Rudolf Aßkamp und Tobias Esh (Hrsg.) IMPERIUM – Varus und seine Zeit Beiträge zum internationalen Kolloquium des LWL-Römermuseums am 28. und 29. April 2008 in Münster 2010 VERÖFFENTLICHUNGEN DER ALTERTUMSKOMMIsION FÜR WESTFALEN LANDSCHAFTSVERBAND WESTFALEN-LIPPE BAND XVIII herausgegeben von Torsten Capelle Arhaeological Traces of the Pannonian Revolt 6 –9 AD: Evidence and Conjectures* Ivan Radman- Livaja – Marko Dizdar I. Sed nuntiata Illyrici defectione transiit ad curam novi belli, quod gravissimum omnium externorum bellorum post Punica, per quindecim legiones paremque auxiliorum copiam triennio gessit in magnis omnium rerum diicultatibus summaque frugum inopia1. With these words, Suetonius did not leave us muh information about the Pannonian revolt but one must admit that this dramatic sentence gives to the modern reader a rather clear idea how the Romans themselves perceived that war. One might argue that Suetonius was prone to exaggeration and that it is unlikely that the Pannonian and Dalmatian insurgents ever represented a real threat to the Italian mainland. Nevertheless, considering the manpower and the material resources engaged to quell the revolt, one cannot deny that Augustus took that threat very seriously. The gravity of the situation in 6 AD is ater all conirmed by other sources, most notably by one historian who personally took part in that war, Velleius Paterculus, who also claims that the rebel forces, whih according to his estimates numbered hundreds of thousands of men, planned to launh an ofensive against Italy2. It would thus appear that this belief was widespread among Romans at that time. When the revolt began in 6 AD and when the grim news reahed Rome, the authorities as well as the Italian population probably shared the belief that a large host of Illyrian barbarians was heading towards Rome and that desperate measures were to be taken in order to save Italy. One should not underestimate the strength of rumours, whih could easily and rapidly spread over a large area, causing panic among civilians and military alike. The troops raised by Pannonian and Dalmatian insurgents could certainly not math in any way the army led by Hannibal but in the eyes of the Italian population it was the irst time ater two centuries that a foreign enemy seemed to be strong enough to defeat Roman legions and marh towards Rome. This is why Suetonius’ comparison is actually quite pertinent and understandable. The gravity of the threat is not the only point of comparison between the Punic wars and the Pannonian revolt. The scarcity of arhaeological traces * The abbreviations of the ancient sources are listed in Der Neue Pauly III (1997) pp. XXXVI –XLIV. 1 Suet. Tib. 16. is common to both wars and as arhaeologists we must sadly admit that the events described in the sources, whih took place in Illyricum between 6 and 9 AD, cannot be substantiated with a high level of certainty by arhaeological inds. Actually, until recent times there were no excavated sites in Pannonia, whih could be atributed beyond doubt to the activities of the Roman military during the Augustan age. Fortunately, the discovery of the large camp at Obrežje on the Slovene and Croatian border hanged that picture and it is to be hoped that future inds will shed a new light on the earliest period of the Roman occupation of Pannonia. When we were asked to give an overview of the available arhaeological evidence of the Pannonian revolt, we had to admit that there are no inds whih would undoubtedly corroborate the sources. No batle nor siege sites were identiied, no marhing camps, no inscriptions, in fact not a single ind or arhaeological site whih would point to any known event of the Pannonian revolt. Indeed, the Obrežje camp still awaits a complete publication and since it can be dated from 14 to 13 BC to the early Tiberian age, i. e. a somewhat larger time span, one must admit that its role from 6 to 9 AD, if it had any, is not easy to determine. Nevertheless, it is likely that the fort was also occupied during the Pannonian revolt. There are also stray inds in Northern Croatia, dating from the Augustan period, whih might be traces of the events that took place in Pannonia from 6 to 8 AD, the year when the revolt in Pannonia itself was quelled. II. In 6 AD, Tiberius’ campaign against the Marcomans was about to start when the news about a revolt in Illyricum reahed him in his headquarters in Carnuntum. It seems that the Daesitiates, a tribe from the northern part of what would become the province of Dalmatia, under their leader Bato, instigated the rebellion. The moment was particularly well hosen since most Roman troops normally garrisoned in Illyricum were part of Tiberius’ army. The Daesitiates were immediately joined by what appears to have 2 Vell. Pat. 2, 110; cf. Crook 1996, 106 f.; Gruen 1996, 176; Seager 2005, 33; Eck 2007, 133. 48 Ivan Radman-Livaja – Marko Dizdar been the largest tribe in southern Pannonia, the Breuci, led by a hief bearing the same name as the leader of the Daesitiates, Bato. Many smaller tribes allied themselves to the rebels and Roman citizens who happened to be in that region were massacred in large numbers by the locals. Velleius explicitly mentions traders and military detahments unfortunate enough to ind themselves isolated in enemy territory. The rebels did not content themselves by exterminating every Roman they could ind but also launhed raids against neighbouring setlements and provinces. Velleius was rather vague when he wrote that the rebels raided Macedonia and planed to atak Italy but Cassius Dio makes it clear that they ataked Salona as well as Sirmium. It would thus seem that they had a muh more ambitious strategic plan than just killing and expelling Romans from their tribal territories. The Roman reaction was swit and quite eicient considering the circumstances. Obviously, not muh could be done for Romans stranded in small Pannonian setlements and it is obviously their grim fate that contributed to the spread of panic even in Italy. However, large setlements could still be saved and, above all, the insurgents had to be contained in the areas they already controlled and not allowed to conduct ofensive operations against other provinces, especially not against Italy. The two key strategic positions for the control of Pannonia were Siscia in the west and Sirmium in the east. Both sides were well aware of that fact and the campaigning in the irst year of the war was marked with a lot of manoeuvring all over the Illyricum. The sources do not tell us if the rebels directly threatened Siscia when the uprising started, but it would seem that Tiberius dispathed one legion – presumably the 20th legion led by Valerius Messala, the governor of Illyricum – to reinforce the defence of that town as soon as he became aware of the gravity of the situation. Tiberius himself and the rest of the army soon followed and a large number of troops secured Siscia well before the end of the year 6 AD. At the very beginning of the uprising insurgent troops simultaneously mounted ofensives against Sirmium and Salona but those operations failed eventually. Sirmium, ataked by the Breuci, was saved by Roman troops led by the governor of Moesia, Caecina Severus. It is not entirely clear from Dio’s account if Caecina Severus broke a siege already underway or simply prevented the Breuci to get close to Sirmium since Dio explicitly states that he defeated them at the Dravus lumen (Drava), whih is not quite close to the city of Sirmium. A likely possibility would be that Caecina Severus effectively broke the siege of Sirmium and pursued the Breuci until the Drava river where they made a stand but sufered a defeat. Salona was put under siege by the Daesitiates but the city was well defended and the main rebel force with their wounded leader Bato soon retreated bak to the hinterland although raiders pillaged coastal areas as far as Apollonia. According to the sources, ater those defeats the insurgents quikly renewed the ofensive. The Daesitiates, abandoning the siege of Salona, tried to intercept the forces led by Valerius Messala, who, although outnumbered, managed to defeat them. It would seem that those were the troops sent by Tiberius to reinforce the defence of Siscia but it is unclear if they were ataked on their way to that city or later on when they presumably started offensive operations from Siscia. That defeat did not discourage the Daesitiates because they moved to the east in order to reinforce the Breuci and renew the assault against Sirmium. They put their camp at the Alma mons, today the mount Fruška Gora and started preparing for the atak. Caecina Severus, arguably a rather competent leader, sent the allied Thracian cavalry led by the king Rhoemetalcus to dislodge them, what the Thracians managed to do. It seems that this was the last major operation of the year 6 AD. The Romans were now in undisputed control of Siscia, garrisoned by Tiberius’ troops and reinforced by many units sent from Italy, and Sirmium, whih served as a base to Caecina Severus’ army, as well as the coastal towns. The rebels controlled the area in-between, but the strategic advantage was deinitely on the Roman side. The Romans must have felt conident that they would crush the rebellion in the next campaign season. Indeed, they mustered a lot of troops and launhed an ofensive in 7 AD. Three Moesian legions led by Caecina Severus, two legions brought from Asia Minor under the command of M. Plautius Silvanus and the Thracian cavalry contingent started their progression into the enemy territory on their way to Siscia, presumably to secure the main road line across the Illyricum, the Sava valley. The Romans must have been overconident and they were ambushed in the marshes called Hiulca palus, also known as the Volcae paludes, probably in what is nowadays south-eastern Slavonia. Despite sufering heavy losses, they managed to break out and reah Siscia. This reverse must have convinced Tiberius that the war would not be won by open ield batles and that a diferent strategy would have to be applied. It seems that he resorted to what would be best termed as a scorhed earth policy, a methodical destruction of crops and setlements that would eventually starve the indigenous population and break their will to ight. Judging unnecessary suh a large concentration of troops in Siscia, he sent the eastern Arhaeological Traces of the Pannonian Revolt 49 Fig. 1: Map of Pannonia. Although there are many uncertainties in the sources, whih anyway do not yield too many details about the course of the war, we can grasp the broad lines. As far as Pannonia (ig. 1) is concerned, the control of the Sava valley and of the two towns connected by this communication line, Siscia and Sirmium, was of crucial importance. One should not forget that Siscia was connected by the Sava river to Nauportus and further bak to Italy. The importance of that communication can thus not be overestimated. Most of the manoeuvring by larger armies must have taken place in that area and it would be interesting to see if the arhaeological traces corroborate the impression given by the sources. Indeed, this seems to be the case. The Obrežje camp is placed close to the Sava river and among other tasks it must have been used to protect and control the supply line to Italy. Obviously, as long as the results of the excavations are not fully published, one will have to wait for more elaborate and far-reahing conclusions, but for the time being and considering the preliminary report, it seems quite likely that this camp was in use during the period whih interests us more particularly, i. e. between 6 and 9 AD, quite probably as a supply and repair base4. Not too far away, in the Sava river near Podsused, a Roman helmet of the Hagenau type (Coolus C) (ig. 2) was found, dated by H. R. Robinson to the late 1st century BC and the early 1st century AD. Although his early dating might be a mater of discussion, it is not at all unconceivable that the helmet could have been used during the Pannonian revolt. The incised inscription on the nek guard (CENTVRIA) TAVRI C. MESTRI CINNA(E) clearly shows that this helmet belonged to a Roman citizen, presumably a legionary5. Less than 20 km away, another Roman helmet (ig. 3) was found in the Sava river, near the village of Rugvica. It is also a Hagenau type and considering its small nek guard, it is quite probably an early production model. It could thus belong to the Augustan period. It is interesting to point out that four soldiers have let their names incised on the helmet: (CENTVRIA) C. TRVTTIDI, GAL(EA) C. OFELLI; (CENTVRIA) NAIVI, Q. TVL(L)I; (CENTVRIA) SEI, LVCRETI; (CENTVRIA) HEDI, L. PAPIRI. Three of them bear the so called duo nomina of the irst manner, i. e. the praenomen and the nomen, a detail that would also point to an earlier dating6. 3 Cass. Dio 55, 28– 34; Vell. Pat. 2, 110 –116; Suet. Aug. 16. 25; Suet. Tib. 16. 20. For a more detailed overview of the Pannonian and Dalmatian revolt cf. Hirschfeld 1890, 351 –362; Rau 1925, 313– 346; Köstermann 1953, 345 –378; Pavan 1955, 380; Pašalić 1956, 245 –300; Mócsy 1962, 544 –548; Wilkes 1969, 69–77; Mócsy 1974, 37 –39; Barkóczi 1980, 88 f.; Šašel Kos 1986, 178–191; Gruen 1996, 176 –178; Wilkes 1996, 553; Dizdar – Radman -Livaja 2004, 44 f.; Dzino 2005, 138 –157; Seager 2005, 33 –35. 4 Mason 2006, 67 –71. 5 Hoffiller 1910/1911, 184 ig. 22; Robinson 1975, 28. 31 pl. 44; Radman -Livaja 2001, 48 f. 6 Hoffiller 1910/1911, 184 ig. 23; Robinson 1976, 33 f. pl. 62; Radman -Livaja 2001, 50 f. troops bak to Sirmium and Moesia, where their presence was in any case badly needed because of Dacian and Sarmatian raids. Although Augustus seemed to have been displeased with what he must have considered a passive strategy, Tiberius’ plan proved to be the right one. Next year, on August 3, 8 AD, the starving Breuci under their leader Bato capitulated on the river Bathinus lumen. This capitulation was resented by many Breuci as well as by their principal allies, the Daesitiates. Bato the Breucus was overthrown but this last revolt of the exhausted Breuci was shortlived since Plautius Silvanus crushed it deinitely in the same year. Pannonia was paciied and the last stand of the insurgents would take place in Dalmatia, where they would be defeated next year3. III. 50 Ivan Radman-Livaja – Marko Dizdar Fig. 2: Helmet, Podsused. Fig. 3: Helmet, Rugvica. Fig. 4: Helmet, Bok. Fig. 5: Helmet, Martinska Ves. Fig. 6: Decorated Helmet, Sisak. Fig. 7: Helmet, Sisak. Following the course of the Sava river, we arrive to the main Roman base of that time in western Pannonia, Siscia. A large number of Roman military artefacts have been found there and some of them can be dated with more or less certainty to the Augustan period. First of all, there are several helmets. One of them (ig. 4) closely resembles the helmet from 7 Hoffiller 1937, 29 f.; Radman -Livaja 2001, 46 f.; Radman Livaja 2004, 67 f. Rugvica, the only signiicant diference being the crest knob. It was found in the Sava river, near the village of Bok, next to Sisak. Both the Rugvica and Bok helmet ind their analogies with the Hagenau type helmet from Haltern and it is therefore not unlikely that both were worn by soldiers who fought against the Pannonian rebels from 6 to 9 AD 7. Arhaeological Traces of the Pannonian Revolt 51 Fig. 8: Pila, Sisak. Another Hagenau type helmet (ig. 5) was found during the dredging of the Sava river bed at Martinska Ves close to Sisak. Although this piece has a somewhat more developed nek guard compared to the last two specimens, it is also an earlier type and it could have been in use during the Pannonian revolt. It also has an incised inscription, (CENTVRIA) LVCCI(I), VARRONIS8. Hagenau type helmets are not the only type of Roman helmets found in Sisak. Two Weisenau type helmets discovered in the Kupa river at Sisak could perhaps belong to an early period. Both can broadly be dated to the irst half of the 1st century AD, but it is not easy to date them more precisely within that time period. Nevertheless it is not entirely unlikely that they could have been manufactured at the beginning of the century. One, although quite damaged, still shows traces of a lavish decoration (ig. 6). It was restored several times already in the Roman period since it was successively covered with three decorative metal sheets. The irst was silvered with a punhed vegetal decoration, the second layer was of gilded silver while the last was again a silver sheet with a punhed decoration. In all likelihood this helmet was reconstructed and redecorated on three diferent occasions by its owner or even several owners. This would point to a rather long period of use for this helmet but, as we already said, it is not unlikely that it could have been manufactured as early as the late Augustan period. This helmet has another distinctive atribute. It would seem that it was an oicer’s helmet, i. e. a centurion’s helmet since the helmet was equipped for a lateral wearing of the crest. The transversal crest, crista transversa, was a sign of the centurion rank9. The second helmet is beter preserved but shows no traces of rih decoration (ig. 7). H. R. Robinson dated it to the irst quarter of the 1st century AD, but the developed nek guard of this helmet incites us to be more prudent in dating. However, one can not dismiss the possibility that it is a late Augustan model10. There are also other types of weapons from Sisak whih seem to be Augustan in date. At least four (or perhaps ive) pila (ig. 8) belong to a type used during Augustus reign with a narrow tang pierced with one hole for a wedge11. One gladius (ig. 9) can also be dated to that period. It could be considered as an intermediary type between the typical republican sword, the so -called gladius hispaniensis, and the early Imperial gladius, known as the Mainz type. Although broken in two, this sword found in the Kupa river at Sisak in the 19th century, is quite well preserved12. Its overall length is 64 cm, with a blade length of 53 cm. It is shorter than the average gladius of the republican time, whose overall length was usually over 70 cm13. Nevertheless, its shape, with its slim and waisted blade (maximum 5 cm wide) and an elongated point is Fig. 9: Sword, Sisak. 8 Hoffiller 1937, 30 f.; Radman -Livaja 2004, 68. 9 Veg. mil. 2, 13, 16; cf. Hoffiller 1910/1911, 177 ig. 19; Robinson 1976, 56 pl. 121; Webster 1985, 131; Bishop – Coulston 1993, 93; Feugère 1995, 116 f.; Radman -Livaja 2004, 71– 75; Bishop – Coulston 2006, 103. 10 Hoffiller 1910/1911, 179 ig. 20; Robinson 1975, 52 f. pls. 107 – 110; Radman-Livaja 2001, 52 f.; Radman-Livaja 2004, 75. 11 Hoffiller 1912, 85 f. ig. 30; Radman -Livaja 2004, 25 f. 12 Hoffiller 1912, 104 ig. 36, 2; Radman -Livaja 2004, 33; Miks 2007, 59. 725 cat. A662. 13 Feugère 1993, 97 – 99; Feugère 1994, 15; Connolly 1997, 49– 56. 52 Ivan Radman-Livaja – Marko Dizdar Fig. 11: Knob-shaped hapes, Sisak. Fig. 10: Scabbard itment, Sisak. Fig. 12: Pelta-shaped hapes, Sisak. deinitely reminiscent of the republican type. Both sides of the blade have a groove running in the middle, and the blade has a reinforced tip with a square section. Although similar in shape, it is shorter than the gladius from the Ljubljanica river, dated to the mid -1st century BC, and its dimensions are closer to the Mainz type swords14. It could be roughly dated to the second half of the 1st century BC, but its use in the early 1st century AD is quite likely15. There are also parts and fragments of sword scabbards whih are dated to the Augustan period. The net-like scabbard itment (ig. 10) found in the Kupa river at Sisak16 is somewhat similar to several inds: the scabbard itment of the gladius from the Ljubljanica river, the scabbard itment of the Magdalensberg sword, the one from the Comachio wrek, the scabbard itment from Kalkriese, one fragment also found in Sisak, as well as two inds from Saintes and Pîtres in France17. They are dated from the mid -1st century BC until the Augustan period. The itment from the Kupa river was not riveted to the scabbard but simply atahed by bending the bars around the scabbard. It would seem that it was originally placed near the botom of the scabbard because the lower bar is narrower than the upper one, thus following the tapering of the blade. Therefore this net-like itment was not covering the scabbard in its entirety, but only its lower half. According to J. Istenič and her analysis of suh itments, it would seem that net-like itments, whih covered only the lower part of the scabbard, should be dated to the Augustan period, i. e. the last decades of the 1st century BC and the beginning of the 1st century AD. Thus, the Sisak fragment would be dated to this period. Considering the width of this itment, it is quite plausible that it was placed on a scabbard of a Mainz type gladius, whih might be an argument to date it to the late Augustan period. There are also several fragments of hapes from the Kupa river at Sisak18. The three knob-shaped end- 14 15 16 17 Istenič 2003b, 1 – 5; Miks 2007, 550 cat. A72; 561 f. cat. A115; 660 f. cat. A459; 756 cat. A767; 783 cat. B23, 8; 868 cat. B252, 1; 870 cat. B262, 9. 18 Hoffiller 1912, 111 ig. 42; Radman -Livaja 2004, 39; Miks 2007, 870 f. cat. B262, 13 – 16. Feugère 1993, 140; Istenič 2000a, 171 –179; Istenič 2000b, 1 –4. Feugère 1993, 99; Miks 2007, 59 f. Hoffiller 1912, 120 ig. 51; Radman - Livaja 2004, 38. Feugère 1993, 265 f.; Franzius 1999, 577 f. 594 – 598 igs. 11. 16.; Dolenz 1998, 49 –52; Istenič 2000a, 171 –179, igs. 3 –13; Istenič 2000b, 1–4; Feugère 2002, 12 f.; Istenič 2003a, 271 f.; Arhaeological Traces of the Pannonian Revolt 53 Fig. 13: Broken Dagger, Sisak. Fig. 14: Daggers with scabbards, Sisak. ings (ig. 11) were originally placed as on the better preserved fragment whih still has parts of the gutering atahed to it. Those are typical hapes of Mainz type swords, whose scabbards had guterings binding the edges in order to prevent damage to the scabbard. They can be broadly dated to the irst half of the 1st century AD and it is not excluded that they belong to the beginning of the century19. The two pelta-shaped hape fragments (ig. 12) are seldom encountered. To the best of our knowledge, besides those two pieces from Sisak, suh hapes were found in Magdalensberg in Austria, Basler Münsterhügel in Switzerland and Reka pri Čerknem in Slovenia. The specimen from Magdalensberg, found with the scabbard and the sword, is very similar to one Sisak specimen. It is probable that those hapes are the result of local inluence, although the possibility that the Magdalensberg sword might be an indigenous copy of a Roman gladius cannot be excluded20. They are dated to the second half of the 1st century BC and the early 1st century AD but it is diicult to know how widespread their use among Roman soldiers was. Of the six Roman daggers found in the Kupa river at Sisak, three could be dated to the early 1st century AD. All of them are quite well preserved, even the one broken in two since its handle is still largely preserved (ig. 13). The dimensions, the pronounced mid-rib, the shape of the hilt with a iting that had rivets whih ran through the upper part of the blade, all indicate a fairly early dating. It can be placed in the early Principate, i. e. the last decades of the 1st century BC and the irst decades of the 1st century AD. The two daggers with their scabbards 19 Ulbert 1969, 120; Deimel 1987, 83 f.; Bishop – Coulston 1993, 71; Dolenz u. a. 1995, 57 cat. 70 – 72; Deschler -Erb 1999, 27; Bishop – Coulston 2006, 82; Miks 2007, 224 – 228. 20 Dolenz 1998, 49 – 52; Radman -Livaja 2004, 39; Miks 2007, 660 f. cat. A459; 783 cat. B23, 8; 870 cat. B262, 11. 12. 54 Ivan Radman-Livaja – Marko Dizdar Fig. 15: Armour iting, Sisak. preserved (ig. 14), in the shape of their blade and by the hilt typologically correspond to the previously mentioned specimen. They can certainly be broadly dated to the irst half of the 1st century AD and it is not excluded that they might have been used during the Pannonian war21. One small lorica segmentata iting (ig. 15) from Siscia also deserves to be mentioned in this context since it undoubtedly belonged to the earliest model of that type of armour, the so called Kalkriese type 22. Ater Siscia, inds are muh rarer but following the course of the Sava, we can still encounter some artefacts that might have seen service in the irst decade of the 1st century AD: One of them is a Mainz type sword (ig. 16) found near Stara Gradiška23. Its overall length is 59,5 cm, and the blade, whih is not waisted, is 46,5 cm long. The blade width tapers from 7 cm to approximately 5 cm, with the point being 12 cm long. Being a typical Mainz type, it can broadly be dated to the early Imperial period, i. e. the irst half of the 1st century AD 24. Its use during the Pannonian revolt is not unlikely and we believe that this artefact deserves to be mentioned in this context. It is interesting to point out that a rivet is still in place on top of the tang. The copper alloy rivet is cast in one piece and has an irregular rectangular section. It has two suspension loops on its sides, and one of the loops still holds a small ring from a bronze hain. A small oval copper plate, pierced by a rectangular hole, is placed underneath the rivet. It is an interesting detail, reminding us of the custom to atah the handle of the sword to the wrist, atested among gladiators25. Similar rivets were found in several places, among them Magdalensberg, and seem to have been standard itments of some Mainz type sword hilts26. The last piece of military equipment to be presented here is a Weisenau type helmet (ig. 17) found in the Sava river near the village of Klakar, also called Klakarje in older publications27. H. R. Robinson considered it to be an early type (Imperial-Italic B) and dated it to the second quarter of the 1st century AD. Considering its nek guard, an even earlier dating does not seem unlikely and it is therefore not impossible that it might have been used during the Pannonian war. IV. When looking for arhaeological traces of a war that occurred in the distant past, for obvious reasons, inds of military equipment are usually the best available evidence. However, this does not necessary always need to be the case. There is at least one very interesting ind from eastern Croatia whih can be linked to the events that occurred during the Pannonian revolt. In 1886, a hoard of Roman denarii (ig. 18) was found on an unknown spot somewhere between Osjek and Valpovo. The head of the Arhaeological Department of the National Museum in Zagreb, J. Brunšmid was informed about the discovery but unfortunately only ater the coins had been sold to several private collectors and antiquities dealers. He did his best to acquire at least some of the coins from that hoard: he convinced two collectors to sell their coins to the Museum but he also tried to gather some data about the coins kept in private collections. It is diicult to estimate how many coins were in that hoard, perhaps over hundred and certainly not muh less. According to the data Brunšmid managed to gather about that hoard, most coins were late Republican, dating to the 1st century BC with a few pieces from the 2nd century BC, as well as at least ive coins minted by Augustus from 19 till 2 BC28. Those are obviously the most interesting coins for our subject because they provide us with a terminus ante quem non for the deposition. The hoard was cer- Fig. 16: Sword, Stara Gradiška. 21 Hoffiller 1912, 117 f. igs. 46. 48; Thomas 1971, 48 f.; Scott 1985, 192. 198; Obmann 2000, 8. 26; Radman -Livaja 2004, 50 – 54. 22 Radman -Livaja 2004, 82 f. 23 Hoffiller 1912, 104 ig. 37; Miks 2007, 732 cat. A687. 24 Ulbert 1969, 120; Bishop – Coulston 1993, 69 –71; Feugère 1993, 139– 141; Bishop – Coulston 2006, 78; Miks 2007, 60 f. 25 Bishop – Coulston 1993, 71; Junkelmann 2000, 39; Bishop – Coulston 2006, 78. 26 Deimel 1987, 455 f. pl 110, 18. 19; Božič 1999, 30. We use the opportunity to thank Dragan Božič (Ljubljana) for this information. 27 Hoffiller 1910/1911, 180 ig. 21; Robinson 1976, 66 f. pls. 152 – 154. 28 Brunšmid 1895, 108 – 114. A more detailed paper about this hoard is currently being prepared by our colleague Tomislav Bilić (Zagreb). Arhaeological Traces of the Pannonian Revolt tainly not hidden before 2 BC and considering the dramatic events whih struk that area in 6 AD, we are inclined to believe that it is precisely at that moment that the owner of that money was compelled to hide it, in the hope to retrieve it later, when the situation calms down. This obviously did not happen and we may wonder about the fate of that man. One cannot forget the words of Velleius Paterculus when he describes the massacres of Roman citizens, most notably traders and military personnel who found themselves isolated at the outbreak of the revolt in the areas controlled by the rebels. One of those unfortunate souls might have tried to preserve his savings while trying to escape the wrath of the vengeful Breuci. It is not at all unlikely that this hoard was hidden by a Roman trader or perhaps a Roman soldier. It is ater all not a negligible amount of money. Of course, it might have also been hidden few years later by a local, when Roman troops were retaking the area, probably by the most brutal means imaginable. Be it as it may, the years 6 to 9 AD seem to be the most plausible time frame for the deposition of that hoard. Fig. 17: Helmet, Klakar. Fig. 18a: Coin hoard, Averses, Valpovo-Osjek. Fig. 18b: Coin hoard, Reverses, Valpovo-Osjek. 55 56 Ivan Radman-Livaja – Marko Dizdar V. What can we say ater this short survey and can any conclusion be reahed at all? The answer is not an easy one. The arhaeological inds are actually not that scarce but their interpretation must be very cautious. The inal results of the excavations of the Obrežje camp are not available yet and we will have to wait for their publication before we put forward any hypothesis about the role of that camp and the exact time frame of its existence. Nevertheless, according to the preliminary reports, it is quite likely that the Obrežje camp could have served during the Pannonian revolt as a logistic base on the main communication line between Nauportus and Siscia. It seems likely that another Augustan military camp was discovered in the same area, approximately 10 km further west from the Obrežje camp. Salvage excavations conducted in Čatež exposed the traces of what would seem to be Roman dithes, as well as some metal inds whih could be dated to the Augustan period. Since only a small area was excavated, one can only guess its size but it was quite likely a marhing camp and one can presume that it was occupied during the Pannonian revolt29. According to aerial photographs, several other Roman military camps might be located in that area30. The helmets from Podsused and Rugvica, being isolated inds, might indeed represent traces of river crossing by Roman troops but it is nevertheless hard to prove that hypothesis. Despite being numerous, the inds from Sisak are perhaps the least likely to be connected with any certainty to the Pannonian revolt from 6 to 9 AD. It is not due to hronological issues because we believe that all of them might have been in use during the Pannonian revolt. It should also be pointed out that we did not mention many inds from Siscia whose dating is more uncertain or very broad, like spears, javelins, arrows, lead shots, fragments of hain and scale armour, harness equipment pieces, etc., although they could have been used during the Pannonian war as well31. The problem with inds from Siscia, or we should beter say the diiculty of connecting inds from Siscia to the Pannonian revolt lies in the fact that this place was an important military garrison from 35 BC till at least 43 AD when the 9th legion let for Britain. An artefact dated to the Augustan period found in suh a place can obviously be dated in a somewhat larger 29 Guštin 2002, 70– 74. 30 Grosman 1996, 65 f. time frame and is not necessarily connected to the Pannonian revolt. Isolated inds, found on spots where military activity was never recorded might therefore be a more credible proof of the passage of Roman troops. The sword from Stara Gradiška and the helmet from Klakar could thus be a more plausible trace of the Pannonian war then, for instance, inds from Siscia. In our opinion, the hoard found between Valpovo and Osjek, would be the most credible arhaeological trace of the Pannonian revolt, not only because we have a terminus ante quem non but also because we can link the deposition of that hoard to the information provided by the sources. The circumstances in whih people hide treasures are well known and considering the events that marked the outbreak of the revolt, one can easily imagine why someone would have been forced to hide his goods. However, one must admit that the available arhaeological evidence for the Pannonian revolt is scant at best, especially when we compare it to the data gathered in Germany for the same period. The only way to improve our knowledge is to increase researh in that particular ield. A thorough prospection of given areas might permit us to ind other Roman camps. At the same time, considering that no late Iron Age setlement has ever been extensively excavated in Northern Croatia by modern methods, it becomes obvious why no destruction layers were ever discovered. Puting an accent on excavations of setlements and hill forts would presumably give us the opportunity to ind how those setlements ceased to exist. The Roman punitive expeditions in the years 7 and 8 AD must have let some traces and one can only hope that we will discover them sooner or later. One project of underwater prospection of the rivers in Northern Croatia has started 2007 and, among other things, the spots where Roman military equipment was found in the past will be investigated again more thoroughly. That way, perhaps more inds will see the light of the day in Rugvica, Podsused, Klakar or even some other places. The Roman conquest of southern Pannonia and the early period of Romanization of that area are an exciting ield of researh and one can only hope that we will not have to wait too long before increasing considerably our knowledge about that fascinating period. 31 Radman -Livaja 2004. Arhaeological Traces of the Pannonian Revolt 57 Bibliography Bishop – Coulston 1993: M. C. Bishop – J. C. N. Coulston, Roman Military Equipment from the Punic Wars to the Fall of Rome (London 1993). Bishop – Coulston 2006: M. C. Bishop – J. C. N. Coulston, Roman Military Equipment from the Punic Wars to the Fall of Rome 2(Oxford 2006). Božič 1999: D. Božič, Hat man in der augusteishen Werkstat auf dem Magdalensberg auh Standartenteile produziert?, Instrumentum 10, 1999, 30. 32. Brunšmid 1895: J. Brunšmid, Nekoliko našašća na skupu u Hrvatskoj i Slavonji II. Našašće rimskih obiteljskih denara između Valpova i Osjeka, Vjesnik Hrvatskog Arheološkog Društva 1 (N. S.), 1895, 108–114. Connolly 1997: P. Connolly, Pilum, Gladius and Pugio in the Late Republic, JRomMilSt 8, 1997, 41–57. Crook 1996: J. A. Crook, Political History, 30 B. C. to A. D. 14, in: A. K. Bowman – E. Champlin – A. Lintot (eds.), The Cambridge Ancient History 10. The Augustan Empire, 43 B. C. – A. D. 69 2 (Cambridge 1996) 70–112. Deimel 1987: M. Deimel, Die Bronzekleinfunde vom Magdalensberg, Kärntner Museumsshriften 71 = Arhäologishe Forshungen zu den Grabungen auf dem Magdalensberg 9 (Klagenfurt 1987). Deschler -Erb 1999: E. Deshler-Erb, Ad arma. Römishes Militär des 1. Jahrhunderts n. Chr. in Augusta Raurica, FiA 28 (Augst 1999). Dizdar – Radman -Livaja 2004: M. Dizdar – I. Radman-Livaja, Nalaz naoružanja iz Vrtne ulice u Vinkovcima kao prilog poznavanju rane romanizacje istočne Slavonje, PriloziZagreb 21, 2004, 37–53. Dolenz 1998: H. Dolenz, Eisenfunde aus der Stadt auf dem Magdalensberg, Kärntner Museumsshriten 75 = Arhäologishe Forshungen zu den Grabungen auf dem Magdalensberg 13 (Klagenfurt 1998). Dolenz u. a. 1995: H. Dolenz – C. Flügel – C. Öllerer, Militaria aus einer Fabrica auf dem Magdalensberg (Kärnten), in: W. Czysz – C.-M. Hüssen – H.-P. Kuhnen – C. S. Sommer – G. Weber (eds.), Provinzialrömishe Forshungen. Festshrit Günter Ulbert (Espelkamp 1995) 51–80. Dzino 2005: D. Dzino, Illyrian Policy of Rome in the Late Republic and Early Principate (Diss. Adelaide 2005). Eck 2007: W. Ek, The Age of Augustus 2(Oxford 2007). Feugère 1993: M. Feugère, Les armes des Romains (Paris 1993). Feugère 1994: M. Feugère, L’équipement militaire d’époque républicaine en Gaule, JRomMilSt 5, 1994, 3–23. Feugère 1995: M. Feugère, L’équipement des oiciers dans l’armée romaine, in: Y. Le Bohec (ed.), La hiérarhie (Rangordnung) de l’armée romaine sous le Haut-Empire. Actes du Congrès de Lyon (15–18 septembre 1994) (Paris 1995), 113–126. Feugère 2002: M. Feugère, Militaria de Gaule méridionale 18. Applique de fourreau de glaive augustéen de Saintes (F, Charente-Maritime), Instrumentum 15, 2002, 12–13. Franzius 1995: G. Franzius, Die Römishen Funde aus Kalkriese 1987–95 und ihre Bedeutung für die Interpretation und Datierung militärisher Fundplätze der augusteishen Zeit im nordwesteuropäishen Raum, JRomMilSt 6, 1995, 69–88. Grosman 1996: D. Grosman, Antično Posavje. Uporaba nedestruktivnih arheoloških metod, in: M. Guštin – M. Novakovič – P. Grosman – D. Mušič – B. Lubšina-Tušek (eds.), Rimsko podeželje (Ljubljana 1996) 43–82. Gruen 1996: E. S. Gruen, The Expansion of the Empire under Augustus, in: A. K. Bowman – E. Champlin – A. Lintot (eds.), The Cambridge Ancient History 10. The Augustan Empire, 43 B. C. – A. D. 69 2(Cambridge 1996) 147–197. Guštin 2002: M. Guštin, Il campo militare Romano a Čatež presso Brežice (Slovenia), QuadFriulA 12, 2002, 69–75. Hirschfeld 1890: O. Hirshfeld, Zur Geshihte des germanisch-dalmatishen Krieges, Hermes 25, 1890, 351–362. Hoffiller 1910/1911: V. Hoiller, Oprema rimskoga vojnika u prvo doba carstva, Vjesnik Hrvatskog Arheološkog Društva (N. S.) 11, 1910/1911, 145– 240. Hoffiller 1912: V. Hoiller, Oprema rimskoga vojnika u prvo doba carstva II, Vjesnik Hrvatskog arheološkog društva (N. S.) 12, 1912, 16–123. Hoffiller 1937: V. Hoiller, Nove rimske vojničke kacige, Časopis za zgodovino in narodnopisje, Kovačičev Zbornik 32, 1937, 29–32. Istenič 2000a: J. Istenič, A Roman Late-Republican Gladius from the River Ljubljanica (Slovenia), AVes 51, 2000, 171–182. Istenič 2000b: J. Istenič, A Roman Late-Republican Gladius from the River Ljubljanica (Slovenia), JRomMilSt 11, 2000, 1–9. Istenič 2003a: J. Istenič, Augustan Sword -Scabbards with Net-Like Fitments, AVes 54, 2003, 271–279. Istenič 2003b: J. Istenič, Fragmenat okova korice mača iz doba cara Augusta iz Siska (Fragment of an Augustan sword scabbard itment from Sisak 58 Ivan Radman-Livaja – Marko Dizdar – Siscia), Godišnjak Gradskog muzeja Sisak 3/4, 2003, 1–10. Junkelmann 2000: M. Junkelmann, Familia Gladiatoria. The Heroes of the Amphitheatre, in: E. Köhne – C. Ewigleben – R. Jakson (eds.), The Power of Spectacle in Ancient Rome. Gladiators and Caesars. Exhibition Catalogue London (London 2000) 31– 74. Köstermann 1953: E. Köstermann, Der pannonischdalmatinishe Krieg 6–9 n. Chr, Hermes 81, 1953, 345–378. Mason 2006: P. Mason, The Augustan Fort at Obrežje, Slovenia, in: D. Davison – V. Gafney – E. Marin (eds.), Dalmatia. Researh in the Roman Province 1970–2001. Festshrit John Joseph Willkes, BARIntSer 1576 (Oxford 2006) 67–71. Mócsy 1962: RE S IX (1962) 516–776 s. v. Pannonia (A. Mócsy). Miks 2007: C. Miks, Studien zur römishen Shwertbewafnung in der Kaiserzeit, Kölner Studien zur Arhäologie der römishen Provinzen 8 (Rahden 2007). Obmann 2000: J. Obmann, Studien zu römishen Dolhsheiden des 1. Jahrhunderts n. Chr. Arhäologishe Zeugnisse und bildlihe Überlieferung, Kölner Studien zur Arhäologie der römishen Provinzen 4 (Rahden 2000). Pašalić 1956: E. Pašalić, Quaestiones de Bello Dalmatico Pannonicoque, Godišnjak Istorjskog društva Bosne i Hercegovine 8, 1956, 245–300. Pavan 1955: M. Pavan, La provincia romana della Pannonia Superior, MemLinc (Serie 8) 6, 5, 1955, 373–574. Radman -Livaja 2001: I. Radman-Livaja, Tipovi izloženih rimskih kaciga, in: I. Ružić – Ž. Škoberne (eds.), Kacige u Hrvatskoj (Zagreb 2001), 38–59. Radman -Livaja 2004: I. Radman-Livaja, Militaria Sisciensia. Militaria Sisciensia. Nalazi rimske Ivan Radman-Livaja M. A. Arheološki muzej u Zagreb Trg Nikole Šubića Zrinskog 19 10000 Zagreb Croatia iradman@amz.hr Dr. Marko Dizdar Institut za arheologju Ulica grada Vukovara 68 10000 Zagreb Croatia marko.dizdar@iarh.hr vojne opreme iz Siska u fundusu Arheološkoga muzeja u Zagrebu, Katalozi i monograije Arheološkoga muzeja u Zagrebu 1 (Zagreb 2004). Rau 1925: R. Rau, Zur Geshihte des pannonischdalmatishen Krieges der Jahre 6–9 n. Chr., Klio 19, 1925, 313–346. Robinson 1975: H. R. Robinson, The Armour of Imperial Rome (London 1975). Scott 1985: I. R. Scot, First Century Military Daggers and the Manufacture and Supply of Weapons for the Roman Army, in: M. C. Bishop (ed.), The Production and Distribution of Roman Military Equipment. Proceedings of the Second Roman Military Equipment Researh Seminar, BARIntSer 275 (Oxford 1985) 160–213. Seager 2005: R. Seager, Tiberius 2(Oxford 2005). Šašel - Kos 1986: M. Šašel - Kos, Zgodovinska podoba prostora med Akvilejo, Jadranom in Sirmjem pri Kasju Dionu in Herodjanu. A Historical Outline of the Region between Aquileia, the Adriatic and Sirmium in Cassius Dio and Herodian (Ljubljana 1986). Thomas 1971: E. B. Thomas, Helme, Shilde, Dolhe. Studien über römisch-pannonishe Wafenfunde (Budapest 1971). Ulbert 1969: G. Ulbert, Gladii aus Pompeji. Vorarbeiten zu einem Corpus römisher Gladii, Germania 47, 1969, 97–128. Webster 1985: G. Webster, The Roman Imperial Army of the First and Second Centuries A. D. 3 (London 1985). Wilkes 1969: J. J. Wilkes, Dalmatia (London 1969). Wilkes 1996: J. J. Wilkes, The Danubian and Balkan Provinces, in: A. K. Bowman – E. Champlin – A. Lintot (eds.), The Cambridge Ancient History 10. The Augustan Empire, 43 B. C. – A. D. 69 2(Cambridge 1996) 545–585. Inhaltsverzeihnis Torsten Capelle, Münster Geleitwort ............................................................................................................................................................... 5 Rudolf Aßkamp, Haltern am See – Tobias Esh, Haltern am See Vorwort ................................................................................................................................................................... 7 Teilnehmer des Kolloquiums „IMPERIUM – Varus und seine Zeit“ ............................................................ 9 Varus und Arminius Werner Ek, Köln P. Quinctilius Varus, seine senatorishe Laubahn und sein Handeln in Germanien: Normalität oder aristokratishe Unfähigkeit? .................................................................................................. 13 Dieter Salzmann, Münster Dokumentation der Münzen des P. Quinctilius Varus aus Ahulla und Hadrumetum ............................ 29 Detlef Liebs, Freiburg i. Br. Die Strabarkeit des Arminius nah römishem Reht .................................................................................... 37 Pannonien und die Nordprovinzen Ivan Radman-Livaja, Zagreb – Marko Dizdar, Zagreb Arhaeological Traces of the Pannonian Revolt 6–9 AD. Evidence and Conjectures ................................. 47 Hans Ulrih Nuber, Freiburg i. Br. Militärishe Einsatzstrategien im Alpenfeldzug des Jahres 15 v. Chr. .......................................................... 59 Werner Zanier, Münhen Der römishe Alpenfeldzug unter Tiberius und Drusus im Jahre 15 v. Chr. Übersiht zu den historishen und arhäologishen Quellen ........................................................................ 73 Mihael Gehter, Overath Neue Forshungen zu den augusteisch-tiberishen Militäranlagen am Niederrhein ............................... 97 Reinhard Wolters, Tübingen Die Okkupation Germaniens im Liht der numismatishen Quellen ........................................................... 105 Afrika und der Osten Luisa Musso, Rom Leptis Magna unter Augustus und Tiberius: Romanisierung und Umgestaltung einer punishen Stadt ............................................................................. 115 Konstantinos L. Zahos, Preveza – Evangelos A. Pavlidis, Preveza Die frühen Bauten von Nikopolis. Bemerkungen zu den Bauphasen und - tehniken der Kaiserzeit ................................................................... 135 Mehmet Önal, Şanlıurfa Die Arhive von Zeugma und die Siegelabdrüke der iulisch-claudishen Zeit ........................................ 153 Joseph Patrih, Jerusalem The Praetoria at Caesarea Maritima .................................................................................................................... 175 Josef Wiesehöfer, Kiel Augustus und die Parther .................................................................................................................................... 187 Kunst und Arhitektur des ‚Goldenen Zeitalters‘ Wolf-Dieter Heilmeyer, Berlin Die augusteishe Arhitektur – Klassik und Politik ......................................................................................... 197 Eugenio La Rocca, Rom Der Frieden der Ara Pacis .................................................................................................................................... 211 Eric M. Moormann, Njmegen Die augusteishe Wandmalerei in Rom am Beispiel der Fresken der ‚Villa della Farnesina’ .................... 225 Martin Tombrägel, Leipzig Der fundus Quintiliolus bei Tivoli im Spannungsfeld zwishen historisher und arhäologisher Überlieferung ............................................................................. 237 Abbildungsnahweise .......................................................................................................................................... 245