Acta Ant. Hung. 44, 2004, 115–122 PÉTER KOVÁCS
THE LATE ROMAN ARMY IN PANNONIA
GREEK TT Summary: In this paper the author gives a short summary of the question of the late Roman army in Pannonia with special regard to the Notitia Dignitatum and the changes in the military organisation during the 4th–5th centuries. The presence of the foederati in this region was examined again on the basis of antique sources. Key words: Roman military history, archaeology of Pannonia.
The question of the late Roman army has been discussed several times but because of the shortage of data most of the problems are hard to solve.1 The most important source, the Notitia Dignitatum (and Ammian’s works) failed to give comprehensive information about the structure of the late Roman military organisation in the Pannonian provinces during the whole late Roman period. The Notitia shows only the situation around 395 (and later to 425–430), when it was compiled and later partially corrected.2 The earlier system cannot be exactly identified on the basis of the NotDig. 1 A shorter version of this paper was published in The Roman army in Pannonia. An archaeological guide of the ripa Pannonica. Ed. ZS. VISY. Budapest 2003, 31–36. MOMMSEN, TH.: Das römische Militärwesen seit Diocletian. Hermes 24 (1889) 195–279 = Gesammelte Schriften VI. Berlin 1910, 206–283; ALFÖLDI, A.: Der Untergang der Römerherrschaft in Pannonien I–II. Berlin–Leipzig 1924– 1926; SZENDY, K. (szerk.): Budapest története I. Budapest 1942, 514–518; VAN BERCHEM, D.: L’armée de Dioclétien et la réforme Constantinienne. Paris 1952, 58, 93–100; VÁRADY L.: Későrómai hadügyek és társadalmi alapjaik. Budapest 1961, 82–83; VÁRADY, L.: New evidences on some problems of the late Roman military organization. ActaArchHung 9 (1961) 333–396, 389–391; NAGY, T.: Die Militärbezirke der Valeria nach der Notitia Dignitatum. ActaArchHung 7 (1959) 183–194; MÓCSY, A.: Pannonia. PWRE IX. Suppl. Stuttgart 1962, 515–776, 629–630; VÁRADY, L.: Das letzte Jahrhundert Pannoniens. Budapest 1969, 376–476; HOFFMANN, D.: Das spätrömische Bewegungsheer und die Notitia Dignitatum. Epigraphische Studien 7. Düsseldorf 1970; VISY, ZS.: Megjegyzések Valeria védelmi rendszereinek kérdéséhez. AntTan 15 (1978) 246–252; SOPRONI, S.: Die spätrömische Limes zwischen Esztergom und Szentendre. Budapest 1978, 156–184; Pannonia régészeti kézikönyve. Ed. MÓCSY A. – FITZ J. Budapest 1990, 84–85; DIETZ, K.-H.: Cohortes, ripae, pedaturae. Zur Entwicklung der Grenzlegionen in der Spätantike. In Klassisches Altertum, Spätantike und frühes Christentum. Adolf Lippold zum 65. Geburtstag gewidmet. Würzburg 1993, 279–329, 295–296, 313–314; FITZ, J.: Die Verwaltung Pannoniens in der Römerzeit I–IV. Budapest 1993–1995, 1189–1193, 1266–1304, 1315–1324, VISY, ZS.: A ripa Pannonica Magyarországon. Budapest 2000, 129–130. 2 POLASCHEK, E.: Notitia Dignitatum. PWRE XVII (1937) 1081–1097.
0044-5975 / 20.00 © 2004 Akadémiai Kiadó, Budapest
116
KOVÁCS, PÉTER
The military inscriptions belonging to the late Roman period are very rare3 with the exception of stamped tiles (bearing stamps of troops, duces, tribuni, magistri (figlinarum), centuriones and OFAR-types).4 During the tetrarchy the two Pannoniae were divided into four: Pannonia I, Savia, Valeria and Pannonia II.5 The Pannonian troops were also divided, the riparienses were under the command of duces,6 the comitatenses under the command of the comites rei militaris or magister militum (per Illyricum), later comites Illyrici (Occ. VII,40–62).7 As suggested by the NotDig, the provinces of Pannonia I–Noricum and Pannonia II–Savia were governed by the same duces. The earliest dux, Aurelius Ianuarius is mentioned in the RIU 699 altar from Almásfüzitő in 303.8 The official name of the Valerian dux appears in the RIU 770 fort building inscription of Solva in 364–367: dux Valeriae limitis. In Pannonia II 2 new legions were set up, the legiones V Iovia and VI Herculia. One of the most important (and very rare) late Roman epigraphic sources is the ILJ = AÉp 1964, 226 building inscription of the legio VI Herculia from 307 found at Ad Militare.9 This inscription shows that the first garrison of the legion was here and ad Militare belonged to Pannonia II. However, in the NotDig Ad Militare belonged to the Valerian ducate (Occ. XXXIII,45. Cf. Rav. Geog. IV,20) and the legion was mentioned as serving in Aureus Mons, Teutoburgium and Castellum Onagrinum (Occ. XXXII,45, 46, 48). Similarly, there is no mention in the NotDig of the milites Histricorum10 or the numerus equi13
Gravestones: Mannersdorf: UBL, H.-J.: Gräberfeld und Siedlung von Mannersdorf. In Actes du IXe Congrès international d’études sur les frontières romaines. Bucureşti–Köln–Wien 1974, 419, T. 66, Poetovio: ILJ 1134, Savaria: RIU 80 (†), 81 (†), 84 (†), Scarbantia: RIU S 38, Mosonszentjános: RIU 237, Levél: RIU S 44, Arrabona: RIU 273 = CSIR Ungarn II. No. 86, Brigetio: RIU 507, 600, 604, 622, 658 (BARKÓCZI, L.: Beiträge zur Steinbearbeitung in Pannonien am Ende des 3. und zu Beginn des 4. Jahrhunderts. FolArch 24 (1973) 67–77), Solva: RIU 789, S 106, Ulcisia Castra: RIU 903b, 906, Aquincum: CIL III 3523, 3529, 10509, CBFIR 391, Csákberény: RIU S 166, Intercisa: RIU 1161, 1165, 1167, 1168, 1191, 1205, 1217, 1239, 1228, 1261 (Intercisa I, 193–196), Alcsutdoboz: RIU 1365, Pusztavám: RIU 1411, Sirmium: CIL III 10232 (†), 10686, ILJ 272, 273, Taurunum: AÉp 1964, 256. Altars: praefecti legionis: RIU 871 (Ulcisia Castra 264–267), CIL III 3524 (Aquincum 267), RIU 385 (Brigetio 269), CIL III 3469 (Aquincum 283–285), CIL III 10406 (Aquincum 290), duces: RIU 699 (Azaum 303), CIL III 4039 (Poetovio, beginning of the 4th c.). Officers: CIL III 3424 (Aquincum 267), AIJ 362 (Poetovio 264–267), CIL III 3228 p. 2328 (Sirmium 260), ILJ 1134 (Poetovio after 262), CIL III 3249 (Sirmium, end of the 3rd c.), RIU 178 (Scarbantia, 1st tetrarchy), RIU S 139 (1st tetrarchy), CIL III 3522 (Aquincum 305–307). Building inscriptions: Poetovio: CIL III 4039: Mithraeum, Forts: 305 A.D.: Batina: ILJ 1072, 375: Carnuntum: 1435811, 364–367: Esztergom: RIU 770, bath of the legionary fort of Aquincum (Thermae Maiores): 268 (CIL III 3525 = 10492 = AÉp (1944) 85 = (1956) 7), schola: beginning of the 4th c.: Aquincum: unpublished, burgi: 371 A.D.: Esztergom: RIU 771, Visegrád: RIU 804, 372 A.D.: Visegrád: RIU S 133. 14 LŐRINCZ, B.: I bolli laterizi militari in Pannonia. Risultati delle ricerche cronologiche. In Pannonia e l’impero romano. Roma 1994, 115–138, 126–131. 15 FITZ: Die Verwaltung Pannoniens 1175–1183. 16 FITZ: Die Verwaltung Pannoniens 1266–1275. 17 FITZ: Die Verwaltung Pannoniens 1189–1193, 1241–1255, 1316–1319, 1390–1402. 18 FITZ: Die Verwaltung Pannoniens 1267–1268, Nr. 904. 19 NAGY, T.: Zu den Militärreform und Verwaltungsreformen Diokletians im pannonischen Raum. In Akte des IV. Internationalen Kongresses für Griechische und Lateinische Epigraphik. Wien 1964, 274–280. Cf. KOVÁCS, P.: Adatok a tetrachia-kori katonai építkezésekhez Pannoniában. AntTan 45 (2001) 141–168. 10 FITZ: Die Verwaltung Pannoniens 1285–1286, Nr. 935, 1303, Nr. 962a, HOFFMANN: op. cit. 74. Acta Ant. Hung. 44, 2004
THE LATE ROMAN ARMY IN PANNONIA
117
tum scutariorum in Intercisa which were mentioned in the RIU 1205 and 1365 gravestones. Meanwhile a few cohortes which were based on the earlier auxiliary troops of the principate are mentioned in the NotDig: cohors I Thracum c. R.–Caput Basentis (Bassiana), III Alpinorum–Siscia, I Aelia sagittaria–Klosterneuburg.11 The origin of the only ala (Sirmiensium) is unknown (Occ. XXXII, 54). As indicated by its honorific title the cohors prima Iovia was established during the tetrarchy (Occ. XXXII, 58).12 The cohortes without ordinal numbers and epithets under the command of tribuni mentioned in the lists belonged most probably also to the earlier, but unidentifiable troops of the principate (NotDig Occ. XXXIII, 59–64, XXXIV, 29–30). The remnants of the earlier structure of the legions can still be observed in the NotDig.13 The 1st fact to be mentioned is that the system of the legions in the NotDig is fragmentary because several parts of the legions did not exist and they have been deleted from the lists when the laterculum was compiled or renewed (and not because of the fragmentary manuscripts). In my opinion the earlier structure was preserved in the Lower Danubian provinces, esp. in Scythia and Moesia II. There were two parts of the legion, a pars (or pedatura) superior and a pars inferior with five cohorts in each under the command of the praefectus ripae.14 The existence of this system also in the other Danubian provinces was pointed out by Dietz on the basis of the data of stamped tiles as well.15 This separation of the legio IIII Flavia from Moesia I is the best example of par(s) sup(erior) and cit(erior) stamps being known16 denoting a unit which was not mentioned even in the NotDig. The system used at the time when the NotDig was compiled, and described in it, came to be abolished later, with only fragments left for observation. It is noteworthy to mention that from the Pannonian lists all (and only) the partes inferiores of the legions disappeared with the exception of the legio II adiutrix (XXXIII, 53). This fact can probably be connected to the participation of the Pannonian troops in the battle of Hadrianapolis.17 In the Notitia only 11
LŐRINCZ, B.: Die römischen Hilfstruppen in Pannonien während der Prinzipatszeit I. Die Inschriften. Wiener Archäologische Studien 3. Wien 2001, 101; Pannonia régészeti kézikönyve 85; ROXAN, M.: Pre-Severan auxilia named in the Notitia Dignitatum. In Aspects of Notita Dignitatum. Papers presented to the conference in Oxford December 13–15, 1974. BAR SS 15 (1976) 59–80, 66, 73, T. I/25, 75, T. II/17, 75, T. III/16. In my opinion there is no evidence that the equites sagittarii in Intercisa (Occ. XXXIII, 38) would have based on the earlier cohors I mill. Hemesenorum sag. eq.: ALFÖLDI: Der Untergang der Römerherrschaft 87–89. 12 SOPRONI: Die spätrömische Limes (n. 137) 136. 13 POLASCHEK: op. cit. 1089–1090, 1095–1096. 14 MOMMSEN: op. cit. 222–225; VÁRADY: Későrómai hadügyek 60–61; VÁRADY: New evidences 368. It is unnecessary to modify Mommsen’s emendation c(o)h(or)t(ium) V to c(o)h(or)t(i)u(m): DIETZ: Cohortes 309–312, because the earlier cohortes (with their ordinal numbers) did not cease to exist in this system as it was proved by DIETZ (Cohortes 282–291), and the expansion c(o)h(or)t(i)u(m) seems to be very unusual, and most probably and logically five cohorts belonged to a pars/pedatura. 15 DIETZ: Cohortes 295–296; DIETZ, K.-H.: Zu den spätrömischen Grenzabschnitten. In Germanen bei derseits des spätantiken Limes. Köln–Brno 1999, 63–68. 16 DIETZ: Cohortes (n. 53–55) 293–294. In Pannonia there is no data on this separation on the basis of the stamps (there is no p(ars) s(uperior) stamp of the legio X gemina, cf. DIETZ: Cohortes (n. 66) 296; LŐRINCZ, B.: Die römischen Ziegestempel des Territoriums von Vindobona. (forthcoming) Typ 186, T. XII/2 (p(ia) f(idelis). 17 Amm. XXXI.7.3. JONES, A. H. M.: The later Roman empire. Oxford 1964, 153–154; MÓCSY: Pannonia 577. For an analogy to the lack of seniores-iuniores pairs, see: HOFFMANN: op. cit. 449–458. Acta Ant. Hung. 44, 2004
118
KOVÁCS, PÉTER
the partes superiores were mentioned. The case of the Valerian list is more difficult because besides the cohorts belonging to the pars inferior there are some belonging to the pars superior (XXXIII, 52, 54 [Alisca, Aquincum]) and there are only praefectus legionis without any data (XXXIII, 55–57 [Contra Tautantum, Cirpi, Lussonium]). Part of line XXXIII, 54 was emended or deleted18 because of the expression praef. leg. II ad. tertiae partis superioris. In my opinion the text can be interpreted without the missing words: the tertia pars (one third) of the pars superior was under the command of this praefectus in Aquincum. According to Várady, the praefectus without any further specifications were leaders of reestablished cohortes, while others tried to explain these scanty entries of titles with the fragmentary character of the text.19 The latter would be logical because they were inserted in the text after the earlier praefectus regardless of the earlier topographical order. To the very end of the list one more (the latest) praefectus was added (pr. legionis Transaquinco [XXXIII, 65]). It appears logical that the legion should be the legio II adiutrix but in the counter forts a mixed detachment of the legions was stationed: e.g., in the castellum Onagrinum (XXXII, 48).20 The building activity of a detachment of the legio I Martiorum under a praepositus is known from Valeria (at the Danube bend) from the burgus building inscriptions from the years 371–372 (RIU 771, 804, S 133: Solva and Visegrád).21 The auxilia were stationed in the forts along the limes, the inner fortifications (Vincentia, Iovia, Quadriburgium)22 and in towns (e.g., Scarbantia: XXXIV, 30; Sirmium: XXXII, 49, 54; Siscia: XXXII, 57; Bassiana: XXXII, 59) as well. The classes of Pannonia II also garrisoned towns (Sirmium, Mursa, Siscia: XXXII, 50, 52, 56) elsewhere along the limes (XXXIII, 58; XXXIV, 26–28). One of the most debated questions of the NotDig is the coexistence of the equites and auxilia-cunei troops.23 It is clear that the equites must have survived Constantine’s reform because there are several forts where only equites were stationed and these forts were used in the postValentinian period as well.24 From among these forts I will mention only two, Matrica and Annamatia where I had the chance to execute excavations recently. Here the
18
Ed. SEECK 194; NAGY: Die Militärbezirke 188; DIETZ: Cohortes 311–312. VÁRADY: Későrómai hadügyek 83; VÁRADY: New evidences 391–392; NAGY: Die Militärbezirke, DIETZ: Cohortes 313–314. 20 SOPRONI, S.: Die letzte Jahrzehnte des pannonischen Limes. [Münchner Beiträge zur Vor- und Frühgeschichte. Band 38.] München 1985, 78; DIETZ: Cohortes (n. 50) 313. 21 The legion was stationed at Kaiser August (GANSCHINIETZ: legio. PWRE XII (1924–1925) 1418–1419; DRACK, W. – FELLMANN, R.: Die Römer in der Schweiz. Stuttgart 1988, 280–281) and it has got nothing to do with the Pannonian army: ALFÖLDI: Der Untergang der Römerherrschaft 81–83; VÁRADY: Későrómai hadügyek 273–274; HOFFMANN: op. cit. 348–349. 22 SOPRONI: Die spätrömische Limes 1328–146, 172–177; SOPRONI: Die letzte Jahrzehnte 100– 102; TÓTH, E.: Az alsóhetényi 4. századi erőd és temető kutatása 1981–1986. Eredmények és vitás kérdések [Vorbericht über die Ausgrabung der Festung und des Gräberfeldes von Alsóhetény 1981–1986 – Ergebnisse und Umstrittene Frage]. ArchÉrt 114 (1987–1988) 22–61. 23 Stratum hypothesis: ALFÖLDI: Der Untergang der Römerherrschaft 88–91; VAN BERCHEM: op. cit. 93–100; NAGY: Die Militärbezirke; MÓCSY: Pannonia 630; coexistence: SOPRONI: Die spätrömische Limes 156–168, VÁRADY: Későrómai hadügyek 82–83; VÁRADY: Das letzte Jahrhundert 389–391; Pannonia régészeti kézikönyve 85; FITZ: Die Verwaltung Pannoniens. 24 MÓCSY: Pannonia 630, SOPRONI: Die spätrömische Limes 156–168, 159–160. 19
Acta Ant. Hung. 44, 2004
THE LATE ROMAN ARMY IN PANNONIA
119
equites Dalmatae and promoti were stationed (Not. Dig. occ. XXXIII, 36–39). On the basis of the excavations and the finds (Valentinian stamped tiles, money circulation) the building activity can be pointed out under Valentinian in both forts and in Matrica a post-Valentinian garbage pit was unearthed in the shrine of the HQ-building.25 The equites could not have belonged to a later period (after the cunei) either26 because the equites Dalmatae are attested as having served in Pilismarót already during the tetrarchy (RIU S 139). This can only mean that the troops mentioned in the NotDig existed at the same time. The co-garrisoning of riparienses does not exclude the simultaneous validity of the NotDig either. Just as the system of legions, the auxiliary troops failed to preserve their intactness after 378. Another fact is that we do not know exactly the strength of the troops (the earlier hypotheses are not convincing) and after 378 the concentration of the remnants of the earlier troops in a castellum can be supposed as well.27 Therefore it is not necessary to suppose Intercisa, for example, to have been fortified with 2500 men (XXXIII, 25–26, 38).28 This is proved by the fact that later troops were posted in a burgus (2nd period counter landing-place) as well (XXXIII,44).29 Another fact must also be considered. From the 360s the auxiliary vici were given up and the rest of the civilian population moved to the forts. This is supported by the observations that the area of the earlier vici was used as part of the Late Roman cemeteries and by civilian (female and puerile) finds recovered inside the forts (e.g., ear-rings, pearls).30 On the grounds of archaeological data further guesses can be made about the lists. There are three Pannonian forts which are not attested in the NotDig but they were surely still used in the 370s. The 1st is the fort at Esztergom-Hideglelőskereszt (to this fort could have belonged to the RIU 770 building inscription31) built between 364 and 367, the 2nd one is the Roman counter fort at Izsa (Celamantia) which, on the basis of coin evidence (Theodosius’s coin) and stamped tiles of the Valentinian period, was given up in the 380s.32 The last one may be the 4th c. inner fortification at Tác. The earlier Roman settlement was rebuilt as an inner fortification in the 4th c. but in contrast with the other forts (Alsóhetény, Ságvár, Környe, Fenékpuszta) only the 1st section was built (with fan- and U-shaped towers). The other forts had 2nd
25 KOVÁCS, P.: Matrica – Excavations in the Roman fort at Százhalombatta (1993–1997). [Studia Classica – Series Historica 3.] Budapest 2000, 54, 60, 88, 90, 108, KOVÁCS, P.: Annamatia (Baracs) – A Roman auxiliary fort in Pannonia. Acta Antiqua Hungarica 41 (2001) 55–80, 74–76, 79–80. 26 MÓCSY: Pannonia 630. 27 JONES: op. cit. 683; VÁRADY: Die spätrömische Limes 358, 361; MOMMSEN: op. cit. 263; SOPRONI: Die spätrömische Limes 166–168; DEMANDT, A.: Die Spätantike. München 1989, (n. 3) 256. 28 MÓCSY, A.: Pannonia-Forschung 1969–1972. ActaArchHung 23 (1971) 57; FITZ: Die Verwaltung Pannoniens (n. 1) 1316; SOPRONI: Die spätrömische Limes 166–168. Line XXXIII, 26 of the list indicates a later date as suggested by the adverb nunc. 29 VISY: A ripa Pannonica 130. 30 KOVÁCS, P.: Castellum és vicus kapcsolata az alsó-pannoniai limes mentén. [Studia Classica – Series Historica 1.] Piliscsaba 1999, 123–131, 166–169. 31 SOPRONI: Die spätrömische Limes 26–29; SOPRONI: Die letzte Jahrzehnte 61, 107–108. 32 KUZMOVÁ, K. – RAJTÁR, J.: Bisherige Erkenntnisse zur Befestigung des Römerkastells in Iža. SlovArch 34 (1986) 204, 220.
Acta Ant. Hung. 44, 2004
120
KOVÁCS, PÉTER
sections with circular towers which date back to around 375.33 It means that Gorsium had been given up as a military fort before that date. In the Notitia Gorsium is not attested. These data clearly show that the Valerian list was compiled after Valentinian’s reign. The next point under discussion is the exact date to be associated with the time adverb nunc in the Pannonian lists.34 The adverb is a special Pannonian term of the NotDig (XXXII, 56; XXXIII, 26, 27, 44, 47; XXXIV, 28). Outside Pannonia it was mentioned only in Raetia (XXXV, 15, 17). In Pannonia it can be found in the case of the Praefectus classis Aegetensium sive secundae Pannonicae, … nunc Sisciae, the Cuneus equitum Constatntianorum, Lusionio, nunc Intercisa, the Cuneus equitum stablesianorum, Ripa Alta, nunc Concadcuha, the Equites sagittarii, Altino, nunc in burgo contra Florentiam and the Auxilia Ursarensia, Pone Navata, nunc ad Statuas and it means that the original list was later supervised. The only question is when it happened: in 380, 409 or between 425 and 430? The 1st and the 3rd possibilities can be ruled out, because the 1st date has been arrived at on the basis of the supposition that the Pannonian lists were compiled under Valentinian (which, as we could see, is not true). The 3rd one (the date of the last registrations in the Notitia) is absolutely unlikely because the province Valeria was evacuated and reestablished as Valeria Media after 425.35 The case to be mentioned last is that of the Pannonian foederati. It is generally supposed that the Goth–Alan–Hun people of Alatheus and Saphrac settled down here in 380 under a foedus.36 The sources that might be referred to in this context are passages by Jordanes and Zosimos (Jordanes Get. 27.40–141, Zosimos IV.34.2–4). The main problem with this hypothesis is that Zosimos continues the story and they are said to have left the Pannonian provinces.37 The archaeological finds from the end of the 4th c. do not even show their presence and the foederati did not serve in the Roman army, only their leaders became duces or comites or magistri militum.38 The only gens mentioned in the NotDig is the gens Marcomannorum in Pannonia I (Occ. XXXIV, 24).39 Naturally, this fact does not mean that the widely known Barbarisation
33
TÓTH: Az alsóhetényi 4. szd. erőd. VÁRADY: Későrómai hadügyek 83; SOPRONI: Die spätrömische Limes 160–164; SOPRONI: Die letzte Jahrzehnte 95–100; FITZ: Die Verwaltung Pannoniens 1322–1324; VISY: A ripa Pannonica 130. 35 TÓTH, E.: Provincia Valeria Media. ActaArchHung 41 (1989) 197–226; BÓNA, I.: Das Hunnenreich. Budapest–Stuttgart 1991, 48–50: Greg. Tur. II,8; Prosper 1288; ChronMin I,471, 658; Philost. XII,14. 36 VÁRADY: Későrómai hadügyek, SOPRONI: Die spätrömische Limes 200, 206; SOPRONI: Die letzte Jahrzehnte 86–93. 37 DiapleÚsantej oân ™pˆ toÚtoij tÕn ”Istron, dianooÚmenoi te di¦ Paion…aj ™pˆ t¾n ”Hpeiron diabÁnai, peraiwqÁnai d™ tÕn Acelùon kaˆ ta‹j `Ellhnika‹j pÒlesin ™piqšsqai, trof¦j por…sasqai ò»qhsan prÒteron, Aqan£ricon pantÕj toà basile…ou tîn Skuqîn ¥rconta gšnouj ™kpodën poi»sasqai d™ prÕj tÕ mhdšna kat¦ nètou tÕn kwlÚsanta t¾n aÙtîn ™pice…rhsin œcein. ™piqšmenoi to‹nun aÙtù sÝn oÙdenˆ pÒnJ tîn tÒpwn ™n oŒj Ãn ¢panšsthsan. Cf. KOVÁCS: Matrica 129–147. 38 VÁGÓ, E. B.–BÓNA, I.: Die Gräberfelder von Intercisa. Der spätrömische Südostfriedhof. Budapest 1976, 196–206; DEMANDT: op. cit. 270; VÁRADY: Későrómai hadügyek 39–48. 39 NAGY, T.: Die gens Marcomannorum in Pannonia Prima. MIA 12–13 (1982–83) 113–121. 34
Acta Ant. Hung. 44, 2004
THE LATE ROMAN ARMY IN PANNONIA
121
of the late Roman army was unknown in the Pannonian provinces. The best evidence of this process is the CIL III 3576 grave inscription of Francus from Aquincum with the following words:40 Francus ego civis Romanus miles in armis Egregia virtute tuli bello mea dextera semper The Pannonian troops of comitatenses were under the command of the comites rei militaris or magistri militum, and later (after the distinction Illyricum was made) they were commanded by the comites Illyrici (Occ. VII, 40–62).41 The NotDig only indicates this situation but gives lists only for after 396.42 Among the comitatenses there were several troops which were established under Honorius as testified by their names: Honoriani victores, Mauri Honoriani seniores, Mattiari Honoriani Gallicani (Occ. VII, 48, 51, 52). In 406–407 new pseudocomitatenses legiones were set up like the Lancearii Lauriacenses and Comagienses (VII, 58–59) and the Secunda Iulia (VII, 60).43 In 409 5 troops (6000 soldiers) from Illyricum under the command of the comes Valens44 were sent to Italy against Alarich (Zos. V, 45,1–2; VI, 7,2). Zosimos mentions him as tîn kat¦ Dalmat…an tagm£twn ¹goÚmenoj which may indicate that at that time the main garrisons of the comitatenses were the Dalmatian cities.45 This is supported by the circumstance that some of the fabricae were transferred to Dalmatia.46 It is noteworthy that no cavalry under the command of the Illyrian comes was listed in the NotDig (Occ. VI).47 From the 2nd half of the 4th c. (and at the beginning of the 5th c.) several Pannonian troops were transferred to areas outside Illyricum.48 The last major effort to stabilize the defence system was taken by dux
40 Budapest története 704, BARKÓCZI, L.: Beiträge zur Steinbearbeitung in Pannonien am Ende des 3. und zu Beginn des 4. Jahrhunderts. FolArch 24 (1973) 78–80, Abb. 8–10. 41 SOPRONI: Die letzte Jahrzehnte 19–20, HOFFMANN: op. cit. II 207–215. 42 FITZ: Die Verwaltung Pannoniens 1318. 43 HOFFMANN: op. cit. (n. 479) 408. 44 FITZ: Die Verwaltung Pannoniens 1318, 1392–1393, Nr. 1041. 45 For that they were not Dalmatian cunei see: VÁRADY, L.: Additional notes on the problem of the Late Roman Dalmatian cunei. ActaAntiqua 11 (1963) 391–406; VÁRADY: Későrómai hadügyek 247– 250. For Zosimos mentioning them see V,45,1–2: tîn ¢pÕ Delmat…aj t£gmata. Similarly HOFFMANN: op. cit. II 213–214. 46 Procurator gynaecii Bassianensis Pannoniae secundae – translati Salonis (NotDig. Occ. XI, 46), procurator gynaecii Iovensis Dalmatiae Aspalato (NotDig. Occ. XI,48). FITZ: Die Verwaltung Pannoniens 1320. 47 PWRE XVI (1936) 1095. 48 Acincenses (pseudocom.) NotDig. Occ. VII, 101 = XLI, 25; Antianenses (pseudocom.) NotDig. Occ. VII, 2 = 162, Cornacenses (pseud. In Gall.) NotDig. Occ. VI, 22 = 272 = VII, 102; Pannoniciani seniores (leg. Pal.) NotDig. Occ. V, 6 = 149 = VII, 7 (intra Italiam), Pannoniciani iuniores (leg. com.) NotDig. Or. VIII, 16 = 48 (Thracia), Lanciarii Sabarienses (leg. Pal.) NotDig. Occ. V, 9 = 152 = VII, 82; Taurunenses (pseud.) NotDig. Occ. V, 111 = 161; Quartodecimani NotDig. Or. VII, 42 = VIII, 39 (leg. com.); Secundani NotDig. Or. IX, 35 (leg. com.). Cf. VÁRADY: Későrómai hadügyek 262–277, HOFFMANN: op. cit. 73, 228–229, 322–323, 340–342, 740; FITZ: Die Verwaltung Pannoniens 1321.
Acta Ant. Hung. 44, 2004
122
KOVÁCS, PÉTER
Generidus in 409 (Zos. V, 46,2).49 The status of the Pannonian troops in the 410s and 420s may have been very similar to the situation of Noricum in the 470s (Eug. s. Sev. 4,20,22).50 Péter Kovács Pázmány Péter Catholic University H-2087 Piliscsaba Egyetem u. 1.
49 ALFÖLDI: Der Untergang der Römerherrschaft 72–83; VÁRADY: Későrómai hadügyek 250– 252; HOFFMANN: op. cit. II 213–214; SOPRONI: Die letzte Jahrzehnte 103–105; FITZ: Die Verwaltung Pannoniens 1318–1319, 1391, Nr. 1040. 50 Eugippius: Szent Severinus élete. A bevezető tanulmányt írta BÓNA ISTVÁN. Budapest 1969, 282–285.
Acta Ant. Hung. 44, 2004