.C.).

⁷⁶ Note the word oerder ‘a Regio ad Caupuram’ in line 1 of the elogium. The official numbering of miles was from Capua (line 8, and the Vibo milestone), which makes the inversion more significant: a man operating from Sicily would naturally start at the south.

⁷⁷ Cic., legg. ii, 15Google Scholar (‘nostri clientes, Locri’), Planc. 97. De sen. 41 for ‘Nearchus Tarentinus, hospes noster,’ Att. xvi, 6.1, 7.1 for Cicero's hosts at Vibo and Rhegium in 44; Sicca of Vibo, with whom Cicero tried to shelter in 58, was no doubt the Sicilian Vibius whom he had made praefectus fabrum in 63 (Att. iii, 2, 4, Fam. xiv, 4.6, Plut., Cic. 32.1–2Google Scholar; Münzer PW s.v. Sicca). For Cicero's, Sicilian clients, cf. Verr. i, 16, iv, 25Google Scholar, div. Caec. 2- Att. ii, 1.5, xiv, 12.1, Fam. xiii, 30, 32, 34–39.

⁷⁸ See above, n. 57, on Rufus' office. As for the date, there is no good reason why Rufus, with at most only one previous consul in his family, should have obtained his own consulship suo anno. I have deliberately avoided the difficult question of the connection between the Polla elogium and the Gracchan assignments in Lucania, which, if soluble, might help to fix in the date. See the bibliography in ILLR p. 258, plus Mommsen CIL i² p. 509, Bracco '54, 26 f., Ferrua 240 and Degrassi '56, 38. The ager publicus in question may well have been in Sicily.

⁷⁹ See above, n. 57, for the most serious objection.