The Augustan Empire 43 BC

The evidence is summarized in Wilkes 1969 (e 706) 92—5, and Mocsy 1974 (e 677) 42—4.

Emona: Ŝaŝel 1968 (e 691) 562-5. Carnuntum: Kandler, in Stiglitz, Kandler and Jobst 1977 (e 695); cf. Zabehlicky-Sheffenegger and Kandler 1979 (e 710) 1 j. Tilurium: Wilkes, 1969 (e 706) 97. Poetovio: Klemencand Saria 1956 (e 671) 56; cf. Curk 1976 (e657) 64. Siscia: ŜaSel 1974 (e 693) 734. Oescus: Gerov 1967 (e 667) 87-90. Naissus: P. Petroviĉ, IMS iv (1979) 30-1.

Later alterations in legionary deployment were caused by events elsewhere in the empire. Legion IX Hispana departed finally for Britain in a.d. 4j and was not replaced at Siscia, leaving the garrison of Pannonia with two legions. When VIII Augusta moved to the lower Danube in a.d. 44/5 its place at Poetovio was taken by XIII Gemina, transferred from the upper Rhine. With VIII Augusta possibly at Novae on the Danube below Oescus the army of Moesia now comprised three legions. Late under Claudius IV Scythica was moved to the East and its place taken by VII Claudia, perhaps first stationed at Scupi, then later on Viminacium on the Danube above the gorge, and the army of Dalmatia was now reduced to a single legion. In a.d. 62 a crisis in Armenia saw two legions withdrawn from the Danube, XV Apollinaris from Carnun- tum, its place being taken by X Gemina from Spain, and V Macedonica, which was not replaced at Oescus, leaving Moesia temporarily with two legions until, late under Nero, III Gallica arrived for its brief sojourn on the lower Danube.

Though perhaps yet to be fully organized with permanent bases, Roman fleets on the Danube and its tributaries played a major role in military operations and their logistics. The attack on Siscia (Segesta) in 35 B.C. (see p. 550) was effected with ships provided by the allies, but Roman fleets participated in expeditions against the Dacians under Augustus and also, slightly later, in the incidents on the lower Danube described by Ovid. The west coast of the Black Sea was also patrolled by a Roman flotilla stationed at Tomis. Under Claudius the Roman fleet patrolling the Danube was on hand to rescue Vannius from his kingdom, and the reported activities of Plautius Silvanus Aelianus on the lower Danube under Nero (see above) would not have been possible without a fleet in control of the river, not to mention the excursion to the Crimea. The Pannonian and Moesian fleets, later based at Taurunum and Noviodunum — in each case the last harbour proceeding downstream — will have functioned quite separately as long as there was no through passage at the Danube gorge and the Iron Gate. In the Black Sea the Pontus fleet was based on the coast of Asia Minor, and for the Adriatic Ravenna on the coast of Italy remained the principal naval base, with stations elsewhere, including one at Salona.⁶⁷

Most of the auxiliary units in Dalmatia were placed in the territory of the Delmatae. Several were in or close to coastal colonies, with two cavalry alae, one of which was a regiment of Parthian refugees, and two infantry cohorts at Salona, with cohorts also at Iader, Narona and Epidaurum. The latter are also found at the legionary bases, two at Burnum and one at Tilurium, with an ala at the latter perhaps being a part replacement under Claudius for the departed VII Claudia. The four cohorts named on early epitaphs at the camp of Bigeste near Narona are

⁶⁷ Starr i960 (d 237) 23 and 125-41.

unlikely to have been in garrison there simultaneously. Other stations along the road linking the legionary bases Burnum and Tilurium were Promona (a cohort), Magnum (aid) and Andetrium (cohort). In the case of some units, for example a cohort of Syrian archers (cohors II Cjrrhestarum), no base can be identified since serving members are found in several places. In Pannonia a larger number of cavalry units was deployed along the main roads leading to the Danube, notably on the Pannonian Highway at Sala (Zalalovo), Savaria, Scarbantia and Carnun- tum. Roman pottery indicates a military base at Mursa on the Drava, similar to that known to have existed at Sirmium. By Claudius, if not earlier, some cavalry units had been placed near the Danube termini of other roads in the north and east of Pannonia, at Arrabona, Brigetio, Aquincum, Gorsium, Mursa and Teutoburgium (Dalj). Under Augus­tus the military presence in Noricum included a detachment from the Pannonian VIII Augusta at Magdalensberg and perhaps there was another at Celeia (Celje) which, although within Noricum, lay on the Pannonian Highway. Around the end of Augustus' reign a locally recruited auxiliary unit (cohors Montanorum prima) had replaced the legionaries at Magdalensberg. By the time of Claudius the auxilia in Noricum, which in a.d. 69 comprised an ala and eight cohorts, had been moved up to the Danube bank, to bases at Lentia (Linz) and Lauriacum in the west, and Augustiana (Traismauer) and Zwentendorf in the east.[709]

In Moesia auxiliary units may have preceded the legions in their later bases at Singidunum (Belgrade) and Viminacium. Signs of early occupa­tion have been reported in the forts of the Danube gorge at Boljetin and Donji Milanovac. On the lower Danube some early epitaphs, though no precise dating is possible, indicate cavalry units at Augustae (Hurlec), Securisca, Variana, Utus, Oescus and Nikopol. Infantry units were stationed on main roads in the interior, at Timacum Minus (Ravna) in the Timacus (Timok) valley, at Naissus and possibly already at Montana (Mihailovgrad), the later station of the cohors Claudia Sugambrorum veterana, a unit that was already serving in Thrace under Tiberius. Finally the cavalry veteran buried at Tomis may have been serving in the newly occupied Dobrudja under Claudius or Nero.[710] Julio-Claudian military deployment in the Danube lands saw the legions mainly held in the rear before Claudius, with cavalry regiments pushed out as far as the Danube crossings and infantry cohorts patrolling the intervening roads. Under Claudius and Nero a gradual move towards the river is discernible but the date when several of the later known legionary bases were first occupied, for example Carnuntum, Viminacium and Novae, remains uncertain. Before the Flavian period there was no Roman frontier, at least in any military sense, along the Danube.

V. ROMAN COLONIZATION AND THE ORGANIZATION OF THE NATIVE PEOPLES

Long before the time of Caesar, Roman merchants and settlers had reached Macedonia and Illyricum but the formal institution of Roman colonies in both areas began only in the aftermath of civil war between Caesar and Pompey. Colonies were established following the decisive battles at Pharsalus in 48 B.C., Philippi in 42 в.с. and Acdum in 31 в.с. Subsequently, new colonies of Roman citizens were rarely instituted and then only for legionary veterans from the same or adjacent provinces.⁷⁰ Foundation dates of the early colonies remain uncertain, especially of those in Achaea and Macedonia where the evidence often consists of a few locally minted coins. Several colonies were evidently refounded with an infusion of new settlers along with the conferring of new titles. No overall strategic scheme is evident in the places chosen for new settlements, though major harbours and overland routes were doubtless a consideration. Caesar's foundadon at Corinth (Laus Iulia Corinthien- sis) was more a commercial enterprise than a settled colony and later dominated the rest of Achaea. Patrae (colonia Aroe Augusta), a veteran setdement from legions X Fretensis and XII Fulminata and streng­thened by deportations from southern Aetolia, was the main port for traffic with Italy. In spite of more than one attempt at settlement, a colony at nearby Dyme was later absorbed by Patrae. The new city of Nicopolis on the Gulf of Ambracia, founded to commemorate the victory at Actium, was not a colony but rather a concentration of several existing settlements to form a new city. Further north, Caesar's new setders may have contributed to the later prosperity of Buthrotum (Butrint) on the coast opposite Corcyra and in the same area the Augustan foundation at Byllis (Gradisht) overlooking the river Aous also flourished.

The five colonies in Macedonia originated in reparations following civil war.⁷¹ Cassandrea on the Pallene isthmus of Chalcidice and Dium on the Thermaic Gulf were first setded on the orders of Brutus, Philippi with veterans by Antony after the battle. After Acdum Octavian permitted Antonians dispossessed in Italy to settle at Dyrrhachium, Philippi and other places. The titles Iulia Augusta suggest that these may have included Cassandrea, Pella and Dium, in addition to Philippi. Dyrrhachium, formerly the Corinthian colony Epidamnus, lay at the

⁷⁰ Vittinghoff 1952 (c 259) 85-7 and 124-9; Brunt 1971 (a 9) 597-9­⁷¹ Papazoglu 1979 (e 682) 357-61.

western terminus of the Via Egnatia and, like Philippi, Dium and Cassandrea, possessed a large territory. The exceptional privilege of 'Italian status' (ius Italicum), carrying immunity from taxadon, reported for the colonies Dyrrhachium, Cassandrea, Philippi and Dium was evidently a recompense to refugees from Italy and was extended also to those settled in the later municipium at Stobi in Paeonia. Among other communities, Thessalonica, residence of the proconsul, enjoyed the status of 'free city' (civitas libera) probably from 42 B.C., while the 'freedom' of Amphipolis may go back to the institution of the province. Elsewhere, nothing is recorded of the 'free people of Scotussa' or of privileges conferred on Amanda near the border of Epirus and Illyria. Existing federations (koina) of the native peoples were retained to give an impression of an autonomy that persisted for centuries.

Along the Adriatic coast of Illyricum the few Greek colonies, Issa, Pharos, and Corcyra Nigra being the principal settlements, had been threatened by the growth of Roman settlement. By the time Pliny wrote of 'several Greek cities and powerful communities of fading memory' the early Roman settlements (conventus civium Komanorum) had grown into flourishing cides enjoying the status of colonia.¹² The colonia Martia lulia at Salona, and the coloniae Iuliae at Narona and Epidaurum were likely creations of Caesar to strengthen and reward Roman settlers of that area for conspicuous loyalty in the civil war. In the pre-colonial period at Narona there is a record of the civic organization of the conventus, a college of two magistri and two quaestors, one of each being a freedman.⁷³ The new colonies possessed large territories, that of Salona including not only settlements on the mainland that had once belonged to Issa but also the island Pharos (Hvar), which was administered as a prefecture. Uncertainty persists over the status of several smaller Roman settlements on the Dalmatian coast described by Pliny as 'towns of Roman citizens' ('oppida civium Romanorum'), Risinium (Risan), Acruvium (Kotor), Butua (Budva), Olcinium (Ulcinj), Scodra (Shkodĉr) and Lissus (Lezha). Risinium had the epithet Iulium and Scodra is called colonia on a later inscription but most likely they were irregular settlements later constituted as municipia.

In Liburnia the colony at Iader boasted of Augustus as its creator (parens coloniae) and the donor of its defences.⁷⁴ The occasion was probably following Agrippa's seizure of the Liburnian navy in 35 в.с. The same event may be the occasion for the foundadon of a colony at the Liburnian port of Senia (Senj), and in Istria at Pola (colonia lulia Pola Pollenda Herculanea) and Parentium (Poreĉ), while the slightly earlier foundation at Tergeste (Trieste) received the benefit of walls following a

Pliny, HN in. 144. Wilkes 1969 (e 706) 192-261.

CIL hi 1820, Wilkes 1969 (e 706) pi. 28. '« CIL 111 2907, Wilkes 1969 (e 706) pi. 24.

destructive raid by the Iapodes. The Italian status enjoyed by several Liburnian communities may have been conferred in recompense follow­ing their inclusion in the province of Illyricum in n b.C., following a period when Liburnia had been administered along with north-east Italy. Those with ius Italicum are Alvona (Alutae) and Flanona (Flanates) on the west of Istria, Lopsica (Lopsi) south of Senia, and Varvaria (Varvarini) near the border with the Delmatae of Illyricum. A similar reason may explain the exemption from tribute (immunitas) of Curicum (Curictae) and Fertinium (Fertinates) on the island Curictae (Krk) and Asseria (Asseriates) in the south near Iader. The presence of enfran­chized native Iulii suggest that several of these places were organized as municipia under Augustus and it seems certain that most had acquired that status by the end of the Julio-Claudian period: Alvona, Flanona, Lopsica, Ortopla, Vegium and Argyruntum along the coast; in the gulf of Flanona (Kvarner), Fertinium and Curicum on Curictae, Crexa and Absortium on Apsorus (Osor), Arba (Rab) and Cissa (Pag). On the mainland behind Iader lay Nedinum, Corinium, Asseria, Alveria and Varvaria, and, less certain, Clambetae, Sidrona, Ansium and Pasinum (the last two not located).⁷⁵

The postponed discharges of veterans from the armies of Illyricum caused by the wars of Augustus' later years are reflected in the high totals of years of service (stipendia) among veterans settled near Burnum and at Pagus Scunasticus in the territory of Narona.⁷⁶ In Dalmatia many veterans moved to the coastal colonies nearby. The mutiny of a.d. 14 in Pannonia was set off in part by the unappealing prospect of settling at the newly organized colony of Emona, whose defences were being com­pleted in a.d. 14/15.⁷⁷ New colonies to accommodate Danubian veterans were instituted under Claudius. Savaria lay on the Pannonian Highway a few km south of a major settlement of the Celtic Boii, Aequum in Dalmatia near the vacated legionary base at Tilurium, and Aprus or Apri in Thrace near the Sea of Marmara. Legionary veterans were evidently the dominant group in these places, from VIII Augusta and XV Apollinaris in Savaria (though here civilians may have been among the original settlers), VII and XI Claudia at Aequum and VIII Augusta at Aprus. Well-placed smaller settlements also attracted veterans, evidently with official encouragement. In Pannonia along the road north of Savaria, the mixed veteran and civilian settlement Scarbantia boasted the title Iulia though it was not formally instituted as a municipium until the Flavian period.⁷⁸ The settlement of veterans by Claudius at a village in

⁷⁵ Wilkes 1969 (e 706) 107-15. ⁷⁶ Alfoldy 1987 (d 159) 298-512.

Sa5el 1968 (e 691) 564-5.

Mocsy 1974 (e 677) 74 (Scarbantia), 76-9 (Savaria). Pliny, HN iv.47-8; ILS 2718; cf. Velkov 1977 (e 703) 122 (colonia Claudia Aprensis). Veterans were also settled at strategic places on the main highways of Thrace, Gerov 1961 (e 666).

the territory of Salona may have preceded the foundation at Aequum and have been a special provision for members of V Macedonica after service on the bleak wastes of the lower Danube.⁷⁹

In Celtic Noricum five of that province's eight municipia were instituted under Claudius.⁸⁰ The establishment of Virunum in the Zollfeld brought an end to the commercial settlement on the Magdalens­berg, though other factors, including the imposition of an imperial monopoly on the Norican iron workings, may have contributed to the demise of what seems to have been a centre of unbridled free enterprise. Virunum remained the seat of the provincial administration for more than a century and was the leading city in the province. Other municipia were Teurnia and Aguntum in the upper valley of the Drau/Drava, the former on a steep-sided hill above the river that ensured its survival in later centuries. Celeia in the south east had been a Celtic oppidum on the main highway between Emona and Poetovio. Iuvavum (Salzburg) lay north of the Tauern, where the Salzach emerges from its gorge. Though three new municipia were created in Noricum after the Julio-Claudian era, Flavian Solva in the Mur valley and Hadrianic Cetium and Ovilava near the Danube, the Claudian urbanization of Noricum marks the first external assimiladon to Roman ways of the bulk of the native peoples in a Danubian province.

The third book of the Elder Pliny's Natural History includes lists of native communities (civitates peregrinae) of the Danube provinces which in part appear to be based on official lists drawn up following the Roman conquest. The lists of peoples in Ptolemy's Geography, which although compiled in the second century a.d. uses earlier information, differ at several points. Both accounts nevertheless furnish a reasonably compre­hensive account of the native peoples as organized, divided or amalga­mated following the formal imposition of Roman rule.⁸¹

In Illyricum an earlier scheme of administration had included a judicial district (conventus) based on Narona that included as many as eighty-Six separate communities. Later the peoples of Dalmada were grouped into three such districts, based on Scardona, Salona and Narona. The first was the smallest and contained the Iapodes and fourteen civitates of the Liburnia, evidently some smaller inland groups of whom Pliny deems only the Lacinienses, Stulpini, Burnistae (the native inhabitants of Burnum), and Olbonenses worth naming. To the lists of communities in the districts of Salona and Narona are added numerical totals of decuriae as an indication of their strength, and which may have been a unit of the Roman census roughly equivalent to existing native groups. Some of the peoples named are known from earlier times

⁷⁵ Pliny, HN 111.141; CIL hi 8753 (2028); cf. AE 1984, 228.

⁸⁰ Alfoldy 1974 (e 65 2) 91-6. ⁸¹ Mocsy 1974 (e 677) 5 3-4; Wilkes 1969 (e 706) 482-6.

and can be located with reasonable precision, while others were new Roman groupings of several smaller communides, some of whom are also named by Pliny.

The Delmatae, with 342 decuriae, belonged to the Salona conventus and were the largest people in the province that was named after them. Their territory extended along the Adriatic between the rivers Titus (Krka) and Narenta (Neretva) and extended inland across the watershed to include the high plains around Livno, Glamoĉ and Duvno. Deprived of much of the coast through Roman settlements, they had, for ease of communication, virtually the entire garrison of the province based within their territory and even when a legion was transferred its place was taken by a colony of veterans. There are indications that after a.d. 9 some of the Delmatae were transported to new settlements in the interior. The Ditiones (239 decuriae) lay north west of the Delmatae and occupied the forests and valleys of western Bosnia around the river Unac. Their territory was the initial terminus of one of the military roads constructed following the conquest, 'to the foot of mons Ulcirus of the Ditiones' (see above). North of these were the Pannonian Maezaei (269) in the Sana and Vrbas valleys, against whom Germanicus had led an expedition in a.d. 7. The Sardeates (52), possibly to be connected with the place Sarnade or Sarute on the main road between the Sava and the Adriatic, perhaps dwelt around Jajce in the Pliva valley, while the Deuri (25), the Derrioi of Ptolemy and perhaps the Derbanoi of Appian, dwelt around Bugojno in the upper valley of the Vrbas.

The thirteen communities of the Narona conventus represent a major reorganization of the earlier eighty-nine. They include the (V)ardaei, 'once ravagers of Italy but now reduced to a mere 20 decuriae', and the D(a)uersi (17), or Daorsi, who also figure in the warfare of the second century в.с. The Deraemestae (30) were a new formation from several smaller peoples in the hinterland of Epidaurum, including the Ozuaei, Partheni, Hemasini, Arthitae and Armistae. The peoples who had formed the core of the old Illyrian kingdom of the third to second centuries b.c., the Labeatae, Endirudini, Sasaei, Grabaei, the Illyrii 'properly so-called', Taulantii and Pyraei (the former Pleraei) were grouped to form the Docleatae based at Doclea, later a Flavian municipium, at the confluence of the rivers Zeta and Moraĉa. Many of these communities were the inhabitants of a single settlement, for example Enderon (near Nikŝiĉ) of the Endirudini or Kinna (on the east of Lake Scodra) of the Kinambroi, who figure in the list of those who surrendered to Octavian in 33 B.C. The much diminished Daesitiates (103), who had begun the uprising in a.d. 6, inhabited central Bosnia around Sarajevo and the river Bosna. Their fortress (castellum) of Hedum, perhaps in the east of their territory near Breza, was the terminus of another of the military roads driven across the province after the conquest (see above). The Narensii (102) were evidently another new formation of peoples and from their name were centred on the river Naron or Narenta, perhaps the middle and upper course and including the plain around Mostar. Since they are listed among those peoples who submitted in 33 в.с. the Melcumani (24) are not likely to have lived any great distance from the coast. It has been suggested that they may have been inland of the Deraemestae, in the plains around Gaĉko and Nevesinje in eastern Hercegovina.

East and south of the Daesitiates, among the mountains around the upper Drina, Piva, Тага, and Lim valleys, dwelt the formidable Pirustae, 'almost unconquerable on account of the position of their strongholds in the mountains, their warlike temper and, above all, the narrow defiles in which they lived'.⁸² Though named by Ptolemy they do not appear in the list of Pliny and, for reasons of security, had evidently been broken up into the hitherto unknown Siculotae (24) and Cerauni (24). The former may have included Delmatae transferred from the coast and perhaps occupied the area of Pljevlja in what is now northern Montenegro. Perhaps also once part of the Pirustae, though Ptolemy lists them separately as the Skirtones, were the more numerous Scirtari (72) who dwelt close to Macedonia, probably in northern Albania around the middle Drin. Also part of the Pirustae may have been the Glintidiones. As they are recorded also to have surrendered in 33 в.с. they were evidently more accessible than the rest and could have occupied the region of Foĉa in the upper Drina valley. The possible record, dating to the second century a.d., of a princeps at Skelani seems to locate the Dindari (3 3) in the middle Drina valley. Celtic names on epitaphs in that area suggest that they, like the Celegeri just across the border in Moesia, were really a group of the powerful Celdc Scordisci, whose northern communities survived as a civitas with their original name across the border in Pannonia.⁸³ Like the Pirustae it may have been deliberate policy to break them down into smaller groups and, in the case of the Scordisci, to divide them between three different Roman provinces.

Nothing on reladve strength or conventus organization appears in Pliny's list of Pannonian civitates, which corresponds closely with the account of Ptolemy. As in Dalmatia several new formadons appear, some named after rivers or places, while along the Danube in the north the Romans appear to have wrought major changes through the movement of whole communities on either bank of the river. The following communities can be located, downstream along the three

и Veil. Pat. и.115.4.

⁸³ Alfoldy 1964 (e 646). The reconstruction has been rejected, on various grounds, by Papazoglu 1978 (e 681) 371-8.

major rivers: along the Danube the Boii, Azali, Eravisci, Hercuniates, Andizetes, Cornacates, Amantini and Scordisci; along the Drava the Serretes, Serapilli, Iasi, Andizetes and, between Serapilli and Boii, the Arabiates; along the Sava, the Catari, Latobici, Varciani, Colapiani, Osseriates, Breuci, Amantini and Scordisci. The Belgitae named by Pliny cannot be placed. A later addition was the civitas of the Cotini, perhaps in the low-lying ground south of Lake Balaton. Some of the above were well-known peoples before the conquest, notably the Boii, Breuci, Andizetes, Amantini, Scordisci and Latobici. Others are named from single places, the Cornacates from Cornacum (Ŝotin on the Danube above Belgrade), the Varciani from Varceia (attested but not located) and the Osseriates from a place somewhere on the middle Sava. Colapiani and Arabiates are named from the rivers Colapis (Kulpa) and Arabo (Raba), while perhaps the Hercuniates recalled in some way the Hercynia Silva, the great German forest beyond the Danube. Breuci and Amantini, prominent during the rebellion in a.d. 6-8, are likely to have been broken up into several civitates. Possibly the Cornacates belonged to the latter, while the Osseriates, Colapiani and Varciani were all created from the powerful Breuci. Similarly the Arabiates and Hercuniates in the west perhaps belonged to the Boii. The Illyrian Azali may also have been detached from the Breuci and transported north to a new home on the Danube between the Celtic Boii and Eravisci, possibly after the ЪеИит Pannonicum of 14—19 B.C. Beyond the river such changes were matched by the eastward migration of the Suebic Marcomanni and, somewhat later, the move of the Sarmatian Jazyges into the plain between Pannonia and Dacia (see above).

The identification and location of native communities in Moesia is hindered by an almost total lack of inscriptions earlier than the Flavian period. It can be assumed that Roman occupation and organization of Moesia was attended by less drastic measures towards the native population than had been the case in Illyricum. Pliny's list of peoples derives from the period before Moesia was extended to the Black Sea following the annexation of Thrace under Claudius and comprises Dardani, Celegeri, Triballi, Timachi, Moesi, Thraces and Scythiae 'adjacent to the Black Sea'.⁸⁴ Since the arrangement is geographical rather than alphabetical it may not be the official register of civitates, and seems to identify individual communities only as far east as the Triballi. Among these the Celegeri in the north west may, it has been suggested above, have belonged to the Celtic Scordisci, while the Timachi are the inhabitants of the Timacus (Timok) valley. The account of Ptolemy, which corresponds with Pliny's only in respect of the Moesi and Triballi, described arrangements following the Claudian reorganization. The

^u Pliny, HN hi. 149, Ptol. Geog. ni.9.2; Mocsy 1974 (e 677) 67-9.

Tricornenses of Tricornium (Ritopek) replaced the Celegeri, the Picensii of Pincum (Gradiŝte) at the mouth of the Pincus (Рек) the Timachi, but the Dardani in the south and the civitas of the Moesi (around Ratiaria) continued as before. New civitates on the lower Danube were the Oetenses of Utus, at the river Utus (Vit), the Dimenses of Dimum (Baline), the Obulenses (who cannot be located), the Appiarenses of Appiaria (Ryahovo) and the Peucini named from the island Рейсе (Chilia) in the Danube delta. Conditions were far from stable along the lower Danube under Claudius and Nero, and there is some evidence for a short-lived civitas of Dacians in the area, probably the result of depor­tations from across the river.⁸⁵

Like that of Moesia the organization of Noricum as a Roman province appears to date from Claudius but a much earlier record of the native peoples under Roman rule are the dedications set up at the Magdalens­berg in 10/9 b.c. to the three ladies of Augustus' family, Livia and the two Iulias.⁸⁶ The eight peoples involved were the Norici, Ambilini, Ambidr(avi), Uperaci, Saevates, Laianci, Ambisont(es) and Elveti. Ptolemy's list of the Norican peoples is broadly similar but adds the name of the Alauni. The Norici occupied the heartland of the old kingdom around Magdalensberg, perhaps the ancient capital Noreia, in Carinthia and part of upper Styria. The Ambilini, whose name suggests that they lived on both sides of a river, have been placed in the Gail valley, and may be linked with a place Ilouna somewhere in south-west Noricum. The Ambidravi were obviously along the Drau/Drava, and the Uperaci perhaps on their east in the direction of Pannonia, where they may be connected with a place named Upellae somewhere north of Celeia. A place named Sebatum appears to locate the Saevates in the Pustertal. These at first were grouped in a single civitas along with the Laianci, who may then have been their neighbours on the west in the area of Lienz, where the municipium Aguntum was later created. The Ambisontes, who appear also among the list of defeated Alpine peoples on Augustus' monument near Monaco (La Turbie), occupied the long valley of the Isonta or Ivarus (Salzach). Beyond them the Alauni dwelt around Salzburg and the Chiemsee, where dedications were erected to the local deities Alaunae, Alounae and Alona. The Elveti were doubtless somehow connected with the Helvetii far to the west, and may originate from the Helvetian Tigurini who entered Noricum in the second century b.c. From their place in the order of the peoples on the Magdalensberg dedications they were neighbours of the Ambisontes and possibly dwelt on the upper Mur or lower Salzach. These nine civitates will not have been the full total of Norican peoples since they cover only the

⁸⁵ Mocsy 1970 (в 676) 29 n. 32, citing CIL xvi 13, a military diploma issued to a 'Dacus' on 9 February a.d. 71. ⁸⁶ Saiel 1967 (e 690); Alfoldy 1974 (e 632) 67.

south and west of the province. A suggestion that the total may have been thirteen, to match the number of niches in the 'meeting-hall' at Magdalensberg has been received with some scepticism.

The civitates peregrinae of the Danube provinces, perhaps totalling more than eighty by the period of Claudius, remained under military control for at least a generation. Some peoples not direcdy involved in the fighting under Augustus were administered through ad hoc com­mands, such as a prefecture of Iapydia and Liburnia during the war against Bato in a.d. 9. Local leaders fought on the Roman side, such as the leading citizen of Aenona in Liburnia awarded a 'greater torque' by Tiberius for service in the 'Dalmatian war' of the same year.⁸⁷ With legions and auxilia now in more or less permanent bases the decades of relative inactivity under Tiberius and his successors furnish some evidence of how the military administration of the native peoples was organized. Under Claudius or Nero the chief centurion of XIII Gemina at Poetovio is found in charge of the neighbouring Colapiani. The Boii and Azali in northern Pannonia were under the commander of the auxiliary regiment at Arrabona, who was also charged with responsibi­lity for that section of the Danube bank. The Pannonian Maezaei and Daesitiates in northern Dalmatia were administered by the chief centur­ion of the XI Claudia at Burnum. The first recorded procuratorial governor of Noricum had earlier in his career administered the 'civitates of Moesia and Triballia', either after or along with the post of chief centurion of V Macedonica at Oescus, indicating the pre-Claudian administration of what later became the provincia Moesia.⁸⁸ The communities of the Dardani may, in like fashion, have been the charge of senior officers of the other legion in the Balkans, IV Scythica at Scupi or Naissus. Among some of the peoples in Dalmatia there are signs that native chiefs may have been entrusted with power not long after the conquest, perhaps even avoiding altogether the unpleasantness of a military administration, for example among the Iapodes, some of the Delmatae and the Docleatae.⁸⁹ That stage may have been a preliminary to the later creation of cities, though in some cases long after the Julio- Claudian era. All the recorded tides of rank, such as princeps, or social and family organization, gens, cognatio, centuria, decurio and decuria, are of Roman origin, though the structures they denoted already had a long history and were to persist in some areas throughout the Roman era.⁹⁰

All valid indicators combine to testify that Romanization, that much

" ILS 1673 3520 (probably from Aenona, VAHD 52 (1939-45), 5 5 fig. i).

^и ILS 9199 (Colapiani); 2737 (Boii and Azali); CIL ix 2564 (Maezaei and Daesitiates); ILS 1349 (Moesia et Treballia).

" CIL hi 14325-8; 15064-5 ('prinapes' and 'praepositi' of the Iapodes at a shrine of Bindus Neptunus near Bihac in western Bosnia); 111 2776 ('princeps' of Delmatae with Claudian citizenship); ILlug 185 ('princeps' of the 'civitas Dodeatium'). ⁵⁰ Wilkes 1969 (e 706) 185-90.

observed process of material and cultural diffusion during the early Principate, made little or no headway among the Illyrians in the Julio- Claudian era.⁹¹ The same holds good for most of the Thracians, notwithstanding their contacts with the Greek and hellenistic world, and perhaps also for many of the Celdc peoples in the north west, where their early adopdon of what has been called the 'epigraphic habit' may have led to an overestimation of Roman influence as a whole.

It is a fact that around the middle of the first century в.с. hellenisdc and Roman coins were entering the Danube lands in some quantities, while several local groups among Thracians, Dacians, Illyrians and Celts were producing their own coins to imitative standards. On the other hand, it seems reasonable to accept the view that neither imports nor local issues appear in sufficient quantities and nor do they exhibit a range of denomination to indicate that there was a genuine economy based on a circulating coinage. The many coins of Dyrrhachium and Apollonia that appear in the area from c. 100 в.с. onwards may, as has been recently suggested, relate to a slave trade, perhaps to meet the demands of a Roman slave-based pastoral economy which had existed in the south­west Balkans since the defeat of Macedon in 167 в.с. Similarly, the many Roman denarii which appear in Dacia around the middle of the first century в.с. may also derive from a traffic in slaves, in this case Burebista's Dacia acting as a much-needed procurement agency after Pompey's suppression of Mediterranean piracy in 67 B.C. Moreover, when Burebista's powerful Dacia had gone and Rome had advanced to the Danube, the amounts of Roman silver found beyond the river suggests that supplies of slaves had then to be sought from beyond the river. Roman coins came first to Illyricum with the armies and their followers. Hoards are found along the Pannonian Highway, at Emona, Celeia and Poetovio, and in the area of Mursa and Sirmium on the lower Drava and Sava, all undoubted military centres in the time of Augustus. A similar origin is likely for hoards found among the Delmatae, at Bastasi and Livno, and among the Iapodes at Ribnica in the Lika, though a more authentic economy is indicated by coin hoards from the more settled areas near the coast, Zadar and Kruievo in Liburnia, Ĉapljina and Narona in the Narenta valley and on the island Pharos at Hvar and Gajine.⁹²

Italy's commerce with the north east was based on Aquileia and the road from there across Pannonia to the Danube. Across the Julian Alps a Roman trading settlement (vicus) had already existed at Nauportus (Vrhnika) in the late Republic, where once the native Celts had

" Note, however, Velleius' comment on the widespread knowledge and use of Latin among the Pannonians, 11.110.5; discussed by A. Mocsy in Hartley and Wacher 1983 (c 274) 235-7.

⁹² Mirnik 1981 (в 345); Crawford 1985 (в 320) 235-7.

maintained their own customs station.⁹³ In addition to the traffic in slaves, cattle, hides and amber from the Baltic, Aquileia was also the focus for the wholesale import of finished metal products from Nori- cum. By around 50 B.C. a terrace (920m) below the summit (1058m) of the Magdalensberg was the site of a flourishing Roman emporium. Its prosperity is perhaps best signified by the lifesize bronze of the Celtic god Mars Latobius, dedicated by merchants from Aquileia, including one of the well-known Barbii family. Iron, copper, lead, zinc and brass (an alloy of copper and zinc) were all traded in quantities of finished utensils. Some of the timber-framed houses of Roman merchants exhibit a high standard of interior decoration. On the walls of some of the cellars, which were filled with debris c. 35 B.C., each with its own shrine of Mercury in a niche, were scratched inventories of finished wares; of iron or steel, rings (anuli), axes (secures), anvils (incudes), and hooks (unci); of brass or copper, jars (cafi), cups (cumbae), plates (disci), goblets (scifi) and jugs (urcei). After the annexation of Noricum Magdalensberg became the centre of a Roman administration and parts of the emporium were levelled to make space for a complex of official buildings. On some of the walls were scratched informal greetings to the emperors Augustus and Tiberius, whose features appear in caricature, along with notices of sacrifice, in addition to the more formal dedicatory plaques set up in 10/9 в.с. to ladies of the Augustan house by eight peoples of Noricum. Close to these buildings a classical temple, 30m by 18m and still unfinished when the settlement was abandoned, had perhaps been intended for a newly instituted cult of Roma et Augustus

Far from being precursors to Roman political and economic domina­tion, the Roman settlements in Illyricum of the late Republic had little or no impact on the native peoples. Some latifundia may have existed around the lower Neretva on lands seized from the Delmatae but elsewhere the coastal setdements rather seem to have turned their backs on the interior, as has often happened in Dalmatian history. When the proconsul P. Vatinius responded to an inquiry by Cicero, addressed to his predecessor, regarding a runaway slave last seen at Narona, the proconsul's headquarters, the report that the fugitive had last been heard of among the Ardiaei implied that that was really the end of the matter, though Vatinius promised to do his best to find him if he was still within the province.⁹⁵ Veteran and civilian settlements in Achaea and Macedo­nia in the period up to Actium contributed little to urban development in those areas, save for the major centres of Patrae, Corinth, Nicopolis and Philippi. In Illyricum colonies around this time were also a mix of civilian and military setdement but with barely any trace of a native

^я Tac. Ann. 1.20; Pliny, HN ш.128; ILLRP 55-4 ('magistri' of riciu); ŜaSel 1966 (e 689).

⁹⁴ Piccoctini 1977 (e 68)) and for the graffiti, Egger 1961 (e 662). « Cic. Fear. v.9.

component. By contrast the Julio-Claudian urbanization of Liburnia and Noricum owed litde to Roman setdement, civil or military.

In the matter of town-planning and civic architecture the early Roman cities were far from uniform. Narona (Vid) retained the character of an emporium on a hill enclosed by pre-Roman walls but containing some fine buildings and monuments, many erected by prosperous freedmen. Here the landowning class, if it figured at all in the life of the city, chose to reside in the elegant and well-appointed residences known to have existed in the surrounding country during the first century a.d. At Salona a new forum was planned within a street-grid at the centre of the old conventus, though in the grandeur of its architecture it cannot compare with the impressive double-precinct forum and Capitolium at Iader, which occupied a large block, 180m by 150m, at the centre of the city's street-grid. At nearby Aenona the Capitolium stood within a new forum, in which were placed several larger than lifesize statues of the Julio- Claudians carved in Carrara marble. The symmetrically planned defences and street grid of Emona, 524m by 435 m, recall those of Augustan foundations at Augusta Praetoria (Aosta) or Augusta Taurinorum (Turin). The later veteran colonies at Aequum and Savaria were also planned cities, as was the municipium Virunum in Noricum, though the latter lacked defensive walls. Not all Roman cities were on new and level sites: the Claudian municipia in Liburnia saw native hill-settlements physically transformed into Roman cities, for example Asseria and Varvaria, with a regularforum and other public works inserted within the defended precinct. The territory of several colonies in Illyricum is known to have been surveyed and divided by roads and paths into grids of square centuriae. The systems so far known, at Salona, Iader, Narona, Epidaurum, Pola and Savaria, had centuriae measuring 20 by 20 actus (c. 710m by 710m) giving an area of c. 124 acres (c. 51 ha), the prevailing standard of the early Principate.[711]

Though some vestiges of hellenistic traditions survived in the Adriatic cities, the Roman cities in the Danube lands as a whole exhibit a wholly Roman and Italian character. Throughout the Julio-Claudian era bricks and roof-tiles produced in large factories around Aquileia, at least one of which (the Pansiana) was imperially owned, were shipped down the Adriatic, although the army began to make its own bricks and tiles locally under Claudius.[712] The Danube armies stimulated local produc­tion of ornate tombstones, especially in the fine limestone of the Dalmatian coast. Some early legionary monuments in Dalmatia are in the style of the 'door-stone', a type originating in Asia Minor favoured by recruits of eastern origin, notably in legion VII. The most popular form, both among soldiers and in the cides, incorporated the 'window- portraits' of the metropolitan Roman fashion within an architectural frame of pediment and columns in relief on a standing tombstone, with the framed panel for the epitaph below. A similar version became popular in Noricum and Pannonia, where Celtic and Roman funeral images appear in combination. Roman epitaphs are found on the Liburnian circular tombstones, a native tradition which remained popular in the new municipia of the Julio-Claudian era. It is a relief sculpture in Dalmatian limestone which provides perhaps the most authentic image of Rome in the Danube lands at this time, a monument at Tilurium which depicts the trophy (tropaeum) or Roman victory with two native Illyrians chained to its base, awaiting a fate that was all too certain.⁹⁸

Before the conquest was completed Thracians, Illyrians and Celts were being recruited for service in the Roman auxilia, both as cavalry and infantry. Several units appear bearing the names of such peoples as Breuci, Delmatae and Pannonii. The many Dalmatians who served in the imperial fleets at Ravenna and Misenum came it seems as much from the inland peoples as from the seafarers along the Adriatic.⁹⁹ No conse­quence of this recruitment is discernible before the end of the Julio- Claudian period in respect of the spread of Roman ways and habits. Doubtless there were some, their origins concealed, who rose from these lands to high positions in the Roman hierarchy.¹⁰⁰ No Roman governor praises the Danubians for their eager embrace of Roman mores-, indeed the contrary was for long to prevail.

Conquest and retention of the Danube lands was, in the military sphere, the distinguishing achievement of Augustus' Principate. A harsh, underdeveloped and for long intractable part of Europe brought no profit and much loss. Yet completion of the task was essential for a strategy -which deployed the new standing armies around the borders and far from the centre of affairs where their presence nearly always posed a threat to order. The Via Egnada no longer saw the passage of armies to fight civil wars, and only the fall of a dynasty drew the legions back to the heart of the empire from their remote bases along the Danube.

* Illustrations in Wilkes 1969 (e 706).

" Kraft 19)1 (e 67a); Starr i960 (d 237) 7).

¹⁰⁰ Certainly Liburnia had links with some leading senators in the first century. The consul of 16 B.C. L. Tarius Rufus may be of Liburnian origin, and the distinguished jurist of the Flavian era L. Iavolenus Priscus had Liburnian family connexions. See Alfoldy 1968 (e 6; 1) 100-16.

ROMAN AFRICA: AUGUSTUS TO VESPASIAN

C. R. WHITTAKER

I. BEFORE AUGUSTUS

If the province of Africa under the Roman Republic was not quite a land without a history, as Mommsen described it, it was certainly not central to Roman interests. The administration from the Punic town of Udca was rudimentary, largely a matter of supervising the local communities and contracting out the taxes. Nor is there much evidence of a military garrison apart from the small contingent with the governor. This did not, of course, prevent Roman and Italian immigrants from coming, whether as settlers on the land or as businessmen and tax-farmers. But the impression we get is that the numbers were not great, even in the coastal towns, where Roman enclaves formed.¹ The official foundadon of the colony of Carthage in 122 в.с. had been a disaster that had left stranded we do not know how many on its territory. Conservative Roman sentiment had resented the expense of the province and had feared to send out colonists. Evidence of Romans and Italians being settled by Marius is so thin that it is unwise to guess too much about their numbers, although some immigrants probably did arrive.

The only exception to this was the Gaetulian veterans of Marius, settled beyond the far borders of the province, who proved a valuable aid to Iulius Caesar in his campaigns in Africa in 46 B.C., and who were to be an important element in the new Augustan dispensation.² During the civil wars between Pompey and Caesar a fair number of Romans took refuge in Africa. But even so, the Pompeians were hard put to it to raise 12,000 men and, even after reinforcements of 10,000 from Cyrene, they almost certainly had to include native Africans, slaves and freedmen to raise a force of 40,000.

If immigration was relatively light, economic interest in the Roman province of Africa and the adjacent territories of the Mauretanias was considerable — in particular because of the fertile land, the corn and (probably) the slaves. By Cicero's day Africa was regarded as a 'bulwark' of Rome's food supply. Beyond the provincial borders Libyan cities like Vaga (mod. Beja) and Cirta (mod. Constantine) were teeming with

¹ Cf. Caes. BAfr. 97.j. 2 Cats. BAfr. 55.4. j 86

Italian negotiatores in the second century B.C. Archaeological sites in modern western Morocco are reported to contain the relics of as many Italian republican amphorae as those in southern France.[713] Sales of land in the African province are recorded in the Agrarian Law of 111 B.C. to have taken place on several occasions, probably to absentee owners in Rome. There may have been further sales thereafter. All this interest was to have its influence on Augustan policy.

Precisely what Iulius Caesar intended or achieved during the brief period of his dictatorship between 46 and 44 в.с. is not always clear. Massive indemnities were laid upon the coastal cities of Byzacium (south-eastern Tunisia) and Tripolitania, the latter being required to pay an annual tax of 1 million litres of oil, which probably continued until the third century a.d. The adjacent territory of Numidia was organized into a second province named Africa Nova, which Caesar announced would pay 8,000 tonnes of corn in tax, to the acclaim of the Roman people. New settlers came, too, not only to the province of Nova, with its curious annex around Cirta, but also to other places in the old province. Many were veterans of the civil war, hastily demobilized to avoid trouble. But many were surely some of those 80,000 inhabitants of the city of Rome whom Caesar sent abroad. Africa's land and food continued to excite Roman interest.[714]

Here we run into intractable problems of identifying and dating the colonial foundations which absorbed many of these settlers. While there can be little doubt about Caesar's intentions to reorganize the African province, there is no way of proving whether the final act of foundation was Caesar's or his heir's. The best evidence we have of Caesar's work is an inscription from the colonia of Curubis (mod. Korba) on Cape Bon, recording an urban magistrate in 45 B.C. But in a sense it hardly matters. Both Caesar and Octavian acted under similar pressures and it seems perfectly possible that what was de facto begun by Caesar was formally completed by Augustus. Those who perceive grandiose hellenistic schemes in the settlements[715] perhaps forget the simple logic of what took place. Civil wars left confiscated land available for allocation to the victors. Veterans and the Roman poor could reanimate and control some of the most productive territory known in its day.

The foundation of the colony of Carthage illustrates perfectly the difficulty we have in separating Caesar and Augustus. By the end of Augustus' rule Carthage had become the administrative capital of the united provinces of Africa Vetus and Nova and a city of some size and

Rornem road

Provincial boundary

— — Appcoxirnete limit oleorrtrol

® Town witti colonial w municipal flatus Ъу AD 1+ о Other town ■ Military sire

Land over 1.00C metr-as

no 30Ĵ коьп

4

4

⁴ \ -

AFRICA-

¥

TaucHn mnino*\j

T

TRIPOLITANIA

CYRENE

j fWdrtOfWII

vt

IS

Ш

(PROCONSULARISI^ ~f — V- ^

Map t). Africa.

splendour. The territory or pertica of Carthage stretched at least as far as Thugga (mod. Dougga), 100km down the Bagradas valley,⁶ incorporat­ing a whole series of communides of both Roman settlers and nadves. But who was the founder of Roman Carthage? Most now agree that, although Caesar may have drawn up plans, the actual, physical founda­tion was probably not his, if only because the pertica of Carthage extended into lands which formed part of Caesar's province of Africa Nova - surely an impossibility until after his death, when the two provinces became one.

But when after 44 B.C. this happened is impossible to be sure. An enigmatic statement by the Christian theologian, Tertullian, two centur­ies later, claimed that it was 'after the violent abuse of Lepidus and after long delays by Caesar, when Statilius Taurus set up the walls and Sentius Saturninus pronounced the religious rites'. The most plausible date is perhaps 36 or 35 B.C., when Cassius Dio records that Octavian sent out Statilius Taurus as his agent to win over 'both the Africas'. The old provinces were evidently not yet united and were 'in need of a settlement'. Taurus accomplished both, at a time when Octavian's army was racked with mutinous troops demanding their rewards and just after the two governors of Vetus and Nova had been fighting each other. The grant of municipal status to Utica in 36 B.C., presumably after adjustment of its boundaries, adds some corroboration that this was the period of reorganization for the whole territory.⁷

The most cogent objection to such a date is that the prestigious cult of the Cereres fertility gods in Carthage, for which we have a lot of inscriptional evidence in later periods, adopted a system of dating its priesthoods which probably went back to before 35 B.C., although the evidence is not entirely consistent. It is not, however, compellingly self- evident that the start of the cult, which had had a long Libyan history before this, and the foundation of the Roman colony were linked.⁸ Nor is it difficult to accept the evidence that the final colonial charter and 'freedom' of the city waited until 29 or 28 B.C., since delays between the award of status and the adoption of a charter are not unknown elsewhere.⁹

« NTH 510.

⁷ Tertull. Defalt, i; Dio xlix. 14.6,34.1. Utica-DioxLix.16.1. M. Le Glay in Lancel 198) (E748) a 5 j-48 is the most recent to put the view contradicted here.

* Fevrier 1973 (e 731), contra, the view of Fishwick and Shaw 1978 (e 733). Gascou 1987 (e 740) has radically undermined the accepted dates of inscriptions and favours 44 b.c. as year 1 of the Cereres priesthood.

⁹ Dio ui.43.1. A sensible summary of the evidence is in Van Nerom 1969 (e 734).

II. AFRICA AND THE CIVIL WARS, 44~J I B.C.

The civil wars which broke out after the death of Iulius Caesar in 44 B.C. inevitably sucked in not only the two provinces of Africa Nova and Vetus but also the allied kings of the Maghreb who depended on the favours of Roman politicians but were not above profiting from their rivalries.¹⁰ The Libyan prince Arabion, for instance, returned to central Mauretania in 44 B.C. and, encouraged by the sons of Pompey in Spain, killed Caesar's old ally Sittius, who had been settled with his mixed bag of followers at Cirta. Having arrived at an accord with the remaining Sittiani, he brought them over in support of the senatorial governor of Vetus, Q. Cornificius, against the Caesarian governor of Nova, T. Sextius — only to switch support completely in favour of Sextius against Cornificius as the luck of Caesar's murderers ran out in 42 в.с. He subsequently resisted Octavian's nominee, Fango, but was executed by Sextius (by now a supporter of Antony) on the suspicion of his too great ambition, which caused his supporters to change sides yet again in support of Fango. Sextius finally drove the whole lot out of the African provinces.

Further west in Morocco a similar power struggle was being played out between King Bogud, who supported Antony against Bocchus when the latter gave his support to the revolt of Tingis (mod. Tangiers) against Bogud. For his opportunistic action Octavian rewarded Bocchus with Bogud's kingdom plus the rest of western Mauretania from Tingis to Cirta. This large territory Bocchus ruled undl his death in 3 3 B.C.

The events of the civil war are confusing and confused. After Brundisium the two African provinces were allotted to Lepidus in 40 в.с. as his share of the triumviral dispositions and he built up an enormous army there of sixteen legions for the invasion of Sicily in ₃6 B.C. against the Pompeians. This massive army group certainly included many native recruits and must have denuded Africa of its defences. After the disappearance of Lepidus, Octavian — as we saw - realized the pressing need to restore order and sent one of his iron men, Statilius Taurus, in 36 B.C. to do the job. The archives record three triumphs ex Africa between 34 and 28 B.C., which we may assume to have been won for border wars to secure the newly formed province of Africa Proconsularis and its colonists.

But the wars were also partly the consequence of the death of Bocchus in 3 3 B.C., who had controlled the Mauretanias as Octavian's nominee. Dio claims that Octavian actually annexed this vast territory, and this has been taken as explanation of the anomalous fact that later, after a new puppet ruler, Juba II, had been installed in 25 B.C., we find a number of

¹⁰ The complex narrative is mainly in App. BCiv. rv.jjff; supplemented by Dio xlviii.21-23.

Roman veteran colonies existing within the native kingdom. On balance it seems unlikely that Octavian went this far. There is no allusion to Mauretania as a province in the account of the settlement of 27 B.C.; nor to the name of any governor. Whereas the fact that some new colonies were founded in western Mauretania, probably in 3 3 B.C., is no proof of Octavian's intention, since we know that later, after 25 B.C., the Mauretanian colonies were administered from Spain, which shows that such an arrangement was not an institutional impossibility.¹¹ To install Juba as ruler in 33 B.C., after he had been raised at Rome in Octavian's own household, would have provoked a violent reaction among the Mauri (as indeed happened later) just at a point when the civil war was at its most critical. But so too would annexation. Octavian simply shelved a decision until 25 B.C., when, after his expedition to Spain, he saw the pressing need for action. Juba, as we shall see, was an important agent of what Augustus intended for the whole of the African settlement.

iii. augustan expansion

Very little is known of the details of the Augustan expansion. We have to be content with names on triumphal lists plus a few names in the literary sources, some of them inadequate for positive identification. Wars are recorded in 21 b.c., 19 в.с., c. 15 в.с., c. a.d. 3 and a.d. 6. The end result was a permanent winter camp for the army at Ammaedara (mod. Haidra), at the source of the river Bagradas (mod. Medjerda) on the high plains of Tunisia, and a road completed by a.d. 14 dropping down from the uplands via Capsa (mod. Gafsa) to Tacape (mod. Gabes) on the Tunisian coast.¹²

Much speculation has gone into just how far beyond this line the Roman armies advanced, fuelled by an intriguing report full of myster­ious place-names from the Elder Pliny concerning a desert campaign against the southern Garamantes by L. Cornelius Balbus, who triumphed in 19 в.с.¹³ There are also some briefer references to a victory over the Gaetulians, after they had rebelled against Juba, won by Cossus Cornelius Lentulus in a.d. 6. Between these two dates we also learn of a victory gained by a certain Quirinius over the Marmaridae and Gara-

DioxLix.43.7, liii. 12.4-6; Pliny, HNv.i. Gsell 1930 (e 741) 223,Gascou 1982 (e 738) 144 and Mackie 1983 (e 75 3) accept the brief provincial period of Mauretania from 33 to 2; b.c., perhaps governed from Spain; but the main argument, that Octavian would not have handed over Mauretania because of the propaganda war against Antony, is not persuasive, given the difficulties of annexing a huge, wild territory just when preparing for civil war in 3 3 B.C. For the colonies, see Mackie 1985 (e 753).

CIL viii 10018; EJ² 290; ILAFr 654 - Asprenas ... pr.cos ... viam ex castris bibernis Tatapes municnAam curavit. legio III Augusta (Tаса pes is an indeclinable variation of Tacape — here 'to Tacape').

ч Pliny, HNv.3j-8; Flor. 11.31; Dio lv.28.3-4. Pliny's names are analysed by Daniels 1970 (e 725) 13-16 and Desanges 1957 (e 727).

mantes (see ch. 13/, p. 635-6) and of triumphal ornamenta being granted to L. Passienus Rufus. But wild theories about Balbus' penetration to the Niger Bend via the Tasili and Hoggar Mountains can really not be credited, given the terrible problems encountered by far better equipped French expeditions to the Sahara in the nineteenth century, and we must setde for the more sober judgment that what we are witnessing is the reaction of Libyan tribes to Roman imperialism over the whole of the southern pre-desert.

The appointment of Juba II in 25 B.C. over a huge territory that extended not only to the Mauretanias (roughly central Algeria to Morocco) but also in theory along the whole Gaetulian or Numidian borders of the Roman province as far as Cyrenaica, provoked a chronic and violent response from the various 'nomadic' peoples, as Strabo calls them.¹⁴ Some of these peoples in loosely confederated groups tradition­ally migrated up onto the plains of Constantine and to the Tunisian Dorsal, recognizing no artificial frontiers. An inscription recording disturbances, which was set up by a Roman settler about a.d. 3 near the colony of Assuras (mod. Zanfur) in the rich Tunisian corn-lands, perhaps reflects the problem this caused. At all events, Cossus is said to have 'held back the Musulami and Gaetuli in their widespread wander­ing to a restricted territory and forced them through fear to keep away from Roman frontiers'.¹⁵

References to the Marmaridae, who are normally associated with Cyrenaica, and to the Garamantes of the Fezzan in modern Libya show how far eastwards these African borderlands extended - so much so that there have been hypotheses that Tripolitania was temporarily detached from the province of Africa to that of Cyrene and that there was a joint strategy conducted by the two governors. If so, it was brief and little of permanence was achieved, since the archaeology of the Fezzan and Libyan Valleys reveals no Roman contact with the hinterland before the Flavian period.¹⁶ But we can be sure that Juba's kingdom was regarded as an integral part of the defences of Africa and it was his inability to handle such a large remit that drew the Romans southwards.

The southern tribes saw Juba for what he was, a Roman agent, and they did not in any case recognize the authority of super-kings. It is not hard to see what they were fighting for. The Musulami, one of the principal names mentioned in the campaigns, controlled a region near Ammaedara, and it was here that the legion's headquarters was finally

Strab. xvn.3.7 (828c); cf. vi.4.2 (286-8C). 'Nomades' in Greek can also mean Numidian. For Juba's kingdom, see Desanges 1964 (e 728).

CIL viii 164(6; Oros. Adv. Pag. vi.21.18; Flor. 11.3.

The UNESCO Libyan Valleys Survey has been reported in successive volumes of Lilyaa Studies since 1979. It was Gsell 1930 (e 741) who first suggested Tripolitania may have been temporarily attached to Cyrenaica.

established. The road running from the base to the Gulf of Gabes constituted a check to the traditional, seasonal movements northwards of the Gaetulian Libyans from the oases and Chotts (salt marshes) of south-eastern Tunisia. Further east, Tripolitania needed protection from the Garamantes of the Fezzan and the Nasamones of the Syrtic Gulf and both must have seen the road that was completed soon after Augustus' reign along the eastern Jebel Nefoussa as a threat to their indepen­dence.[716] Whether Augustus really had in mind a grand design or was merely reacting to protect the provincials is discussed later. It is clear that he did not solve the problem.

iv. tiberius and tacfarinas

Armed resistance after the initial conquest was fairly typical of the process of pacification in most western provinces, the revolt of Sacrovir and Florus in Gaul, Boudica's rebellion in Britain, the attempt of Civilis in Germany being obvious examples. A variety of reasons for this resentment against Roman rule is given in our sources; hatred of arrogant or corrupt officials, dislike of military recruiting officers. Often, no doubt, it was sheer opportunism when Rome seemed to be otherwise engaged. But above all it was the imposition of Roman taxation on land which caused the greatest anger.[717]

The revolt of Tacfarinas must be seen within the context of the Roman advances, which brought with them steady appropriation of land, the imposition of an ordered tax system and obligations to provide recruits. Although little is known about the tax arrangements, an undated dedication by forty-four civitates of Africa to a tribune of the III Augustan legion who had conducted the census, shows the hand taken by the military in the operation.[718] Tacfarinas, a chief of the Numidian Musulami, had served in the Roman auxiliaries, no doubt as part of an ethnic unit. So we can see all the elements of imperialism which absorbed the southern Gaetulians into the Roman administration.

The cadastration of southern Tunisia for tax purposes, completed in a.d. 29/30, was probably begun as soon as Ammaedara became the legionary headquarters, since the decumanus maximus, the base line of orientation for the cadaster, was probably fixed on the conical peak of Jebel bou el Hanĉche just north of the camp.[719] The various 'Gaetulian' tribes - the name is used by the Romans loosely to mean southerners - such as the Musulami, the Cinithii, the Nybgenii and the Tacapitani, therefore, found their seasonal movements controlled by the frontier roads and fortifications. Equally provocative, they were probably expected to acknowledge the Roman puppet Juba as their overlord, which gave them common cause for resistance with the western Mauri. Juba's silver coins recording victories in a.d. 16 - a year before the date of Roman intervention — perhaps show that Juba had been trying to deal with the troublesome tribes already.[720]

It is hard to believe that the war between Tacfarinas and Rome, which eventually developed in a.d. 17 and lasted until a.d. 24 was a serious threat to Roman power in Africa. Velleius Paterculus, a contemporary, barely mentions it and, apart from brief references in later epitomators, it is only really Tacitus who gives the episode any prominence, because he was obsessed with the story of the emperor Tiberius, in whose reign the events occurred. He had little interest in the geography of the war and none in its causes. Various place-names are mentioned in the fighting - Thubuscum, Thala, Auzia, Cirta, Lepcis and the river Pagyda; various tribes like the Cinithii and Garamantes are said to have been involved.[721]But how much we can reconstruct out of this is very uncertain. Auzea, if the same as the later town of that name (mod. Sour El Ghozlane) south east of Algiers, lies 1,600 km west of Lepcis Magna. Thubuscum may be later Thubursicum Numidarum (mod. Khamissa) in east Algeria, or Thubursicum Bure (mod. Teboursouk) in Tunisia or one of half a dozen other like-sounding names. The basic fact, however, remains; the war was wide-ranging and it both implicated the Garamantes in the east and extended deep into Algeria in the west.

The fighting, which began with an attack on Thala near Ammaedara, extended to other 'cities'. This probably means that there was a series of hit-and-run raids or ra%%ia, typical of mounted nomadic people, deep into the African province. The Gaetulians eluded Roman reprisals by retreating into the 'desert', until the arrival of Iunius Blaesus, uncle of Sejanus, as the new governor in a.d. 18. His tactics, like those of his successor, P. Dolabella in a.d. 23-4, were to isolate Tacfarinas from his base by what was called 'blockhouse' strategy in the Boer War - the location of permanent castella and fortifications at 'suitable places', most plausibly at points like Kasserine, Sbeitla and Thelepte to control the passes up on to the Tunisian Dorsal, where later Roman towns developed.

The IX legion (or detachments of it) was posted from Pannonia,[722] partly to protect Lepcis Magna from the threat of the Garamantes and pardy, it would seem, to keep the Libyan peasants in check in the old province, since they erupted when the legion was withdrawn about a.d. 22. Blaesus' settlement, for which he was awarded triumphal honours that year, further deteriorated when Juba's son, Ptolemy, succeeded his father in a.d. 23 and alienated many of his Mauri troops. Despite this, Dolabella, an experienced commander on the Danube, finally trapped Tacfarinas at Auzia in Mauretania in a.d. 24, killing him and executing several of his Musulami leaders. Soon after this a Roman military prefect was set up over the nationes Gaetulicae. The end of the Musulami also brought the Garamantes to Rome to beg for peace, for which they had probably to pay by the loss of some of the territory that we now find being allotted to Lepcis Magna.[723]

From now until the end of Tiberius' rule we hear of no more African resistance, although we may suspect there were continual troubles caused by the cadastration that was carried out by the army over a great breadth of land in south-eastern Tunisia. Judging by the existing, numbered marker stones, it extended over at least 27,000 square km, as far as the Chott el Fedjaj. Although this cadastration divided the land into large blocks for the purpose of tax, there is occasional evidence of centuriation into smaller units and probable allotment of land. By a.d.29/30 the main work of survey had been finished and the marker stones, of which we have twenty surviving examples, were set up by the governor С Vibius Marsus. Dolabella, who was almost certainly the initiator of the survey, for which he had recently had experience in Dalmatia, was not much honoured in Rome but he was remembered in Lepcis.[724]

The Tacfarinas episode is less important for the threat that it posed than for the information it provides about the character of African society and frontier relations in this period. Several features need explanation: the width of native territorial alliances, yet the feebleness of the resistance; the close relationship between the desert and the sown and the effect on this of Roman intervention. The use of general terms like tota Gaetulia or Numidia by ancient and modern historians gives a misleading impression of African unity, which did not exist. Modern comparative evidence from semi-nomadic peoples of the southern Tunisian and Algerian marches suggests that 'tribes' are themselves highly unstable alliances of both sedentary and mobile fractions whose unity depends on success in raiding warfare and economic reciprocity. Rights of movement, rights of grazing and rights of exchange, which can take place over quite wide distances, are more important than complete ownership of the land and territorial demarcation.

Tacfarinas was a Numidian or Gaetulian member of the Musulami 'tribe', who, having proved his leadership in war, established far- ranging but fragile alliances with other Mauri, Gaetulian and Gara- mantes groups, based on resentment of Roman rule. His own Musulami apparently maintained specific links with people on the Tunisian and Algerian uplands, as well as operating from winter bases in the regions of the southern oases and Chotts. We are told that peasants of central Tunisia supported him, and Tacitus says that he traded for the corn that grew there. His request for a land concession for his people could mean that he wished to become sedentary; but it could just as well mean a demand for free access to historic grazing grounds. We know from Massinissa's dispute with the Punic Carthaginians that access to the 'Great Plains' of central Tunisia, the fertile uplands where the main production of wheat took place, was regarded by Libyan nomads in those days as their historic right.[725]

Almost certainly Roman property rights and boundaries were con­cepts unknown in customary practice for southern groups like the Tacapitani or Nybgenii who had had little or no contact with either Punic or Roman republican powers. The Roman term for 'marking out' (limitare) the land by boundary stones of the cadaster was a word that came to be used of a frontier and carried with it even at this stage the implication of'limiting' and controlling the movements of the southern tribal groups.[726] If Tacfarinas had used the routes south of the Aures Mountains to reach the plain of Constantine, as seems probable, then the need to control such routes must have been evident already to the Romans.

v. gaius to nero

The emperor Gaius has been credited with two important changes in north Africa: the separation of the army under its legatus from the province of Africa; and the ending of the independent status of Mauretania. Neither is strictly correct. Each stemmed from a single cause. Tacitus and Dio record the first event briefly and with contradic­tory information,[727] but they broadly agree that fear of senior senators in command of an army stimulated Gaius in a.d. 37 into separating the legion, not from the province but from the direct command and patronage of the governor, by placing an imperial legate in charge. The army remained active within the province.[728] Such a form of split command was not unparalleled nor even remarkable, and proconsular governors continued from time to time to take the military command. It is, however, geographically correct that soon after this, in the Flavian period, the army's base moved into the ill-defined region of Numidia south and west of Ammaedara, first to Theveste (mod. Tebessa) and then to Lambaesis (mod. Lambese).

The annexation of Mauretania, the huge territory extending from Algeria west of the Ampsaga (mod. Oued el Kabir) to the Adantic, was the decision of Gaius' successor, Claudius, following the war which broke out after Gaius had executed Ptolemy in a.d. 40. Exacdy why Gaius did this is a matter for debate, since Tacitus, Suetonius and Dio, our main sources for the episode, had no love either for Gaius or for a native king and they let their prejudices show.[729] Suetonius gives us a childish story about how Ptolemy upstaged the emperor at a public spectacle in Lyons by appearing adorned in a purple cloak. Dio more realistically says there were fears that Ptolemy was becoming too wealthy. Tacitus portrays Ptolemy as a weak, unpopular fop, dominated by his freedmen.

The danger of the large kingdom of Mauretania to Rome always lay in a ruler who might become too independent to control. Ptolemy's striking of gold coins, very much an imperial prerogative, suggests his assertion of emancipation, just at the time when Gaius had been badly shaken by a plot of distinguished senators on the northern frontier, one of whose leaders was Cornelius Lentulus Cossus 'Gaetulicus', son of Juba's ally in a.d. 6 and therefore heir to his father's political friends. Ptolemy himself, no doubt fearing Roman penetration further and further into Numidia and Mauretania, became a willing target for conspiratorial plans. He was, after all, Antony's grandson and cousin to the emperor. It was no coincidence that Gaul was the place to which Ptolemy was summoned in a.d. 39, since Gaius had gone there to deal with the northern crisis. The bravado of the appearance confirmed that he must go. With him went the last of the great Libyan kings.

If Ptolemy had been as unpopular as Tacitus described him, his death would hardly have provoked a violent reaction in western Mauretania, much less a rebellion conducted by one of his 'freedmen', Aedemon. One suspects that Aedemon was in reality a vassal, one of the Mauri princes at court, and that many Mauri chiefs saw in Ptolemy a symbol of their

freedom. In later years Roman governors thought it politic to honour the name of Juba II and Ptolemy with commemorative statues and one Roman pretender in a.d. 69 even took the name of Juba to win local favour.[730] From archaeological evidence it would appear that the rebel­lion concentrated on violent attacks on towns of western Mauretania, centres like Tamuda, Lixus and Kouass where Romans were no doubt trading. At Volubilis, an important centre which may have had special treaty status, Roman citizenship had already been extensively granted to local families, as we know from two famous inscriptions commemorat­ing M. Valerius Severus, son of Bostar, who raised a troop of irregular horse and was subsequently able to petition for privileges, including 'Roman citizenship' (meaning, probably, municipal status) for the town.[731]

The Roman campaign was a long and arduous affair, requiring supplies from Spain. The main details come from the Elder Pliny, a contemporary, supplemented by Cassius Dio.[732] It is clear, however, that Dio is correct against Pliny to date the war from a.d. 40 before Claudius' accession. In a.d. 41-2 the theatre extended down the Moulouya gap to the Middle Atlas and into the desert; but by 44 the campaign was over and the whole territory was annexed as two Roman provinces, Maureta­nia Caesariensis and Mauretania Tingitana, administered respectively from Caesarea (mod. Cherchel) and Tingis (mod. Tangiers).

The war against Aedemon may have been the spark to set off the Musulami once again, since we know that the future emperor Galba was appointed governor of Africa extra ordinem for two years in about a.d. 45 to deal with unrest in Numidia, a task that also took him into Juba's former territory of eastern Algeria.[733] But, as before, we should resist the temptation to think in terms of unified, African nationalist, resistance movements and see these as endemic but discrete outbreaks.

One way of checking such outbreaks was by extending Romanization through colonial foundations of Roman veterans and individual grants of citizenship, for which Claudius was celebrated.[734] Tingis was refounded, Lixus was raised to colonial status and probably, though not necessarily, reinforced. A new veteran settlement was located at Oppi- dum Novum to protect Caesarea inland, while Tipasa and Rusucurru on the coast were granted municipal status with Latin rights. Caesarea itself

was given colonial status, though again this does not imply new setders. Volubilis, as we saw, did well out of its loyalty.

Why the two Mauretanias were administered by equestrian procura- torial governors is not easy to say, particularly since senatorial legates were appointed from time to dme (e.g. in a.d. 75 and 144) and procurators occasionally held command over a united territory pro legato. Was it simply because no legion was stadoned in the provinces? Was it because Claudius saw himself as in some sense the heir of Ptolemy (on the analogy of Egypt)? Or was it because the territory was too wild to regard as a setded province (like the Alpine territories)? It is certainly true that the provinces were never much developed, their southern fronders were hard to define and communications between the two were tenuous.³⁶

Nero's contribudon to the history of Africa lay, as far as we know, in a single acdon — the confiscation of a large amount of property in central Tunisia. With colourful exaggeration the Elder Pliny says that six owners had possessed half Africa before their execution by Nero. Some relics of this brutal change may be conserved on inscriptions of the second century a.d. from the middle Bagradas valley, where an estate named saltus Neronianus is recorded in the vicinity of other estates bearing the names of old Roman families, saltus Lamianus, saltus Blandianus, saltus Domitianus?¹ There is no reason to suppose these imperial confiscations were linked to some policy by Nero to increase the supply of grain to the citizens of Rome, as some have argued (see below, p. 616).

Cruel execution of Africans, perhaps on behalf of Nero, was a reputation gained by Nero's last legionary legate in a.d. 68, L. Clodius Macer. Once the secret was out that an emperor could be created outside Rome, he developed imperial ambitions of his own having apparently already taken over as governor. In the rivalries which developed on Nero's death he tried to manipulate the grain supply to Rome for his own advantage, urged on by one of Nero's court friends.³⁸ But he was assassinated on Galba's orders and, after Galba himself fell, the province became a prey to the rival supporters of Vitellius and Vespasian. Oddly enough three of the contenders, Galba, Vitellius and Vespasian, had served in the province, the last being the least popular. But, thanks to the independent power of the legionary commander of Africa, Valerius Festus, who favoured Vespasian, the proconsul, L. Calpurnius Piso, was killed — an act for which Festus received his due reward.³⁹ It was said that

Sec the discussions by J. Marion and M. Euzennat in Bull. Arcb. Maroc. 4 (i960) 442-7,525-7.

Pliny, HN xviii.35. NTH 463 and 464 are translated and discussed by Kehoe 1988 (e 746); Carcopino 1906 (e 723). The best text is Flach 1978 (e 754).

» Plut. Galba6; cf. Tac. Ни/, i.i i. Suet. Galba 11; Tac. Hist. 1.73. J. Burian, Klio 38 (i960) 167-73 implausibly considers that Macer made common cause with senatorial sympathizers.

³⁹ Tac. Hist. rv. 38,48-50; MW 266 (showing military decorations and career under the Flavians).

Piso, like Macer before him, had been tampering with the corn supply for Rome.

In Mauretania Galba during his brief rule had given the governor of Caesariensis, Lucceius Albinus, command also over Tingitana - perhaps to counter the influence of Macer. Having gathered a large force of 12,000 auxiliaries together, Albinus declared his independence after Galba's fall and prepared to invade Spain. Assassination, however, by friends of Vitellius ended his claim.

By the time the civil wars were over Roman rule in Africa was in need of reorganization. Rival sides had offered too many tax concessions. Cities in Africa had used the wars to pursue their own vendettas, like Oea which had called in the Garamantes against Lepcis. The tension between the legionary legate and the proconsular governor had to be resolved. And we may guess that the Mauri and Numidian tribes of the interior had not remained inactive, making it a necessity to increase the security of the frontiers. Above all, the importance of protecting and encourag­ing the production of African grain and, increasingly now, oil was underlined. That was all work for the new Flavian administration.

vi. the administration and organization

of the province

Whatever Caesar intended, there is litde trace of his actual achievement. The province of Nova was, of course, his creation, running probably from the old fossa regia — the. republican boundaries taken over from the Numidian kings - westwards. But whether he incorporated the enclave of Sittiani around Cirta is unclear. Octavian immediately saw the undesirability of having two provinces of Africa, particularly if he wanted the new colony of Carthage to include the Gaetulian veteran setdements beyond the fossa regia. There is just a hint in Dio that Augustus began by giving Juba II the territory of Nova which had formerly been ruled by his father, Juba I.⁴⁰ But given the importance of the corn of Africa, the idea seems implausible.

The single province of Africa Proconsularis was therefore formed in 35 B.C., as argued earlier, to incorporate all the former territories, including that around Cirta. The only legion we know to have been permanendy stationed in the province was the III Augusta, not in fact recorded until a.d. 14, but obviously present earlier and stationed perhaps first at Carthage before moving to Ammaedara. There was almost certainly also a fair number of auxiliaries recruited locally, a normal obligation laid upon native communities in this period. But how many outside or local auxiliary units there were at this stage one can only

⁴⁰ Dio li.ij.6, LIII.26.Z.

guess. At Ammaedara we have a pre-Flavian stela recording a cohors XV Voluntariorum, without doubt Roman citizen irregulars; the ala Siliana is also recorded as serving in Africa.⁴¹ Both were probably recruited locally.

The problem of the relationship of the provincial governor to the army and the relationship of both province and army to Augustus himself is perhaps something that concerns us more than it did Augustus. Although Africa became technically a public province in the setdement of 27 B.C., the emperor's grip was always firmly on the army, where he could — and sometimes did — nominate the legionary legate, despite the theoredcal right of the governor to appoint his own legates. Furthermore, the emperor could always manipulate'appointments of governors when there was occasion for important military campaigns. Tiberius had no difficulty in 'persuading' the Senate of the wisdom of appoindng Iunius Blaesus to the post for the campaign in a.d. 18.⁴² Galba was appointed extra sortem when the need arose under Claudius. So the arrangement remained basically ad hoc, even after Gaius ended the anomaly of a legionary commander who was subordinate to the governor, while selected by the emperor.

It is impossible to talk about precise boundaries under Augustus when the territory was in the process of being defined. Some fifty years after Augustus' death the Elder Pliny preserved in his description of the Maghreb coast two undated and different lists of the colonies and towns, which have been thought to have had their origin in the early formulae provinciae.⁴³ But this seems unlikely. Part of the lists must go back to Iulius Caesar, since there is a reference to the two separate provinces of Vetus and Nova. Other parts are updated to include colonies founded by Augustus or Claudius. Many of the towns are not listed in any strict juridical or tax category but only vaguely in non-technical language as oppida. To reconstruct the Augustan settlement from this is a more or less hopeless task. But Pliny's list does provide some clues.

For the taxes of the province we have only the guidance of the few inscriptions already mentioned recording the new land cadaster in the south. In the northern part of the province there are quite extensive signs of cadastration and centuriation along the lines of the republican orientation, which may have been the work of Augustan governors in distributing lots to new settlers, since the cadaster extends well beyond the fossa regia. Between the two cadasters is a third major orientation and

⁴¹ Legion - EJ²290 (a.d. 14); CIL viii 22786; auxiliaries - CIL viii 23252, 232J j, 23646; AE 1972, 969; Tac. Hiit. 1.70. It is impossible to calculate the numbers in this early period; But see Cagnat 1913 (e 722) 107-10, 140ff; Holder 1980 (d 195) 289, 330.

« Blaesus - Tac. Ann. 111.3 2,3 j. The relationship between the governor and the legate is laid out in Dio liii.14.7 and Tac. Hist, iv.48 and discussed by Benabou 1972 (e 714).

о Pliny, NH v.1-30. Discussed by Teutsch 1962 (e 765), Brunt 1971 (a 9) App. 13.

a few smaller ones in the coastal regions of Byzacium which could be the work of early emperors.[735]

But cadasters do not tell us much about how the tax was collected or assessed. Most probably in the Julio-Claudian period, at least, there was a condnuadon of the republican system, since some of the evidence suggests that the taxation units and the agents of the pre-imperial period persisted. In other words, the mixed system established by the Agrarian Law of 111 B.C. was maintained, whereby a fixed sum (stipendium) was imposed on nadve communities or those with movable property and a tithe (decuma) on Roman purchasers of former public land. This was in addition to the pasture tax on animals. In the absence of other evidence we must assume that new Roman settlers were treated like those already there, unless they were veterans. The latter were granted tax immunities by Augustus for their own lifetime and that of their children, according to a papyrus copy of his edict in 31 B.C. But it is possible that the Gaetulian veterans of Iulius Caesar had been given some form of tribute immunity for their heirs in perpetuity.[736]

If this is correct, taxes would have been farmed out from the quaestor's office to publican entrepreneurs, who bid for the contracts based upon block assessments of the native civitates stipendiariae. It is unfortunate that the only real evidence of such an arrangement — a dedication by the mancupes of the stipendiarii to the quaestor - cannot be dated, although it is probably Augustan. The civitates, however, do not seem, as later, to be independent tax communities but more like villages grouped together into rural districts, as they had been under Punic Carthage. In fact, we have Roman inscriptions referring to these old Punic land divisions — called in Latin pagi. One such records the sixty- four civitates of the pagus of Tuscus and Gunzuzus, recalling Appian's description of the Punic 'land of Tusca' with its fifty towns. We also have an inscription dating from soon after the refounding of Roman Carthage mentioning an administrative district of eighty-three castella under a law officer of Carthage, M. Caelius Phileros, who was responsible for allocating their taxes quinquennially.[737]

As Roman rule extended southwards, the military districts were possibly regionalized in the same way, since an inscription commemor­ates the census of forty-three civitates taken by a tribune of the III legion (p. 593). We also know of the existence of an imperial procurator in the 'plain of Byzacium' under Augustus, perhaps a military supply officer or an agent of imperial estates. None of this, however, is much to go on.

VII. CITIES AND COLONIES[738]

It is a banality that the Roman empire was fundamentally no more than a collection of city-states, around which the emperor provided a protect­ing frontier that was paid for by their taxes. The city or civitas, therefore, was the administrative unit upon which the empire depended. The problem for the Romans in Africa, different from other western provinces, was not to persuade scattered, rural communities to collecti­vize into city units, as in northern Gaul and Britain, so much as to find a formula that would organize the scores of small, independent villages and hill-top forts that already existed into manageable communities. Despite the Carthaginian coastal cities and a few native centres (like Thugga), one hundred years of Roman republican rule had done little to advance this process.

What changed all that was the civil wars and the Principate. The wars created a desperate need to demobilize and provided the land on which to setde veterans. Augustus possessed the will to order such events and the self-interest to know that his political survival depended on satisfy­ing this need and on supplying the volatile population of Rome with regular food. Colonies, communities and corn were the informing principles of Roman imperialism in Africa.

Colonies first. Apart from Carthage, Iulius Caesar and Augustus between them founded some twenty-six to twenty-eight colonies the length of the Maghreb. They cannot all be assigned with certainty to the Caesarian and Augustan periods — some may have been founded just after Augustus' death.[739] Their names and those about which there is greater certainty than others, are marked on Map 13. Some of these setdements obviously had a defensive, military purpose that was usual when veterans were kept together in their original army units. Soldiers of the thirteenth legion were established at Thuburbo Minus (mod. Teboura) and at Uthina (mod. Oudna) as buttresses for Carthage, controlling the southern and western plains of the Medjerda and Miliana rivers. In the same way the colonies at Zuchabar (mod. Miliana) and Aquae (mod. Righa Hammam) protected Caesarea, Juba's capital, from inland raids in Mauretania. The colonies along the Algerian and Moroccan coast were useful ports of communication, just as the colonies along the coast of eastern Tunisia and Cape Bon controlled the former Punic ports. In former Africa Nova, the Punic town of Sicca Veneria is the only colony to commemorate Augustus as conditorf⁹ all the other colonies of Thuburnica, Simitthu and Assuras may have started life as military or veteran satellites of Sicca. Ammaedara, the legionary base by A.d. 14, became a colony under Flavian rule.

After the annexation of the Mauretanias, Claudius continued the policy of colonial foundation (p. 598), although in many cases it was more a matter of raising the status of towns rather than actually sending out new settlers. This is testimony to the Romanization and unofficial immigration during the rule of Juba and Ptolemy. Oppidum Novum, which now became a colony of veterans, may have started life as a garrison of Roman auxiliaries to help Juba, since we hear of a curator of a fort there.⁵⁰

At Iol Caesarea (mod. Cherchel) Juba, in imitation of other hellenistic rulers, deliberately constructed a show-piece city, laid out on an orthogonal plan, with a number of monumental buildings.⁵¹ Most important were the temples, including a temple of Augustus, of which a colossal statue of the emperor survives, showing the deliberate political intention of bringing urban Roman culture to the Mauri, as well as organizing the resources of the countryside. The grant of colonial status did not necessarily involve any new settlement but Italian craftsmen may have come to produce pottery in the city. The many Roman names inscribed in the city, almost certainly from the period of Juba, include some who were probably Italian negotiatores.

As to the numbers of Roman settlers in each colony, best estimates suggest a figure of about 300 to 500 adult males, giving a total for the Augustan colonies of some 8,000-13,000 families. This is not counting Carthage or Cirta, which are discussed below. But a colony's territory was not only occupied by Roman settlers from Italy. The land cadaster from Arausio in Narbonensis and the manuals of Roman surveyors show that native inhabitants remained. Some of the elites were given citizen­ship and formed joint communities, as happened at the veteran colony of Emerita in Spain.⁵² Native wives of veterans were granted citizenship, too, quite apart from the fact that many of Caesar's and Augustus' veterans had themselves been local native recruits for the emergency of the civil wars in both the legions and as Gaetulian auxiliaries.⁵³

« CIL vin 27568. ⁵⁰ AE 1926, 23.

⁵¹ Gseli 1930 (e 741) 206-84. Recent work is in Leveau 1984 (e 752) and Benseddik, Potter forthcoming (e 717).

^s2 Brunt 1971 (a 9) 246-61; Romanelli 1959 (e 760) 207. Grom. agrimens. 155. 6-8 (Lachmann); Strab. in. 2.15 (151c).

⁵³ Grants to Octavian's veterans included citizenship ipsis,parentibus liberisqtte eorum et uxoribus- EJ² 302; and probably the right to join in a colony or remain in a native community; cf. FIRA15 5.

So we must not overestimate the cultural impact of the new founda­tions. Colonial status, Roman citizenship and often large plots of land — a third of a century, 16 hectares was the least an Augustan veteran could expect — created a privileged minority, loyal to the settlement and anxious to prove their Romanness. But they were men who were often linked culturally to the local population by language, religion and custom. We see how commonly the name of Iulius was taken by new citizens in a colony like Sicca Veneria - overall about 20 per cent of recorded names. At another colony, Simitthu, the name Iulius Numidi- cus (which occurs twice) speaks for itself.[740] The oldest inscriptions record religious homage from the Algerian colony of Rusguniae to the Mauretanian king, Ptolemy in a.d. 29 and to the African god, Saturn; but the latter is honoured in his Romanized form and the prominent families who make the dedications also betray their new Roman status by their names.[741]

Carthage was quite different from the military colonies on the coast. Appian, describing its foundation, says that he had 'found out that Augustus gathered together some 3,000 Roman colonists and the rest from those dwelling around (perioikoi) in the region'. The easiest interpretation of this statement is that Roman immigrants plus Romans already in the territory came to a total of 3,000, to which were added native peregrines. There were certainly some veterans of Caesar included but the unusually large proportion of freedmen recorded in civil and religious officies in the early colony suggests that many of the immi­grants came from the city of Rome. Among country folk, Virgil tells us, there was no great enthusiasm to go to 'thirsty' Africa.[742]

Archaeology gives us some idea of what the early colony of Carthage was like.[743] The most interesting feature is that, despite the earlier, different Gracchan cadastration, the city was refounded on the old Punic orientation and made much use of Punic foundations, building material and cisterns that had lain unused or in ruins since 146 B.C. The Punic citadel on the Byrsa was the central point for the centuriation of the town and the hill itself began the first stage of its transformation as the monumental focus for the city. Little remains of the Augustan city but there are signs that a start was made on the dramatic levelling of the citadel summit and infilling on top of Hannibal's city on the south side. It was on this site that a huge new forum centre was to be created in the second century. Near here, appropriately, an altar to the gens Augusta was put up on private ground by one of the first generation of settlers.

The monumental preparation of the Byrsa, however, contrasts with the tentadve and poor buildings of the shore and harbours. Mud brick and unpaved streets suggest that the early colony was quite a humble affair which grew only gradually. Some of the Punic ruins were not rebuilt for two generadons and the early Roman cemeteries were inside what was later part of the city street grid. Virgil's romantic portrait of colonists constructing Dido's first city — the great citadel, the paved streets, the gates and the theatre — which was probably written to celebrate the Augustan colony, did not exactly resemble the reality.

The difficult feature of the foundation for us to understand is how the rural settlements within the territory or pertica of Carthage, which stretched at least iookm down the Bagradas valley, were organized. A series of inscriptions in the earlier imperial period record communities of Roman citizens, confusingly called pagi but nothing like the other pagi districts of native communities (p. 602). These were single enclaves, many of them bunched together just beyond the old fossa regia boundary of Africa Vetera in the fertile middle Bagradas and Siliana valleys at places like Thugga, Uchi Maius and Thibaris.[744] They were part of Carthage, containing citizens and later even magistrates of the colony, administered by a 'prefect of justice' from the city. But at the same time they were quartered on top of native settlements.

Some of these pagi, like Uchi Maius and Thibaris, but not all, are recorded by the Elder Pliny as oppida civium Romanorum. The inscription noted earlier of M. Caelius Phileros,[745] who had been a freedman attendant of the governor before 40 в.с. and had then joined the new colony to become aedile, prefect of justice and officer for taxes in charge of 'eighty-three castella' (native sites), is matched by another inscription from Uchi Maius, damaged but probably of Phileros, too, recording his arbitration between the coloni and the local Uchitani. Not only were the gentile names Marius and Iulius common in these communities but several explicitly honoured Marius when they later acquired a municipal charter.

Pagi are recorded within the old province, too.⁶⁰ Two of them, Saturnuca and Medeli, are not far from the colony of Uthina in the Miliana plain; both have inscriptions stating they were veteran settle­ments; and one claims Augustus es benefactor. A similar settlement was at Hippo Diarrytus (mod. Bizerta) and two others were near Thabraca. Apart from these pagi, other types of Roman communities appear on inscriptions:⁶¹ 'Roman citizens who are in business (negotiantur) at Thinissut' on Cape Bon; 'Roman citizens who are living at (morantur) Suo' in the Bagradas plain; 'a community' (conventus) of Romans and Numidians who live at Masculula, west of Sicca Veneria.

Finally, there is the puzzling relationship between Carthage and a number of sites on the Tunisian coast which eventually became colonies bearing the title of 'lulia' in their names. The inscription of Phileros, examined earlier, records his career not only as a magistrate at Carthage but also twice as chief magistrate (duovir) of Clupea (mod. Kelibia) on Cape Bon. Clupea is listed by Pliny as a 'free oppidum' but it became a Julian colony. That is also the case at Curubis, Neapolis and Carpi (not 'free' in the last case), two of which also had magistrates, freedmen again, who held office at Carthage.

As far as we can tell, Cirta also, was given a very large territory, administered by prefects and subdivided into pagi and other communi­ties where Rpman citizens lived attached to the main colony. Augustus is known to have taken ever the former Sittiani veterans and to have supplemented them with further colonists in 26 b.c. That, as far as we can judge, was also the origin of the special relationship of contributio recorded in the second century a.d., by which three of these sub- communities had rights of interchanging magistracies with Cirta when they later became colonies.⁶²

There is some suggesdon that this was the earliest form of organiza­tion at Sicca Veneria, too — the only one of Augustus' colonies in Africa Nova to figure in Pliny's list, antedating the colonies of Thuburnica, Simitthu and Assuras. Two of these later colonies are called oppida civium Komanorum by Pliny, so may have begun as pagi just as we have records of pagi and other types of small Roman communities near Sicca. The fact that some of these pagi appear to have been on sites that also contained castella may mean that an elaborate system of Roman settlements {pagi) was constructed to supervise native hill-forts. At Cirta, however, the castella were part of thepagus and look like fortified points of security for the early colonists.⁶³ Despite this confusion, however, both Cirta and Sicca look surprisingly like Carthage.

So what can be made of these scraps of information? There is no need to read into Carthage's foundation some sort of new, super-hellenistic model city, since contemporary Augustan colonies with similar exten­sive territorialpertica are known in France (Arausio) and Spain (Emerita)

" EJ² 106; IL Tun 682; EJ² iii.

⁴² ILAtg ti(i) 36; AE 1955, 202; ILA/g 11(1) 3596.

⁶³ The latter is argued by Gascou 1983 (e 759), against Beschaouch 1981 (e 719). The Phileros inscription above (CIL viii 26274) shows him demarcating land of the castellum at Uchi between the native Uchitani and the coloni, where there was a Roman pagus. This suggests that a castellum can be either peregrine, Roman or both.

- the latter also administered by prefects. Pagi as subdivisions of a city's territory were also perfectly normal. But it is difficult to resist the conclusion that at Carthage, in addition to the new, veteran settlers, most or all of the pagi beyond the fossa regia were not new settlements at all but villages of Gaetulian auxiliaries who had been rewarded with land and perhaps citizenship by Marius and Iulius Caesar and whose families were now incorporated as citizens of the new colony. Perhaps the same was the case with the Sittiani at Cirta. As for the coastal Julian colonies, they may have kept some sort of special relationship with Carthage after they became colonies in their own right.

In addition to colonies and the oppida civium Romanorum (probably pagi), Pliny lists other categories of communities within the African province — about which it is difficult to say anything of their organization or physical appearance. There are a number of oppida libera, some simple oppida and civitates, one oppidum stipendiarium, one oppidum Latinum and finally Utica, the former provincial capital, which, Pliny said, had Roman citizenship — presumably as a recognized municipium Romanum. In other words, apart from Utica and the single town with Latin rights, all the rest were what would be juridically classified as peregrine (native) civitates, towns with their own territories which were self-governing and recog­nized within the formulae provinciae. We have the record of two of them, Thysdrus (mod. El Djem) and Hadrumetum (mod. Sousse) in a land dispute about the middle of the first century a.d. Some of them, former royal strongholds of the kings of Numidia, such as Zama Regia, Hippo Regius, Bulla Regia, were probably the creations of Caesar in his new province of Africa Nova. Others along the Tunisian and Tripolitanian coast had a long Punic and republican history of urbanization.[746]

We can only guess whether Pliny's list was complete or what exactly the differences were between his categories. Lepcis Magna, for instance, the rich Punic centre of olive oil export, was only called an oppidum, while its rival Oea was called a civitas. Was Lepcis disgraced for opposing Caesar? If so, it was rehabilitated by Augustus, since it had the right of a 'free town' to strike its own coins and there began to appear a number of spectacular, public Roman-styled buildings as early as 8 в.с. But for Thugga, which had been a royal capital of Massinissa and where a pagus of Roman citizens was installed, Pliny gives no evidence of civitas status, which is not recorded before Claudius. Yet Thugga almost certainly was recognized before that, since we hear of the Thuggenses commemorat­ing a governor in a.d. 3. The same is probably true of Musti nearby.[747]

Puzzlingly, on some early inscriptions we find small villages being called civitates when they were clearly not recognized communides.⁶⁶ So the confusion is considerable.

Many of these small village communides were strongly Punicized and continued with their own Punic magistrate, called sufet (pi. sufetim), long into the Empire without any evidence, in some cases, that they were ever recognized as independent cides by the Julio-Claudian emperors. There was a number of them just west of, and possibly including, the later colony of Thuburbo Maius, in the rich Miliana valley 50km south of Carthage. The interesting suggesdon has been made that these were the original inhabitants of Carthage, exiled when the city was destroyed ini46 в.с. If so, they were appropriately incorporated once again into the great pertica of Carthage, though only as peregrines. Another group of communities, many of them near the sufet villages, possessed governing councils called in later periods undecimprimi which may also have had Punic origins.⁶⁷ We may guess that life in these villages or small towns was much as it had been when Punic Carthage collapsed over one hundred years earlier.

We must not forget the southern territories of Tunisia, the land brought under Roman control by Augustus and Tiberius south east of the legionary base of Ammaedara. Pliny describes some of the communi­ties as 'not so much civitates as natiortes'. In other words, tribes like the Musulami, Capsitani and Cinithii, were recognized but not as urban units, despite the fact that some of the leading families among groups like the Cinithii had been strongly Punicized and quickly accepted Roman urban structures, too. But until their centres of Gigthis (mod. Bou Grara), Tacape (mod. Gabes) etc. were given recognized status, they were probably put under the control of a military prefect. One such person, C. Flavius Macer, prefect of the Musulami in the later first century or early second century may himself have been a native leader, who became an officer of the auxiliaries and was given citizenship by the Flavian emperors.⁶⁸

It is thought, too, that many of the later towns on the edge of the Tunisian dorsal, places like Cilma (mod. Djilma), Sufetula (mod. Sbeitla), Cillium (mod. Kasserine) and Thelepte, became Romanized through soldiers or veteran stationed there to control the routes in this period. That was certainly true of Thala, the former Numidian strong­hold near Ammaedara, according to Tacitus, and perhaps of the oasis

" CIL hi 338.

Discussion of Punic sufet towns by C. Poinssot, Kartbago 10 (1959-60) 93-131. The Punic origin of umltcimprimi is discussed by B.D. Shaw, Museum Afrieum 2 (1973) 1—10 — but this is controversial.

Macer-ILAlgi 285, NTH 260; cf. Pflaum CP no. 98. Gigthis had long been a Punic port and centre - N. Ferchiou in Picard 1984 (e 758л) 65-74.

centre of the Capsitani at Capsa (mod. Gafsa), which had also been Punicized.[748]

Further west in Mauretania, in addition to the colonies, Claudius granted a limited number of municipal rights and recognized a few communities as civitates, mainly places along the coast which had a Punic background. Tipasa, for instance, became a Latin town, Rusucurru (mod. Dellys) became a Roman town. Volubilis was rewarded for its loyalty by getting Roman municipal status and Tucca (mod. Zucca) on the border of the African province became a civitas. But not very much seems to have been done to develop the interior before the Flavian emperors.

viii. romanization and resistance

Two conclusions follow from these administrative arrangements for the provinces of north Africa. First, since the number of new Italian immigrants was relatively small, their impact was less dramatic than has sometimes been supposed. Secondly, the local African elites, including those who were incorporated in the colonies, many of whom had long been Punicized, were those most readily integrated into the urban system. Both these conclusions contribute to our understanding of the process of Romanization.

The precise juridical status of a community made little difference to the realities of life in the small castella and vici of the countryside, which continued to be administered by their own principes, magistri and seniores and which still thought of themselves in terms of their own sub-groups (domus and familiae) and tribal alliances (gentes). In many parts of Africa these categories persisted until the late Empire. Pomponius Mela, a geographer in the middle of the first century a.d., wrote of African society as made up of nomadic wanderers and the rural masses (vulgus), still living in their huts (mapalia) under their own leaders. We have a good collection of Libyan funerary inscriptions from a region of eastern Algeria around Hippo Regius (mod. Annaba), which date in some cases from the Roman period, since they are bilingual. They show that, even in the case of a man with as Romanized a name as C. Iulius Gaetulus, who looks like a veteran, he was also a chief of the Misiciri group and lived in a traditional, Libyan-speaking community.⁷⁰

Many of these Libyans simply continued to be peasants working on the lands of their former chiefs. Iulius Caesar had allowed his friend, C.

Iulius Massinissa, to keep land that probably belonged to his royal ancestor. In the pagus of Abuzza, near Sicca Veneria, an inscripdon commemorates a woman, Maria Plancina, who is called 'foremost of all Numidian women, descended from a royal family', whose daughter married a large landowner, Licinius Fortunatus. Around Sicca especially we have a number of inscriptions of later periods showing that castella under their own seniores continued to function.[749] So, one wonders what difference the peasants working on these lands would have noticed when the owners adopted their Roman names. The continuity of dependent relations between the rich and poor Libyans must explain why there is so little evidence of Roman-style imported slavery on the land.

On the other hand the history of two former Numidian royal towns, Bulla Regia and Thugga, shows how quickly native towns adopted Roman styles of building and culture.[750] Bulla, recognized as a 'free town' in the Augustan province, contained many Romanized families bearing the name of Iulius, including some of the most prominent, that go back to the earliest period. Fairly soon we see the Roman reticulated technique being used for a public building in the centre, showing how urbanization was fostered by Roman rule.

Thugga, which had probably been Massinissa's capital in the second century b.c. and had long ago acquired administrative institutions, much influenced by Punic culture, plus a number of monumental buildings, was now increased by a Roman pagus. The enclave adjoined the old native town and a Roman type of town centre with forum and market began to be laid out in this early period. But the rapid Romanization of the civitas was partly due to the domination of two prominent Numidian families, the Gabinii and the Iulii, who had probably been given citizenship by Caesar or Augustus and prided themselves on it. An inscription dating from a.d. 48 records the dedication by the patron of the pagus, a citizen of Carthage, of a temple which had been paid for by the local magistrate, Iulius Venustus, husband of Gabinia Felicula; Venustus had served as flamen of the Roman imperial cult, like his father, Faustus Thinoba, before him - for which both had been given the honorary Punic title of sufet.[751]

The double communities of pagi and civitates, therefore, speeded up the Romanization of the African elites. The exact constitutional relation­ship between the two groups became complex as time went on, since Roman citizens of the native town sometimes married Romans of the pagus and acquired land there. On some inscriptions the words utraque pars civitatis or uterque ordo give the impression that the two formed a single civic community, even though the pagus continued to exist as a territorial offshoot of Carthage.⁷⁴ The reason why the two did not coalesce is clear. Holding land in the pagus was a jealously guarded privilege, which carried with it the economic advantage of tax immunity granted to the original members. On an inscription of the second century a.d. a man significantly called Marius Faustinus (a Marian veteran family?) at Thugga proudly calls himself 'defender of the immunity of the pertica of the Carthaginians' after a mission to have this immunity confirmed. Apparently, unlike later, Augustus or Caesar had granted these settlers (perhaps limited to descendants of the Gaetulian veterans) tax freedom for their heirs in perpetuity.⁷⁵ If this interpretation is correct, such a valuable asset would have widened the economic gap between Roman colonists and most natives.

The social and political benefits of the Augustan system for the elite had already become apparent in the Julio-Claudian period. Under Tiberius a citizen of Musti, L. Iulius Crassus, reached equestrian status and under Vespasian the first known African consuls, Q. Aurelius Pactumeius Fronto of Cirta and his brother Clemens, were created.⁷⁶ All may have been Italian emigre's, but Iulius Crassus looks like an enfran­chised Libyan. The patronage of high Roman officials, such as the governor, was a valuable asset that encouraged native elites to imitate Roman institutions, as we see when the people of Thugga commemor­ated their 'friendship' with the governor in a.d. 3.

Lower down the scale, too, the patronage of Roman officials must have encouraged Romanization. As early as 12 в.с. we have an inscription set up by 'the senate and people of the civitates stipendiariae in the pagus (of) the Gurzenses' to record their formal clientela links with the ex-proconsul, P. Sulpicius Quirinius. One cannot miss the obvious attempt of these villagers who lived in the region of Hadrumetum on the Tunisian coast to prove their Romanness by their high-sounding institution and it contrasts starkly with the names of the men com­missioned to set up the inscription, and with the native oppida from which they came, Ammilcar of Cynsyne, Boncar of Aethogursa, and Muthunbal of Uzita. A similar inscription records a former officer of the III Augustan legion living at Brixia (mod. Breschia) in north Italy, who had hospitium relations with four tiny African communities in the Miliana valley.⁷⁷

The service of Africans in the auxiliaries, whether in ethnic units like

Thugga - CIL viii 26)91, 26615; Thignica - CIL vm 15 212.

NTH 510; for veterans, see n. 53.

Iulius Crassus - CIL VIII I5)i9and 26475; ILTun 1393. Cirtan senators - MW 298; ILAlgn (i) 642.

Two of the inscriptions are recorded in EJ² 334-5; for the set, see CIL v 4919-22.

the cohors Musulamiorum that were recruited quite early on, or in mixed units, was another way in which Roman ideas were transmitted. In a.d. 69 'a large number' of Mauri were serving in the twenty-four units of Albinus, although we have no details.⁷⁸ Gaetulians and their influence within the pagi have been mentioned several times. But the revolt of Tacfarinas, who had served in the Roman war should warn us not to exaggerate the effect of such indoctrination.

The fact is that the pre-Roman culture of Africa, including the strong Punic and hellenistic elements, inevitably remained embedded in the make-up of the new provincial society, not just at the level of the poor but of the rich, too. The Thugga inscription noted above records a man who was priest of divine Augustus but also honorary sufet. At Volubilis in Morocco, the local dignitary, M. Valerius, son of Bostar, had probably become a Roman citizen, as were many others in the town, before it was incorporated into a province; as such, he held the office of sufet originally and became the first duovir andflamen when the town became a municipium (see p. 598). At Lepcis Magna, which had been an important Punic port, an inscription was put in Latin to commemorate the market built by Annobal Tapapius Rufus, son of Himilcho, one of the leading families of the town. He also held the office of sufet and added to the Latin inscription another one in Neo-Punic.

The emperor-cult, as we can see from these examples, was a vehicle by which local aristocracies demonstrated their Romanness and should not be regarded as insincere flattery or impositions by the state authorities. At Carthage, for instance, the altar of the gens Augusta was explicidy on private land. Elsewhere the dedications look like isolated enclaves of Roman citizens asserting their identity in predominandy Libyan towns - at places like Thinissut, Thysdrus or Vaga. Soon civitates themselves took the initiative, as at Mactar, to set up a temple as part of their civic cult. At Lepcis Magna the imperial cult went together with the monumental transformation of the Punic town into a Roman city; statues of the imperial family were set up in the temple of the new forum and the rich elite who paid for the great new theatre and markets - men with names like Iddibal Tapapius, son of Mago — were also those who took on the priesthoods of the cult.⁷⁹

There has been much debate about the survival and continuity of African and Punic political organizations within the Roman provincial towns.⁸⁰ Apart from the sufet magistrates, who are found in every part of north Africa where Phoenician setders had preceded the Romans, the

⁴ Tac. Hilt. 11.58-9. cf. Benseddik 198» (e 716).

" Early imperial cult inscriptions - ILA/r 306; EJ² 106; some may be only honouring Augustus - 1439a, 22844; city cult - IRT 273; EJ² 105b; IRT 321-3. See Smadja 1978 (e 763).

¹⁰ Gascou 1976 (e 736), against Kotula 1968 (e 747). See also above, n. 67.

Thugga inscription above also refers to a decree voted by 'all the gates' (omnium portarum sententiis) of the town. Whether this organization was the same as the Punic mi^rah - brotherhood with religious affiliations - which persisted into Roman times at Mactar and Althiburos on the Tunisian high plain, or whether it was an exclusively Libyan council meeting hardly matters since pre-Roman African life was already a cultural fusion.

Similarly, the popularity of Afro-Punic cults in Roman Africa shows how a new amalgam of provincial culture was emerging.[752] The pre- Roman cult of the Cereres corn gods became one of the most prestigious in Roman Carthage, its priesthoods dating back to beyond the formal foundation of the colony itself. The cult of the earth goddess Tellus, which was probably practised near the altar of the gens Augusta on the Byrsa, where her statue was found, was elsewhere assimilated with the local African divinity Gilva. The Libyan bull-god Gurzil (compare the name Gurzenses, above) was used as a motif on a Roman lamp in first- century Carthage. And we have already seen how the Saturn cult, a Romano-African version of the important Afro-Punic cult of Ba'al, was popular with the earliest Romanized African elites in the colony of Rusguniae. Above all, the Punic moon-goddess Tanit never ceased to be venerated in Roman Carthage in her Romanized form as Dea Caelestis. The child sacrifice associated with this cult was carried out 'openly', according to the African, Christian writer Tertullian, until 'the procon- sulship of Tiberius' - presumably he meant the emperor Tiberius - and it is clear that thereafter it continued clandestinely.

Whether examples like this represent a form of passive resistance to Rome or the steady progress of Romanization is to some extent a matter of semantics.[753] Advocates of the 'resistance' model regard Romanization like a layer of paint which was easily stripped off later when Roman rule deteriorated to reveal the true Africa lurking below the surface. Modern studies of acculturation, however, demonstrate not only how compatibi­lity varies enormously according to the social class and the isolation of individuals, but how even in indigenous resistance movements (cargo cults and the like) the language is not so much that of the old culture surviving beneath a veneer as that of a new vocabulary which emerges from the fusion of two civilizations, preserving elements of both. Romanization and resistance were two sides of the same coin.

In the Julio-Claudian period there was still an active, physical resistance among the southern, semi-nomadic populations and the montagnards of central Algeria and Morocco which spilled over from time to time even into the heart of the old province. We may assume that some Libyan leaders bitterly resented Roman rule because of its interference in their own power and that many Libyan poor were virtually untouched by it. This continued in remoter communides until the last days of Roman rule. But the success of Roman provincial rule lay in its capacity to capture the allegiance of the African elites by its laissez-faire attitude to administration and local autonomy, while providing financial and social rewards to those who were prepared to participate in the system. Urban government under Roman rule was by and for the rich, reinforcing social inequalities and leaving social relations with the rural poor much as they had always been.

ix. the economy

It is not easy to judge how much the economy - and especially rural production — changed during this period. Presumably the trends already in motion under the Republic continued. Not surprisingly most of the information from the late Punic and republican periods relates to the production of grain, which is also the subject of dominant interest in the early Empire. The extraordinary productivity of the soil of Africa, and notably that of Byzacium — the south-eastern coastal region of Proconsu- laris - was a byword in Rome. But the notion, derived from the Elder Pliny, that Africa was entirely dedicated to the crops of Ceres is a misreading of the text and explicitly contradicted by the many references we have to oil, wine and garden produce.⁸³ Archaeology is increasingly confirming the importance of oil production and its continuity with the Punic tradition, particularly in the region of Tripolitania and, probably, of Byzacium. This also fits the evidence of the Punic period, when the hinterland of Carthage and Cape Bon were noted for mixed farming, and from where, we may suppose, Mago derived the experience for his famous treaty on estate management which enjoyed such respect in Rome in the first century a.d.⁸⁴

The popularity of Mago's treatise suggests the influence of Punic farming methods on early Roman settlers. And that in turn indicates the principal development of this period — the growth of large estates and villas of the sort encountered by Iulius Caesar on the Byzacium coast. Sale of land under the Republic, plus the allocation or sale of confiscated land after the civil wars, must have accelerated the process which led the Elder Pliny to report that before Nero's confiscations half Africa was

Productivity - e.g. Varro, Rust. 1.14.2, Pliny, HN xvni.94-j, Columella, Rust. i.pr. 24. Ceres - 'Cereri to turn id nature concessit, oleum ac vinum non invidit tantum', Pliny, HN xv.8, but tantum means 'almost'; contra, Columella, Rust, xi.2.80, Plut. Cats. 35.

Archaeology, van der Werff 1977/8 (e 769), Aranegui and Hesnard forthcoming (e 713). Mago — Heurgon 1976 (e 744).

owned by six landlords. Whatever the exaggeradon, it was to Africa that imperial writers regularly turned to illustrate a land of large estates. Petronius imagined Trimalchio and his guest as owners of vast proper­ties in Numidia and Africa, while Seneca moralizes about the thousands of tenant coloni working for single landlords in Africa.[754] The emperor himself, of course, was one such estate owner. It was an imperial procurator who brought to Augustus and to Nero prolific ears of corn to demonstrate the fertility of the soil; and the first slave-bailiff of an imperial estate is recorded in the region of Calama (mod. Guelma) in Nero's reign.[755]

Many of these property owners were, like the emperor, absentees and it is not evident how their estates were organized in terms of labour or produce. Petronius talks of an army of slaves and Seneca of tenant farmers. The latter were certainly more common in later periods and there are a priori reasons, given earlier, for thinking this was always the more usual type of farm worker. But in neither case is there any real reason to believe that the growth of large estates radically altered - let alone ruined, as Pliny says - African farming methods or productivity. What it did was to change the social balance, by concentrating wealth in the hands of a minority and by providing them with the means to pay for the growing number of expensive, public buildings in towns such as Thugga, Lepcis Magna or Carthage, which have been noted already. That is, of course, when the profits did not leave Africa to pay for the expenses of the aristocracy and emperor in Rome. By expanding southwards and westwards the Roman-African economy was reaching a point where it was about to become a major supplier of the empire as well as of Rome.

x. roman imperialism

The Roman conquest of the Maghreb in the first century a.d. began as the by-product of civil war and ended up with the acquisition of new territories as African chiefs and princes were swept up in the turmoil. Octavian's defeat of Antony led direcdy to the southern 'Gaetulian' problem, drawing Roman arms as far as the pre-desert. The allied Mauretanian kingdoms of the west were an unstable solution to this involvement which eventually broke down under Claudius and led to the annexation of two more provinces.

The question is, did the Roman emperors have a coherent policy of imperialism that went deeper than this kind of reflex reaction to emergencies? Even in the republican period, when Africa was relatively neglected, we can see that the territory was regarded as a source of private wealth in land and of public food and oil. The climax was Caesar's public announcement of the acquisition for the Roman people of 8,000 tonnes of grain and 1 million litres of oil from the new province he had acquired, 'in order to impress the people with the size of his victory'.⁸⁷

That tradition of public patronage was continued by Augustus who boasted in 23 B.C., for instance, that he had made a grant of one year's ration of corn to 1 million Romans, as well as claiming to have saved the city on various occasions from corn shortages. The emperor was fully aware of the 'fear and danger' which could lead to city riots if supplies broke down. In a.d. 51 the emperor Claudius came uncomfortably close to being lynched when it was correcdy rumoured that the warehouses of Rome were almost empty.⁸⁸

Given this background of propaganda and need, it would have been surprising if the corn of Africa had not figured somewhere when emperors pondered the prudence of military campaigns, even if our sources do not specifically link it to southern conquest. Can it be only chance that the raids of Tacfarinas deep into the African province and the consequent wars between a.d. 19-24 coincided with a sharp rise in the price of corn in a.d. 19, which remained high until about a.d. 23 or 24? Tacitus himself was in no doubt about Rome's dependence on Africa (and Egypt) for her livelihood nor about the strategic importance of African grain in the civil wars.⁸⁹ All Rome knew the value of Africa.

The central importance of African and Egyptian corn in supplying Rome is confirmed by two much discussed ancient texts. The first, referring to Nero's reign, states that Africa maintained the people of Rome for eight months of the year and Egypt for four; the second that Egypt in Augustus' rule provided 20,000,000 modii (about 130,000 tonnes) of grain for Rome. Unfortunately, there is no basis here for a simple mathematical calculation, since a regular annual import of about 40 million modii of grain would have far exceeded any calculable consumption rate of Rome's population, even if this were the only source of supply.⁹⁰ Nor does the need to increase the annona supply or imperial largesse provide a plausible reason for Nero's confiscations of senatorial estates in Africa, since it falsely assumes that productivity

" Plut. Cats. 55; Haywood 1959 (e 743) 21.

•> Augustus, RG j, ij.i;DioLv.26.1 - 27.3. Tac. Ann. хп.₄з;с£ Sen. DeBrev. Vit. 18. Claudius - Suet. Claud. 18.2.

" Tac. Ann. 11.87, rv.6, xii.43; Hist. 1.73,111.48, iv.38.

⁹⁰ Joseph. BJ 11.383; [Aur. Vict.] Epit.deCats. 1.6; Haywood 1939(£743)43; Picard 1936(e756); Lassere 1977 (e 749) 296. Garnsey 1983 (d 130) 118-19.

increased when property passed under imperial management or that free corn distributions dramatically increased under Nero at the expense of the market.

What is clear, however, is that African corn was always a vital imperial asset, a weapon of control in the emperor's hands and a commodity for which there was a chronic need in Italy. The unreliability and wild fluctuation of grain yields in the pre-industrial Mediterranean are well known. 'Poverty and uncertainty of the morrow', says Braudel, were endemic pressures in the Mediterranean world that underlay 'certain, almost instinctive forms of imperialism'.⁹¹ Augustus' and Tiberius' push to the southern pre-desert more than doubled the arable area of Roman Africa. Claudius' annexation of Mauretania added to the source of frumenta fiscalia that could be and was sometimes used, while protecting the western flank of the old province. These may not have been the articulated motives but they were surely powerful and instinctive ones.

⁹¹ F. Braudel, The Mediterranean and tbe Mediterranean world in the age of Philip 11 (London, 1972) 224-5.

CHAPTER 13/ CYRENE

JOYCE REYNOLDS AND J. A. LLOYD

I. INTRODUCTION

Modern Cyrenaica, in the Roman period variously named Cyrenae (from its chief city), the Cyrenaea, the parts around Cyrene, Libya around Cyrene, was bequeathed to Rome in default of an heir by its king Ptolemy Physcon in 15 5 B.C., and inherited by her on the death of his son, Ptolemy Apion, in 96 в.с.¹ Rome freed the Greek cities (we are not told whether or not she also gave them immunity from taxation); she probably accepted ownership of the royal property at once (the estates

¹ The literary evidence for the history of Roman Cyrenaica is limited and often terse and obscure. Archaeological discoveries, including coins and inscriptions add important new information, but it is often fragmentary and insecurely dated. A particular problem arises from the many inscriptions which were dated by reference to an eponymous priest of Apollo, for whose year of office we have no other evidence, or to an era which is not specified. The present writers have conjectured that after 96 b.c. the cities used an era dating from the Roman declaration of their liberty. It would be understandable if they started another era in 75/4 or in 67 (the latter has recently been proposed, although not quite proved, for Berenice). It is certain that Cyrene, and almost certain that Teuchira, took Actium as a new starting-point and likely (as is assumed here) that the other cities did the same. Even at Cyrene, moreover, many inscriptions of the Principate are dated in a year which is patently not Actian and is sometimes explicitly stated to be the regnal year of a named emperor. Unfortunately the texts often fail to specify the emperor whose regnal year they were using, thus making the precise chronology and sequence of events obscure to us. In general, see J. Reynolds, in Gadullah 1968 (e 780л) and on Berenicean practice. Bows к у 1987 (e 776). The main items of ancient evidence are the following:

Inscriptions CIG in 5129-362; CIL 1116-11; SEG ix; items s.v. Cyrenaica in SEG xiii, xviii, xx, xxvi-xxvii and AE 1946,1950,1961-2,1967-69/70,1973,1974,1976-8,1980-3,1983,1987,1989; Smith and Porcher 1864 (e 804л) App. IV; D.M. Robinson, AJA 17 (1915) 137-200; de Visscher 1940 (в 293); G. Oliverio, QAL 4 (1961) 3-54; G. Pugliese-Caratelli, D. Morelli, ASAA 39-40 (1961-2) 217-375; G. Giambuzzi, QAL 6 (1972) 43-104; Lŭderitz 1983 (в 250); and corpora published in the excavation reports on Apollonia (e 785), Berenice (e 793), Cyrene (e 775, e 780, e 79), e 798, e 805, e 807, e 809) and Ptolemais (e 789, e 799).

Coins Robinson 1927(8 347л); Chapman 1968 (в 3 i6a); Buttrey 1983 (в 315),id. 1987 (в 315л); some recent coin discoveries are published in the excavation reports listed above. Literary sources are very scattered; all important ones are collected in the footnotes to Thrige 1940 (е 807л) and some are given in the footnotes below.

Current archaeological discoveries are reported mainly in three journals which specialize in Libyan archaeology, Libya Antiqua (= LA, published Tripoli), Libyan Studies (= LS published London), Quaderni di Arcbeotogia delta Libia (=QAL, published Rome); note also Africa Ramarta (= AR, published Sassari). Mises au point may be found from time to time in these journals, and are occasionally published separately, notably Stucchi 1967 (e 805л), Gadullah 1968 (e 780л), Barker, Lloyd and Reynolds 1985 (e 775л), Stucchi 1990 (e 806л).

619

Map 14. Cytene

are first unequivocally attested in her possession in 63 b.c.). The Libyans of the region were perhaps regarded as dependants of the cities.²

By 75/4 B.C. it was clear that this attempt to exercise suzerainty at no cost had failed. To the literary evidence for Cyrenaican instability in the intervening years inscriptions have recently added vivid detail; there were dissensions and tyrannies within the cities, and sometimes, appar­ently, between them, attacks probably from Libyan raiders and certainly from pirates, famines, sieges, lootings.³ In this context a continuous Roman presence may well have seemed preferable to freedom, bringing a hope of peace and revived prosperity to the local population as well as to the Roman negotiators attested at Cyrene and probably present in all the cities. A senatorial decision to send a quaestor to Cyrenaica is reported of 75 or 74; the first indication of serious administrative activity is of 67, when a cluster of inscriptions records action by Cn. Cornelius Lentulus Marcellinus, a legate of Pompey in the Pirate War. Eutropius in fact dated the annexation in 67 rather than 75/4, and it may have been locally regarded as the first effective year of the new dispensation. We know very little, however, of what was involved in that. Although commonly stated, it is not certain that Cyrenaica was governed with Crete at this stage, nor that governors were invariably of quaestorian standing (although that may seem more likely than not). What we know consists of the names of several quaestors who served there, of references to negotiatores and to publicani there, and to the presence of Cyrenaican silphium in the Roman treasury (some of which, however, was deposited before annexation). The publicani were doubtless managing the royal estates and may also have collected tax, but there-is no evidence; the silphium, a plant which produced a gum-resin used as a condiment and for medicinal purposes, may have come as rent in kind from the estates, or as tax in kind, but we do not know that either.⁴

If there had been hopes that annexation would revive prosperity, they were soon disappointed, for within a very few years Cyrenaica felt the impact of the Roman civil wars. Pompey took Cyrenaican corn to feed the troops he mustered against Caesar; after Pharsalus Pompeian refugees collected there - eventually, it is said, 10,000 of them - under Cato, who forced the reluctant to accept them and, indubitably, to provide supplies for them. It is not surprising that depression is written very clearly in the archaeological evidence for the middle of the first century B.C. recently discovered at Sidi Khrebish, Benghazi, a suburb of ancient Berenice.⁵

SEG ix.7; Livy, Per. 70; Tac. Ann. xiv.18; Cic. Leg. Agr. 1.19.51.

Plut. Lue. 2.2-4, [Plut.] Мог. 2 j j E—2 j 7E; Joseph. A J xiv.7.114; SEG xxvi.1817, xxviii.1540.

Sail. H. 11.fr.4); App. BCiv. 1.111, Eutrop. vi.i 1; for inscriptions, JRS 52 (1962) 97-103. Cic. Plane. 26.63; P'iny. HNx1x.15.30.

Caes. BCiv. 3.5; Luc. ix.39f, ²94ĥ Plut. Cat. Min. 56; Strab. xvn.3.20 (836-7C). Lloyd 1977-8) (E 793)- ii. THE COUNTRY

The eastern and western limits of Cyrenaica were indicated by Ptolemy I as the Great Catabathmos, a steep pass near the modern Egyptian town of Solium, and Automalax, a fort on the Syrtican coast probably at modern Bu Sceefa, a little east of the traditional eastern limit of Carthaginian influence at Arae Philaenorum.[756] These were approximately the limits of the Roman province too in 44 B.C.

How far Ptolemaic and Roman suzerainty penetrated the interior is less clear. The forts established by the Tiberian period (see below) in the Syrtican approaches to the Cyrenaican plateau are clues to the location of the frontier zone there. Very recently Libyan archaeologists have found classical material in the desert south of Mechili, including part of a stone set up in a.d. 5 3/4 to mark the boundary of an estate inherited by the Roman people from Ptolemy Apion; if it belongs where it was found it indicates that Ptolemaic as well as Roman control was much deeper than has been supposed.[757]

The sub-Saharan climate and poor soils of the western and eastern Cyrenaican littorals render them for the most part unsuited to settled cultivation. However, certain areas favoured by underground water, and sometimes also by anchorage facilities, were developed in antiquity as road-stations and minor ports. Systematic survey in the Tripolitanian Syrtica suggests that the intensity of early Roman agrarian activity associated with them (no doubt accompanied by pasturage) has been seriously underestimated; the productivity of the Marmaric region, clo§e to the Great Catabathmos, in the late second century a.d. is illustrated by a cadastral papyrus which records a highly organized landscape given over to cereals, vines, figs and olives. Terracing, water collection and storage systems and irrigation contributed to effective husbandry in these marginal areas. A kinder environment in antiquity has not yet been proven.[758]

The chief cultivable area, and so the zone of the classical cities, small towns and villages, lies in the northern part of Jebel el Akhdar or Green Mountain and its coastal plain. The Jebel is a limestone plateau, cuestaform, which stretches c. 250km as the crow flies from Berenice (Benghazi) in the west to Darnis (Derna) in the east and slopes down to the Sahara in the south. Where it juts northwards into the sea (in the direction of the Peloponnese) it has distincdy Mediterranean qualities in its relief, climate, soils and vegetation. The coastal plain is usually narrow and sometimes interrupted where the mountain reaches the sea; except, therefore, where it broadens at its western end behind Teuchira (Tocra) and Berenice, it offers little room for cities and no possibility of a continuous coast road (the narrowness is accentuated by a rise in the sea level since the classical period but not, apparently, on a scale significant enough to change the essential facts).

The mountain rises steeply on the north, by two main escarpments from sea level to an upper plateau which reaches 500m over much of its length and nearly 900m at Sidi Mohamed el-Hamri a little south of Cyrene. The lower plateau, narrow at its eastern end, broadens towards the west where it accommodates the one extensive fertile plain in the country, controlled by the Greek city of Barka and its hellenistic successor Ptolemais-Barka. Outside this plain the landscape is frequently undulating, with soil often collected in comparatively small depressions and fields surrounded by rocky outcrops. Arable land is, however, quite extensive in the area of Cyrene. In general the soils of the north Jebel are deeper, heavier and more water-retentive than those of the coastal plain, although there are some stretches of thinner soil which are only useful for pasturage. They are also better watered. Rainfall, concentrated in the winter months, may be up to 650mm annually on the high ground; whereas in the coastal plain at Benghazi it is 250—300mm, which is close to the minimum for dry farming; the rate is variable, however, even on the mountain and there is everywhere danger of periodic drought. Moreover, much of the rainfall permeates the limestone and runs underground. It gushes out at points along the edge of the escarpments (as, very notably, at Cyrene); but permanent fresh surface water is rare. There was heavy dependence, therefore, on wells and cisterns, both in the cities and in the countryside; and some construction of aqueducts for cities is attested, at least in the Roman period. There are also a number of water-courses (wadis) which are dry for much of the time but fill briefly on occasions of winter flooding. They commonly run east—west in their early stages but later turn north to reach the coastal plain where they have often deposited good soil which attracted settlements. When broad enough, their beds were cultivated in their upper reaches, although it was necessary to build series of retaining walls across them, to limit the removal of soil by flood water. On their south-north sections, however, they have often cut deep ravines into the limestone, thus providing passes through which movement between the coast and the several levels of the Jebel is comparatively easy, despite the precipitous character of the escarpments. On the upper plateau too they might provide conve­nient routes for movement between the settled areas and the interior. Naturally, settlements often occur on their banks and at their exits at the sea end.

The most spectacular of the wadis, the Wadi Kuf, to the south and west of Cyrene, runs for much of its length in a deep gorge which sharply divides the territory of Cyrene in the east from that of Barka or Ptolemais-Barka to the west. Since it was not bridged until the twentieth century it had a marked effect on the settlement patterns and the system of communications. No doubt some ancient tracks crossed it at much the same point as the modern bridge-builders have chosen, but it seems very likely that the main ancient road from Cyrene westwards turned north to follow the east bank of the wadi which it crossed near the sea, where it becomes broad and shallow;⁹ and, since the mountain interrupts the coastal plain to the west very soon thereafter, the road then turned south again to run approximately parallel to the west bank for some distance before resuming a westward direction. An alternative, but probably minor, route by-passed the eastern end of the wadi by running south from Cyrene through what seems to have been Libyan tribal territory, before turning westwards. Both detours, of course, attracted settlements along their lines and in their proximity.

On the gentle southern slopes of the Jebel the soil is decreasingly rich and the rainfall steadily diminishing as the desert is approached. The main value of this steppe area lay in its production of the wild plant silphium and the pasturage of its scrub. Sedentary occupation was hardly possible beyond the 3 2nd parallel except in the occasional isolated oasis. Both steppe and desert were certainly in the domain of Libyan tribes.

Ancient accounts of the country are schematic and principally concerned with the Cyrene area but they show some appreciation of the configuration and its effects. Herodotus identifies three belts of land, which he says were harvested in succession: the coastal plain, a middle region of hills and the highest country behind. Strabo and Pliny describe a zone extending for about 15 Roman miles south of the coast in which trees could be grown, then a band of similar depth, devoted largely to cereal production. Diodorus notes that the land around Cyrene (which falls within the first zone) grew many crops (wheat, olives, vines and wild trees) and possessed'rivers (by which he probably meant the springs which gush out along the edge of the escarpments). Beyond, Pliny describes an area 30 miles deep and 250 miles across, in which the only crop was the silphium plant. Diodorus makes no mention of silphium but his uncultivated and featureless zone, located south of Cyrene, lacking springs and surrounded by desert, is to be equated with the steppe country in which the plant flourished.[759]

iii. the population, its distribution, organization and internal relationships

The ancient sources encourage belief that the Cyrenaicans were all Greeks or Greco-Romans; but the indigenous Libyan populadon was large and a significant element in regional history. Equally they tend to suggest that all Libyans were nomadic shepherds, little touched by civilization and usually at odds with the Greeks; but the realides were certainly much more complex.

Greeks, mostly Dorians, had come to Cyrenaica in a series of groups beginning in the seventh century B.C. Settling within the cultivable zone they had established, by the hellenistic period, four cities and an unknown number of villages. The hellenistic kings, who ruled Cyrenaica either as a dependency of or an appendix to Egypt, introduced additional settlers; certainly a number of hellenized Jews and perhaps also others of Macedonian, Thracian or Anatolian origin, to judge from the names associable with these peoples that appear in the later inscriptions. The evidence for the hellenistic settlers is clear in the cities, much weaker in the country; that some of them did settle in the country is certain, but it is rash to attempt an estimate of their numbers. There may have been yet more immigrants in the first century B.C., if it is right to deduce from an inscription at Ptolemais that in 67/6 B.C. Pompey authorized settlement of former pirates there. Moreover, there were certainly Italian negotiatores at Cyrene by 67 B.C. and some indication that some men and women, predominantly South Italian in origin, and/or their slave and freedmen employees, may have been established in Cyrenaica more or less permanently.[760]

The indigenous Libyans, depicted by Herodotus as tribally organized, lived both in the cultivable zone and in the steppe to the south, no doubt moving between the two as the need for pasture and tillage required; but the tradition suggests that where geography favoured it some of them developed villages or even agglomerations of dwellings which might resemble towns; and this receives a little support from the discoveries of such Libyan 'townships' recently made in the interior of Tripolitania (no surveys on the same scale have yet been made in the Cyrenaican hinterland).[761] Those living in the relevant areas are said to have helped the first Greek settlers; and although later Cyrenaican history is punc­tuated by Libyan wars, it is probable that peaceful interchange, intermar­riage and cultural influence in both directions were regular. Herodotus already reports it, and the process certainly continued after his time. It was furthered by the trade in silphium, collected by the Libyans but marketed by way of the cities (see below); no doubt also by trade in animals and animal products, of which a trace may be visible in a large enclosure outside the south sector of the hellenistic/Roman city wall of Cyrene, now attractively interpreted as a caravanserai for herdsmen who brought animals from the steppe to the city market.[762] Ptolemy I thought it necessary to rule that the sons of Greek fathers and Libyan mothers were citizens; and it accords with this that already in the hellenistic period portrait sculptures of citizens may show Libyan facial types; no doubt the Libyan names transliterated into Greek which appear in civic inscriptions often indicate men from families of mixed blood (but recent analysis of Greek naming patterns suggests that Libyan names in elite families of Cyrene and Barka may sometimes reflect relations of xettia between these families and Libyan tribal chiefs).[763] Evidence for Libyan cultural influence on Greek cults is particularly clear, but it was certainly much more extensive than that.

The ochloi apparently resident in the cities, and mentioned in an inscription of the first century B.C. from Teuchira, can hardly be other than Libyans. They were, presumably, detribalized and at least partly hellenized, but not absorbed into the citizen body. Similar groups are likely to have existed in all cities and perhaps in the villages too.[764]

At the same time in the first century в.с. many Libyans apparently continued to live very much in their traditional way, even when they had accepted something from the incoming culture. That is doubtless true even within the more highly developed areas of the cultivable zone - traces of them there can be seen, for instance, (from as late as the Roman period) in the upper occupation strata of the cave called the Haua Fteah on the coast near Apollonia;¹⁶ more would certainly be found by systematic survey. Such people were often, no doubt, engaged in agriculture, some as dependent labour on land owned by Greeks, others, more probably, on land communally owned by their own tribes whose main locations were in the steppes but who would bring flocks and herds northward for grazing after the harvest. This system of transhumance was probably practised in ancient Cyrenaica on much the same pattern as was observed in the middle of the twentieth century. For the tribal groups in the steppe there is litde useful evidence. Plutarch shows a tribal chieftain in the area south of Cyrene in c. 87/6 B.C., in touch with aristocrats of the city and clearly able to communicate with them, in fact called in to help in the overthrow of a tyranny. Diodorus Siculus, in a passage which may in part have derived from a knowledgeable hellenis­tic source, wrote of four Libyan tribes in the region of Cyrenaica and of three Libyan life-styles. There were, he said, peaceable farmers and peaceable nomads (presumably transhumants whose seasonal move­ments were on fixed routes), both groups obedient to their chiefs, but a third group consisted of robbers living off the loot of their raids, and sometimes able to coerce the peaceable into joining them. Greeks, it is implied, were aware that many Libyans were acceptable neighbours and that the seriously disturbing element came from further afield.[765] It has become common recently to interpret much of Cyrenaican history as a series of cycles in which Greek expansion of sedentary agriculture threatened Libyan transhumance patterns and led to war, after which an imposed peace opened the way for renewed Greek expansion of sedentary agriculture. Events did sometimes occur in this sequence, but Diodorus' account suggests that it is not the key to all Greek-Libyan clashes. The Libyans were in touch via overland routes with kindred to the south, the east and the west; Libyan raids on Greek lands were certainly sometimes the result of social, political or climatic change outside Cyrenaica.

Some further information can be gleaned from the story of the plant silphium, which grew in a belt of land south of the Greek cities (see above). There is good reason to suppose that it was in normal supply at least as late as 50 B.C., but by the reign of Nero the plant was a rarity; it is generally said to have died out, but is probably a plant found in 1990 still growing in one part of the ancient silphium belt. Strabo explains that barbarian invaders had deliberately destroyed it as an expression of their hostility; his evidence accords with that of Diodorus, for it must mean that tribesmen from a distance were damaging the resources of the peaceful Libyan pastoralists who harvested it. Pliny, on the other hand, blames Roman publicani with a contract for the pascua (presumably the grazing tax collected for use of agerpublicuspopuli Romanics pasture land) who had, he said, found it profitable to encourage grazing on a scale that prevented the plant's survival. We cannot at present make a satisfactory assessment of the two explanations; but they are not wholly incompat­ible. Serious damage to the plants may well have occurred in the Marmaric War of the reign of Augustus (see below); after that there would be little or no silphium to harvest for a time and the only profit to be made would come from the grazing tax. For present purposes however, what matters is Strabo's belief in an interruption of the activities of one group of Libyan pastoralists in normally peaceable relations with Greeks by another from further afield.¹⁸

In the early second century b.c. there had been four Greek cities, Cyrene, Ptolemais (originally the port of Barka, but becoming its centre of government in the third century B.C.), Teuchira (called Arsinoe in the hellenistic period, but reverting to its original, Libyan, name under the Romans) and Berenice (the name given to the new harbour site to which the citizens of Euhesperides had moved in c. 246 B.C.). Between the early second century and 67 b.c. a fifth, Apollonia, was created through promotion of Cyrene's main port; and since hellenistic royal creations were normally given dynastic names it is possible that this was due to Roman intervention. Whatever its date the creation must have been disadvantageous to Cyrene, although perhaps less so than it might seem; it is clear in fact that a good deal of land near Apollonia had already been taken from Cyrenaeans into the possession of the king; and after 75/4 it seems likely that harbour dues there would all be collected for the benefit of Rome. Apollonia and Cyrene were in dispute in 67, but there is no evidence for tensions between them later. Apollonia soon became so much part of the Cyrenaican scene that the whole region acquired the name of Pentapolis, land of the five cities (first attested in the usage of the Elder Pliny).¹⁹

The cities, especially Cyrene, Barca and Teuchira, were sited with a view to exploitation of particularly extensive fertile areas. There were many other fertile and well-watered areas beyond their immediate environs to tempt exploitation, but not of a size to support a city. The settlers were also interested in coastal sites with a view to harbours, for connexions with Greece, for export and import and for the convenience of coastwise shipping by which movement eastwards and westwards was easier than by overland routes (see above). If good harbours are scarce on this coast, quite modest facilities would meet the needs of much ancient shipping; but in the coastal strip even modest harbourage rarely coincides with a sufficient hinterland to support a city. Both in the interior, therefore, and on the coast, there were far more villages than cities; in consequence, most city territories were unusually large. Some villages became substantial places, as road-stations where tracks crossed, for instance, and/or as collecting places for goods to be transmitted between the interior and the coast; but very few ever achieved the status of cities, even in late antiquity when this became easier to do.

There is little information about the government, of these Greek communities either before or after they came under Roman rule. A copy of a constitution established for Cyrene in 322/1 B.C. survives, but we do not know how much, if anything of significance, remained of it by 96,

¹⁸ Strab. xvii.5.2o(8₃6-7c); Pliny, HNxix-i 5.5. « SEG xx 709; Pliny, HN v.5.31.

much less 44 в.с. A decree of the first century в.с. at Teuchira shows that the number of voting citizens there at that time was very small; and that is likely to have been the pattern in all the cides. The Libyan residents in them presumably had no civic rights. Of their other inhabitants the group of Roman negotiatores at Cyrene clearly formed a self-governing community separate from, but within the city, and a community of Jews at Berenice was similarly privileged, as no doubt were the Jews in all the cities; at Berenice they called themselves apoliteuma and conducted their own affairs, non-religious as well as religious, through quasi-civic institutions. That gave them an autonomy which might very easily lead to clashes with the civic authorities.²⁰

Within the city territories many of the Greek villages were too distant from their city centres to allow of day-to-day administration from them. Most, therefore, must have had institutions not unlike those of the Jewish politeuma at Berenice for handling their own affairs; the model was presumably that indicated by the one village decree so far found, which shows something very like a civic organization with localized euergetism and local initiatives in the matter of public building and corn supply, probably in the early first century B.C. Strabo called the Cyrenaican villages noXixvta, and about a century later Ptolemy the Geographer listed a number of them under the heading of iroXeis; they must have seemed rather more than ordinary villages to both. The precise character of their relation to the cities cannot be defined. The only real evidence is that in the territory of Cyrene a number of them used the Cyrenaean dating system by Cyrene's eponymous priest of Apollo; and that in two villages Cyrene's priests of Apollo are known to have taken some responsibility for other cults; in both cases these were cults which were attracting foreign visitors, but we cannot tell whether this had anything to do with the matter or not.²¹

In addition to the Greek villages we must envisage also a number of areas within city territories but not part of them. So 'king's land', which became ager publicus populi Komani, is known to have existed within the territories of all (see below) and in principle was surely outside civic authority. It is probable that there were villages on some of these estates; their inhabitants, in some cases perhaps Libyans, in others probably Jews or other hellenistic immigrants, were surely outside the citizen bodies; they were probably provided with the institutions of a politeuma or something similar. Finally there will have been pockets of Libyan tribal land also outside the civic system, none of which can now be precisely pinpointed.

The relationship of the Greek cities one to another is also unclear.

SEC ix i, xxvi 1817, xx 715, xvi 931, xvii 82J.

SEC ix 354; Strab. xvn.3.21 (837c), Ptol. Gtog. iv.4.7; SEG ix 349 for an example.

Cyrene claimed to be the metropolis of all the others; Strabo called them her irepmoAia. Taken with the ancient use of Cyrene's name for the whole region this has led some to think that they were her dependencies; but that is not consistent with what we know of the independent civic life of Berenice and Teuchira in the first century в.с. Under the Principate, perhaps already in the time of Augustus, there seems to have been a koinon or common council of cities, meeting in Cyrene (see below); it may well have existed earlier and might account better for the language used than dependency.²² As for relations with the Libyans, one might conjecture that an eminent Cyrenean, honoured shortly after the death of Ptolemy Apion for services to Cyrene, the other cities and the tribes of the territories, had negotiated between the cities and the tribes relation­ships that had formerly depended on agreements between the kings and the tribes.²³

It is a natural supposition that all the peoples and types of community described above were comprehended within the four categories, which Strabo is said to have distinguished in 'Cyrenaea' - citizens, farmers, perioeci and Jews. There are, however, obscurities in his formula. Citizens should be the Greeks of the villages as well as of the cities; the farmers might well be dependent labourers, presumably Libyan, on Greek-owned land, but could also, perhaps, be the term for immigrants other than Jews who worked the royal land and sedentary Libyans on Libyan tribal land within city territories; perioeci are even more proble­matic - possibly, but far from certainly, Libyan tribesmen in the steppe.²⁴

IV. FROM THE DEATH OF CAESAR TO THE CLOSE OF THE MARMARIC WAR ( C. A.D. 6/7)

In summer of 44 в.с. Cyrenaica was assigned to C. Cassius, as Crete was to M. Brutus. There is no sign that Cassius ever went near this province. After Philippi it naturally became part of Antony's command and was probably used by him, along with Crete, in the first place as a naval base. There is a series of coins, some minted in Roman denominations, and with parallel issues for Crete and for Cyrenaica, which have often been connected with this; but on present evidence few can be dated precisely enough for the connexion to be certain. By the 'Donations of Alexan­dria' Antony cancelled the Roman annexation of Cyrenaica and gave it as a kingdom to a Cleopatra, either Cleopatra herself or Cleopatra Selene; the discovery at Cyrene of a coin of 31 в.с. from an issue which features both Antony and Cleopatra herself has been taken to suggest that the whole issue should be attributed to Cyrene, with the implication that the new queen was Cleopatra herself. There is no indication that anything

²² Strab. xvii.3.21 (837c). ²³ SEG xx 729. ²⁴ Strabo л/>. Joseph. xiv.7.114.

was done to reconstitute a royal administration however. Antony garrisoned Cyrenaica with four Roman legions under L. Pinarius Scarpus (coins have survived from several of the issues that he made to pay his men); and the cities must have borne the burden of providing supplies for them. After Actium, Scarpus was quick to change alle­giance, refused Antony a landing and, in due course, handed over Cyrenaica and its garrison to Cornelius Gallus as Octavian's representa­tive; surprisingly he had time to issue coins carrying the name of Octavian before he left. The recovery of Cyrenaica for Rome - mentioned in the Res Gestae - was undoubtedly celebrated in Octavian's triumph. But what he recovered was an area in poor shape; the excavation at Sidi Khrebish shows unchecked decay throughout the third quarter of the century.[766]

Octavian/Augustus introduced a new order, which was recognized in Cyrene by the use of a provincial era starting in 31 в.с. At any rate from 27 в.с. Cyrenaica was administered together with Crete, governed by a proconsul of praetorian status. He and the quaestor appointed with him, normally held office for one year, and divided their time between the two parts of the province. The provincial Fasti are full of gaps and uncertainties, so that it would be rash to generalize from them about the kind of men who served in the province and the kind of careers to which they proceeded, at any rate for the reign of Augustus, and indeed, for most of the first century a.d.

The provincial capital was at Cyrene; but it is likely that the governor also held assizes at Ptolemais where there are, as at Cyrene, a number of official inscriptions in Latin. These official texts include prayer formulae of the type used by the Arval Brethren at Rome on 3 January each year and certainly prove that Latin rituals (concerned with the preservation of the current emperor and his family) were conducted in the agorae of each of these cities. There are also from each a few soldiers' tombstones, some probably of the first century a.d., perhaps for men who served as the governors' guards.[767]

Whatever had been the case earlier, the cities had now lost their freedom and the province was certainly taxed. Collection of portoria on goods passing in and out of provincial harbours and frontier stations would normally be let to publicani. Given the large areas occupied by Libyans as well as the complex character of most city territories, it would be understandable if Augustus thought it best to use publicani also for collection of the land tax rather than to entrust it to the cities; but as publicani must also have managed the agerpublicus it is not always possible to distinguish contractors for tax collection from contractors for the public estates in our sparse evidence. A very recently discovered inscription, probably of the Julio-Claudian period, records a dedication to Ceres Augusta in a major precinct of Demeter and Kore at Cyrene, by a promagister publici Cjrenen(sis), who, on the face of it, was the representative in Cyrene of a company of publicani collecting tax. Given his connexion with Ceres, this was probably the land tax, which may well have been collected in kind, and so mainly in cereals. It is a reasonable conjecture that the contract for collection of all Roman taxes in Cyrenaica was let to one company (hence the publicum Cjrenense in contrast to, for example, the quattuor publica Africae), since the profits from each individual tax were perhaps insufficient to tempt bidders. Possibly the contract for management of the ager publicus was let along with that for taxation.[768]

No imperial estates are at present attested in Cyrenaica in the first or second centuries a.d.; certainly no procurator is attested there before the early third century a.d. and there are no adequate grounds for accepting the view that the procurators of Crete also operated in Cyrenaica. One inscription of uncertain date at Ptolemais shows that there were, at some stage, members of the imperial household there; but at present we have no information at all about their function.[769]

The arrangements of Augustus provided, in the long term, for a reasonably stable and prosperous Cyrenaica; in the short term, new problems arose, recovery was certainly interrupted and the period of the reign cannot be regarded as an unqualified success. That is best illustrated at Sidi Khrebish where the district remained in a dilapidated and deserted state throughout it, although the one small temple there was receiving votives, and a channel aqueduct was constructed across it to carry water to a point beyond it, showing that developments were taking place nearer to the city centre. Of those we have a litde positive evidence in two inscriptions erected by the Jewish community of Berenice; they seem to show an active group, possessing a meeting house that is grandly called an amphitheatre, which one of the members could afford to redecorate at his own expense; nevertheless, and despite the inclusion of a few Roman citizens in the community, its financial competence seems to have been modest overall, since the stelae carrying the inscriptions are small.[770]

There is more evidence from the centres of the other cities, and although comparatively little of it can be firmly dated in the first three- quarters of the reign it seems to justify belief that normality was

653

CAESAR TO THE MARMARIC WAR

returning. At Cyrene that is demonstrated by a stela of c. 16-15 B.C. containing the end of a civic decree which conferred the annual priesthood of the cult of Augustus on Barkaeus son of Theuchrestos (we know that he held it in 17/16) and others relating to the will in which he bequeathed one estate to Apollo and Artemis for the use of their priests and another to Hermes and Herakles for provision of oil in the civic gymnasium. Prized amenities of city life were available to cidzens, then, and at least one rich citizen showed his patriotism in the traditional way by benefactions. Civic administration was proceeding as it should. In addition, imperial cult had been quickly established and integrated into the local system of honours (and, no doubt, liturgies); in fact we know from other inscriptions that at this period the name of the priest of Augustus was being used, along with that of the priest of Apollo, to date civic documents.[771]

Nevertheless the dated inscriptions on public Vorks suggest that an extensive programme of repair and new building was still needed in the last decade of the reign and was, in part at least, undertaken by Roman officials; that should perhaps be related to a series of problems that can be detected earlier.

The first of these problems to appear in the record concerns the Jewish communities. At a comparatively early date in the reign they complained that the cities were preventing the dispatch of the money that they offered annually to Jerusalem and harming them in other ways; Augus­tus responded with a letter to the governor confirming both their right to dispatch the money and their isoteleia, which perhaps meant their immunity from the metic tax paid to the cities by resident aliens. By the time of Agrippa's command in the East (17/16—13 B.C.) this decision was being ignored; in Cyrene at least, and probably in the other cities too, informers were accusing the Jews of failure to pay civic taxes due from them and the civic authorities were therefore preventing dispatch of the sacred money again. After hearing a Jewish embassy Agrippa wrote to the city of Cyrene, with a reference to the other cities also, reaffirming the rulings of Augustus. That the provocative factor was clearly financial suggests that the cities were conscious that their means were limited. Nothing more is heard of the matter. By a.d. 3—4, moreover, a few Jewish names appear in a list of ephebes at Cyrene and among the graffiti on monuments in the gymnasium; while in a.d. 60-1 one of the Cyrenaean magistrates called vo/xo^uAa/cc? had a Jewish name. It would appear that some kind of accommodation had been reached between Greeks and Jews, perhaps in order to secure Jewish financial contribu­tions to civic life, as happened in Asia Minor in the Severan period.[772]

When the cities of Cyrenaica, probably acting jointly, sent ambassa­dors to Augustus in 7-6 B.C., there were quite other problems to put to him, mainly concerned with the administration of justice in criminal cases, but involving also the relations of Greeks with Roman citizens and, to some extent, of Greeks with Greeks. The fact that the embassy was sent is, in itself, evidence for some enterprise in the cities (and perhaps for co-operation among them in a koinori). The facts that necessitated it provide unusually sharp insights into the continuing defects of Roman provincial government, as well as into the specific difficulties of this province.[773]

The first striking point is that Roman citizens resident in Cyrenaica (most of whom were probably immigrants, judging by their nomencla­ture) had been successfully ganging up against Greeks, to procure sentences, including death sentences, on innocent men. They were aided by an obviously unsatisfactory system of jury-courts in which prosecu­tion, witnesses and jurors might all be drawn from a very small group of resident Romans. The second edict may add a further insight if, behind its obscurely allusive formulae, we may see a plot by the three Roman citizens it names to involve Greeks in charges of disloyalty to Augustus.

A second point is the implication in Augustus' provisional proposals for reform of the jury system that Greeks could not always trust other Greeks to give them justice; he thought it wise to offer them the option of all-Roman juries in the courts for which he proposed that there should normally be mixed juries, and, in those for which he proposed all-Greek juries, advised that no juryman should be drawn from the same city as anyone directly involved in the case. It must be admitted that it is not certain that this was based on anything in the recent Cyrenaican record rather than on wider experience of Greek feuding, but it is not unlikely, given Cyrene's earlier reputation for violent staseis.

Thirdly, there are now clear indications of financial weakness in the province. The panel from which the Roman jurymen were drawn consisted of 215 names, all that could be found to meet a minimum property qualification as low as 2,500 denarii; Augustus proposed a minimum property qualification of 7,500 denarii for Greek as well as Roman jurymen and was conscious that there might be difficulty in finding enough men who could meet it. We should not, of course, suppose that there were no rich men in Cyrenaica, but must accept that there was no substantial number of reasonably well-off men even among the resident Roman citizens. A similar implication underlies Augustus' decision that a Cyrenaean Greek who received Roman citizenship must continue to fulfil his local obligations unless specifically given exemption from them at the time of his enfranchisement (and then only in respect of property that he owned at that dme).

It is hard to believe that there can have been any perceived threat of attack from outside at the dme when the embassy went to Augustus; and it is still hard when two years later the province received its copy of the fifth edict (setting out a new procedure for certain types of extortion and addressed to all provinces, not specifically to Cyrenaica), for Cyrene then decided, in an apparently carefree mood, to have all five documents inscribed on a marble stela for erecdon in the agora. It is reasonable then to take 5/4 в.с. as the terminus post quern for the next major problem, the raids, on a scale justifying use of the word war, made by the Libyan tribe of Marmaridae (who were located both in the area between Cyrenaica and Egypt and in the Syrtica, between Cyrenaica and Tripolitania). An inscription at Cyrene celebrated the conclusion of this war for the city in a.d. 3, attributing it to the merits of Pausanias, eponymous priest of Apollo in that year. To what happened in between these two dates there is an almost certain reference, without context, in an extract from Dio, Book lv, apparently of a.d.i. Raids, we are told, had gone unchecked by others and by soldiers coming from Egypt, until a praetorian tribune was brought in, when control was established but only after a long period when no senator was sent to govern the cities. A number of Cyrenaican inscriptions can also be associated with these events, most usefully two decrees from Cyrene. The first of these honours Alexandros son of Aiglanor who himself fought in them, killing many of the enemy and taking prisoners. It may well be that when the raids began there were no Roman troops in the province because the cities were expected to deal with that kind of trouble by local militias — using ephebes and neoi (young men just past their ephebic training years), the practice described in the early first century в.с. in a decree at Berenice. If so the system proved inadequate and troops had to be summoned, but they too were, at first, unsuccessful. The second decree details the activities of Phaos son of Kleandros who undertook a dangerous embassy in winter storms during the war and brought back most timely help; the language would accord with a journey to Rome to persuade Augustus to send Dio's praetorian tribune, presumably accompanied by new military forces. The decrees may, of course, exaggerate the weight of responsibility which fell onto the cities at the onset of trouble; but it is certain that Augustus' arrangements had failed to provide for the defence of Cyrenaica. A small gobbet of literary evidence, from Florus, adds that, at an unstated date, Augustus entrusted a war against the Marmaridae and Garamantes to P. Sulpicius Quirinius, a senator whose career is full of problems. Florus' evidence might refer to a governorship of Crete and Cyrene held by him in c. 15 B.C. soon after his praetorship; but Florus seems to suggest that he was associated with the campaign of Cossus Cornelius Lentulus, proconsul of Africa, against the Gaetuli in the Syrtica in a.d. 5—6. By then Quirinius was a rather senior man to hold such an appointment unless, perhaps, we suppose that he was given responsibility not simply for finishing the desert campaign, but for constructing a defence system for Cyrenaica in the Syrtica. It would appear that there was no such system when the Marmaridae began their raids; but a line of forts had been built by the reign of Tiberius, the earliest evidence being perhaps of a.d. 15 and certainly of a.d. 21. There is, unfortunately, nothing at present which firmly links these forts to Quirinius, but the chronology may be thought to favour it.³³

The Syrtican forts provided a screen behind which the province could develop in security from desert raiders and their establishment marks a new phase in the history of Cyrenaica. The screen consisted of a series of strongpoints intended to protect the western and south-western approaches to Cyrenaica, each placed beside a major watering-point for the effective oversight of the populations using it and providing bases for patrols who moved further afield. The garrisons were drawn from auxiliary units of the Roman army and in some cases have left informa­tive graffiti on fort walls and at local shrines. At Sceleidima and Msus (ancient names unknown) there were mounted as well as infantry soldiers, some of the men spoke Latin and several, to judge from their names, were recruited in Spain or Gaul. At Agedabia (ancient Cornicla- num) a number of men came from Syria, chosen no doubt because of their desert experience. At the same time, and along with the graffiti of men who were certainly regular auxiliary soldiers of the Roman army, there are also graffiti of men whose names are drawn from a recognizably Cyrenaican repertoire, and in their mixture of Greek, Libyan and Latin, recall the ephebic graffiti of Teuchira and Ptolemais. Their interpretation is uncertain. They might indicate one episode of military recruiting in Cyrenaica (such as is attested during the Julio-Claudian period), but since they very rarely include any reference to military status, they may be the work of ephebes or neoi from the cities, doing tours of duty alongside, or in substitution for, Roman soldiers.

v. a. d. 4-70

After the Marmaric War reconstruction in the cities was taken in hand quickly. At Cyrene a series of inscriptions of the last decade of Augustus' reign and the early years of Tiberius' shows Roman officials concerned with repairs to public buildings in the agora and its neighbourhood, in

³³ SEG ix 63; Dio. Lv.ioa; AS A A 39-40 (1960-1) 321, no.8; OG1S 767; Flor. ii.ji; Desanges 1969 (e 778); SEG ix 773-95; J. Reynolds, AR 5 (1988) 167-72.

the sanctuary of Apollo, in the temple of Zeus, and perhaps on the defensive walls of the acropolis (but that may have been earlier). In some cases the credit is attributed to a commander of a cohort, suggesting that it began before normal proconsular government was resumed, although it certainly carried on after that for some years.[774] The involvement of Roman officials in building, for which they presumably made funds available, could perhaps be compared with the help that Rome was beginning to give to provincials suffering from natural disasters; although there is no clear evidence that these repairs were necessitated by direct enemy action (failure to maintain the soft local building stone might be sufficient explanation). The inscriptions on the buildings are more often in Latin than might have been expected, which may reflect the presence of Latin speakers, not only soldiers but also the resident Roman negotiators who have left at least one Julio-Claudian record (not precisely dated), apparently from a building which they themselves erected.[775] But it would be mistaken to ignore the part which the Greek citizens were playing too. Fragments of a series of inscribed civic decrees give glimpses of the city's government in operation.[776] The texts of several of these stress the public spirit of the honorands, making it clear that during the Marmaric War men had given very generously indeed in personal effort as well as in money, and continued to do so. Minor monuments show that there were candidates enough for the expensive priesthood of Apollo, and that the ephebic organization was active. Among dedications, the city's large marble altar for the cult of Gaius and Lucius Caesar in the agora is a notable - and surely costly - demonstration of the point.³⁷ Moreover, by the middle of the first century a.d. the lists of Cyrene's priests of Apollo begin to show men with Roman citizenship (usually with the names Tiberius Claudius, implying enfranchisements under the emperors Claudius and Nero); that should mean that the public services of these men were of some note.³⁸ It is possible also that one man from Cyrene entered the Senate at Rome, Antonius Flamma, the proconsul of Crete and Cyrene who was prosecuted and exiled in a.d. 70 for extortion in Cyrenaica. Several men with the names M. Antonius Flamma appear as priests of Apollo and as sponsors of public works at Cyrene in the middle of the first century a.d. and the grandson of one of them (by his daughter) was certainly a Roman senator in the time of the emperor Trajan. It is tempting, therefore, to identify the earliest of the Antonii Flammae of Cyrene with the proconsul; but since it is inherently unlikely that a Greek from this province would have obtained entry to the Senate at so early a date, his family should probably be seen as one of Italian immigrants which accepted Cyrenaean citizenship and probably married into the local aristocracy. If the proconsul was from Cyrene (whether a true Greek or not) we should know at least one really rich man in the province. His prosecution might be held to show feuding within the elite class at Cyrene which, perhaps, expressed itself in support of different Roman factions in the months of civil war after the fall of Nero.³⁹

At Cyrene then recovery is clear. What little we know of Apollonia at this time suggests a similar series of developments there. For Ptolemais and Teuchira there is a different type of evidence. At present building inscriptions, civic decrees and dedications are rare in these cities, but there are plentiful ephebic graffiti and funerary inscriptions throughout the first century a.d.;⁴⁰ that seems to show that there were quite sizeable citizen populations able to afford ephebic training for their sons and a literate, if often modest, memorial for themselves. For Berenice the evidence is different again. Aside from a few statues of Tiberian date which may have come from the city centre or nearby, it consists in what is shown by the excavation of Sidi Khrebish. At approximately the middle of the first century a.d. the whole desolate area was levelled, new paved streets were laid and new houses were built. These had ground plans and external facades like those of their hellenistic predecessors but more substantial foundations and some more elaborate features such as peristyle courtyards, underground cisterns and a little architectural decoration. At the least, they seem to imply that the population of Berenice was growing again and needed more living-space. A Jewish inscription of Neronian date from the city has been used independently of this evidence to argue for an increase in the size of the politeuma population, since the number of its officials is greater than in the earlier inscriptions; it certainly shows reconstruction of the synagogue funded by its members, through subscriptions that were quite numerous although in no case large.⁴¹

In the villages too there appears to have been an increase in the number of funerary inscriptions erected, most of them quite modest, some very much so, but nevertheless evidence that more of the rural people valued a literate funerary record than before, and perhaps indicating an increased rural population. At any rate a military levy was held in Cyrenaica in the fifties suggesting that there was no perceived manpower shortage at that time.⁴²

Evidence for Roman official activity is now limited. We know that

" Reynolds 1982 (e 802).

SEG ix 361-726 (in need of revision). There are also some unpublished texts.

SEG xvи 823. « Tac. Ann. xiv. 18.1.

once in the reign of Tiberius the routine was broken and the tenure of a governor prolonged for three years - but perhaps for reasons connected with the fall of Sejanus rather than with Cyrenaica. Four times we hear that Roman officials provoked Cyrenaeans to prosecute them at Rome, usually for extortion. Only for two Roman initiatives, both due to Claudius, can anything more be said, one concerned with roads, the other with ager publicus.[777]

It is generally held that there was already a good system of communi­cations in Cyrenaica before the Romans came, Unking villages and cities, interior and coast, quite adequately. Its tracks may often be recognized as shallow cuttings in rock-surfaces, perhaps also showing deep wheel- ruts, and sometimes lined by rock-cut sarcophagi and other tombs; there is no sign that any other method of road construction ever superseded it. Neither construction of such tracks nor their repair (a simple process of cutting away a damaged surface) are datable. So although we might expect the Romans to have paid attention to the system quite early, even to have extended it in connexion with the Marmaric War, there would be no indication of that unless their work included the erection of milestones. On present evidence the earliest milestones in Cyrenaica are those erected in thn name of Claudius, on the Cyrene—Apollonia road, the crucial link between Cyrene and the outside world, and on the Cyrene—Balagrae road which led from the city towards some of her most fertile territory, from there on towards the cities of Ptolemais-Barka, Teuchira and Berenice and beyond them to the Syrtican forts.[778] We cannot be sure how much to put to Claudius' credit and especially whether he was responsible for the very important development which involved rerouting the road from Cyrene to Apollonia on a new line which was less steep and less subject to winter flooding than its predecessor. Nor do we know his reasons for action on the Cyrenaican roads; but he may well have been strongly influenced by his concern for the corn supply of the city of Rome, which should have giver, him an interest in Cyrenaican cereal production and in the movement of the grain from the interior to the coast.

'An interest in cereal production may also have been a factor, along with straightforward fiscal considerations, in his decision to appoint a praetorian senator, L. Acilius Strabo, as his legate with a commission to recover ager publicus in Cyrenaica which had been occupied by squat­ters.⁴⁵ Acilius Strabo appears to have spent a good deal of time dealing with a number of small estates in the cultivable zone and apparently with some land in the silphium belt (see above). The series of stelae that he erected after reclamation of land begins in Claudius' reign, when he was at work in country districts east and south of Cyrene; it carries on after Nero's accession and while some of his Neronian stelae are in country districts, a number stand in close proximity to Cyrene and Apollonia. That apparently brought him up against articulate and powerful men in the city elites so that in a.d. 59 he was prosecuted for misconduct. Nero, who heard the case, acquitted him, but nevertheless allowed the squatters to remain in possession, although the survival of many of Strabo's stelae could mean that some of his reclamations were retained.

In the circumstances it would not be surprising if some Cyrenaicans regarded the fall of Nero with regret. Their attitudes and fortunes during the course of the year of the four emperors are not recorded but it is fair to wonder how enthusiastic they felt about the accession of Vespasian, who had once been a quaestor in the province. If they did have doubts they were, in a sense, justified for one of his early acts was to resume the reclamation of ager publicus.

GREECE (INCLUDING CRETE AND CYPRUS) AND ASIA MINOR FROM43 B.C. TO A.D. 69

В. M. LEVICK

I. GEOGRAPHY AND DEVELOPMENT

The area to be dealt with here was in some senses a unity, in others, less important, diverse and falling into three regions, mainland Greece and the islands, western Asia Minor, and the Anatolian plateau.¹ What unified it was geography — common subjection to Mediterranean geology and climatic conditions and the seasonal aridity that governs Mediterranean agriculture; language — it was all predominantly Greek- speaking; history - the entire area had come under the sway of Alexander the Great and then that of Rome; and devotion to common political ideals, those of the city-state (polis). Within these categories came also the variety. In Asia Minor the thin border of arable soil that fronts the limestone mountains of mainland Greece, the 'bare bones' of Attica, as Plato calls them,² was being enriched and extended by accretions brought down by the rivers; to such an extent that cities such as Priene, built like most Greek cities for communication by sea, had already found themselves stranded inland; even Miletus and Ephesus were to lose their position on the coast in the end. Inland and to the east, as the mountains rise into the Anatolian plateau and then into the Taurus range, with its

¹ The most important literary sources are the Geography of Strabo, Books vin-x (5 52-4890) (Greece) and xii-xiv (490-685C) (Asia Minor), the Acts of the Apostles, Pausanias' Guide to Greece, and Pliny's Natural Hittory, especially Books in—vi; historical material is supplied by Cassius Dio's Roman History Books xlvii-lxiii, Appian's Civil Wars Book v, and Tacitus' Annals. A prime contribution has been made by archaeology (e.g. Forschungen in Ephesos veriiffentlicht vom Osterr. arehaolog. Inst, in Wien 1-9 (Vienna, 1906-81); Altertiimer von Pergamon 1-15.i (Berlin, 1911-86); Corinth: Results of Excavations of the American School of Classical Studies at Athens 1-17 (Cambridge, MA, 1932-85); for recent work see the Archaeological Reports published by the Society for the Promotion of Hellenic Studies and the British School at Athens); and inscriptions, of which the main collections are to be found in CIL hi, IG, IGRR, SEG, and SIG¹, ТАМ and the lnscbriften griech. Stadte aus Kleinasien (Кот mission f. d. arch. Erforschung Kleinasiens bei d. Osterr. Akad. d. Wiss., Inst.f. Altertumsk. d.Univ. Kiln), 1- (Bonn, 1972- ); J. and L. Robert, Bulletin epigraphique in RHG 51-97 (1937-84), is indispensable. Coins are hardly less important, and B.V. Head's Historia Numorum is the most succinct guide to them; the main publications are W.H. Waddington, E. Babelon, andTh. Reinach, Recueilgeneraldes monnaiesgrecques£Asie Mineure (Paris, 1904-1912, vol. 1, edn 2 1925), BMC and SNG (notably SNG von Aulock), and RPC, and Burnett et al. 1992(6 512).