Studies in the History and Language of the Sarmatians
J. Harmatta


6. The Evacuation of Dacia and the Roxolani

We can see, therefore, that a series of problems of the 3rd century A. D., solve themselves when we admit that the Roxolani, being driven by the Goths, settled in Hungary. It remains to be decided when and how that was possible.

It we examine the historical events, we shall find that the first great blows Dacia received, were inflicted upon her under Philippus and Traianus Decius. [195] According to our sources Dacia was ravaged by the Carpi, while the invasion of the Goths was directed rather against Lower Moesia. [196] We shall best understand what a terrible blow this was to Dacia, if we consider that from that time on nearly no Roman coins at all were found in that country. [197] It seems rather obvious that the Roxolani must have moved into Hungary during this unsettled period.

It is not at all difficult to find some traces of this event. Though the invasion of the Carpi was mainly directed against Transylvania, while the Goths broke into Lower Moesia, yet the territory of Oltenia did not remain intact either. Under the emperors Philippus and Traianus Decius the chain of front-line fortifications were lost in the east of the Olt, and it was at this time that the Romans withdrew their occupying forces behind the Olt limes. [198] In view of all this we may assume that this province was also visited by invasions. As it is not very likely that these invasions were in any way connected with either the Carpi or the Goths, it seems much more probable that they hung together with the earliest arrival of the Roxolani on Hungarian territory. That such an invasion was not impossible through the Oltenian-Banatian narrow Roman corridor, is proved by the fact that even in Moesia permanent raids and invasions of the barbarians were the order of the day, so that fortifications had to be built against them far in the interior of the province as is attested by the inscription of Kutlovica dating to 256 A. D. [199] It is not very likely, however, that the entire people of the Roxolani had reached Hungary during these few years. Other tribes such as the Carpi also reached Roman territories only in several waves. [200]

Additional Notes

Addendum to p. 53. On the invasion of the Goths under Traianus Decius cf. B. Gerov: Die gotische Invasion in Mösien und Thrakien unter Decius im Lichte der Hortfunde. Acta Ant. Philippopolitana, Studia Hist. et Phil. Sofia 1963. 127—146.

195. See EPhK 54 (1930), 2.

196. Schmidt, Die Ostgermanen, 207: EPhK 53 (1929) 163. That the Gothic raids were not directed against Dacia, see EPhK 54 (1930) 92.

197. See EPhK 54 (1930) 3. and Magyarok és románok (Hungarians and Roumanians) I. Budapest, 1943. 70.

198. Magvarok és románok (Hungarians and Roumanians), I. 70.

199. See EPhK 54 (1930) 90.

200. See Schmidt, Die Ostgermanen, 221, 224.


It is highly probable, therefore, that the Roxolani reached Hungary not in one body but that they arrived in various groups. Very probably this infiltration and settling down in small numbers, came to an end only after Dacia had been completely given up, and thus there were no more obstacles in the Roxolani's way. We might infer that this movement towards Hungary, beginning under Traianus Decius, was stopped for some time by the consolidation under Gallienus. [201] Though it is true that the bulk of the army, stationed in Transylvania, was withdrawn under Gallienus, [202] yet on the other hand, the country between the Danube and the Timiul was under a stronger military occupation than ever before. [203] The reason of this interesting military reshuffle was performed, according to recent research, in order to leave a route open in case of any threatening invasion and to isolate from Moesia those barbarians that had been settled by Gallienus obviously in the east or north-east of Dacia to guard the frontiers. On the other hand, taking into account that Dacia had not been entirely given up as yet, and that military troops and state administration had been left behind in some places, we might as well suppose that the chief reason for this strong military occupation of the corridor connecting Dacia with the Empire was to guard the contact between the province and the other Roman territory. That such military measures were necessary is shown by the fact that the contact had been for some time in considerable danger. Very likely this danger was due to the Roxolani.

Which route might namely the Roxolani have taken? If we consider all the possible traffic lines in use in those days, then it will be clear to us that the most likely route taken was through the Iron Gate, Mehadia, the Porta Orientalis and through the valley of the Timişul, a route that has been much favoured ever since classical time. [204] It is very interesting to note in this context that the southern part of this road was guarded by two units under Gallienus: the cohors III Dalmatarum between Mehadia and Plugova and a detachment of the legio XIII Gemina at Băile-Herculane. [205] It is hardly probable that these troops were guarding the road between Dierna and Sarmisegethusa at this particular spot since from a strategic point of view it could hardly be imagined to hold up an attack from the north or east by guarding the last stretches of the road. Such a disposition of the troops could habe been effective only when the idea was to secure the road between Dierna and Sarmisegethusa, from an attack that was expected from the south. It is equally unlikely that these troops had been placed here to intercept an attack against Moesia from the east or north-east. It is hardly conceivable either that the enemy could traverse over the Godeanul or the Retezat as both mountains are over 2000 metres high. An attack coming from the Transylvanian Basin was possible only through the Iron Gate pass and the Bistra valley. Had these Roman forces been kept there to defend against such an attack, they ought to have been stationed somewhere about Caransebe. From a strategic point of view it seems more likely that these units were meant to secure the connections between Rome and Dacia against an attack expected from Oltenia. The task of these troops was very likely to guard the road leading from Oltenia through the valley of the Timiul to the Hungarian Plain in order to prevent the Roxolani from breaking through the Dacian corridor and cutting the communications between Rome and her province already partly evacuated.

201. See on this subject Magyarok és románok (Hungarians and Roumanians) 73.

202. EPhK 54 (1930) p. 8, 11 f.

203. See EPhK 54 (1930) 10.

204. See about this road C. Patsch, Der Kampf um den Donauraum unter Domilian und Trajan. SWAW 217 (1937) I. Abh. 108.

205. See EPhK 54 (1930) 12; Schmidt, Die Ostgermanen. 211.


It is very possible that all attempts at an invasion by the Roxolani must have been stopped for a time by the military reforms of Gallienus and the consolidation following it, as well by closing down the military road from Oltenia to the Hungarian Plain by stationing troops there.

An episode from the life of Regalianus, a rival of Gallienus, has hitherto been neglected and not fully understood: this episode will help to prove that a part of the Roxolani had already been settled in Hungary. Regalianus after defeating Ingenuus was proclaimed emperor by his troops in 260 A. D., according to one information in Moesia and according to another at the initiative of the Moesians. [206] Recent investigations resulted in proving that the legions taking part in the revolt of Regalianus were the X and XIV Gemina from Upper Pannonia, the XIII Gemina of Dacia, and the XI Claudia legion from lower Moesia. [207] It looks rather probable that the information of both sources meant not more than that Regalianus was proclaimed emperor on the initiative of the Moesian legion. The proclamation must have taken place somewhere in Pannonia since the defeat of Ingenuus took place in the same province, probably near Mursa or Sirmium. where all the rebellious troops must have been concentrated. [208] The power of Regalianus could hardly have spread as far as Moesia since his coins have not been found outside Pannonia. [209] According to our informations during his short reign Regalianus had to fight against the Sarmatians, but at the instigations of the Roxolani a conspiracy by his own soldiers put an end to his life. [210] According to the usage of the time the name Sarmatians here means Iazyges and therefore the fight against them also connects Regalianus to Pannonia once more. But what does it mean that the Roxolani took part in the plot against Regalianus? This information makes no sense, if we assumed that the Roxolani were settled on the Roumanian Plain near Lower Moesia, since Regalianus could not have visited this province during his short reign. Everything will be clear at once, if we assume that some of the Roxolani had already been settled in Hungary. It might have been they who were responsible for the Sarmatian disturbance which Regalianus had to quell, and this again was an important factor since it ted to his final destruction.

The defeat of Regalianus seems to point to the presence of the Roxolani in Hungary and this needs not be doubted. Another information in the Historia Augusta, cleared up only lately, tells that Regalianus was proclaimed by the Moesians; this information has also been interpreted to prove that it was the Claudia legion of Moesia that took part in the revolt leading to the proclamation of Regalianus. [211] Now if the Historia Augusta gave evidence that was for once exceptionally reliable about the beginning of Regalianus' reign, then we might trust that the story of his fall might equally be true.

Additional Notes

To p. 55. On Regalianus and the Roxolani cf. now J. Fitz: Ingenuus et Régalien. Collection Latomus. Vol. LXXXl. Bruxelles–Berchem 1966. 49 foll.

206. Epitome de Caesaribus 32, 3; Script. Hist. Aug. trig. tyr. 9. On the revolt of Regalianus see Stein, RE II. R. I. Bd. pp. 462 ff.

207. See NK 25 (1926) p. 71 f.

208. See Stein, RE II. R. I. Bd. 461.

209. See Stein, RE II. R. I. Bd. p. 462. The explanations given by B. Saria, Kilo, 30 (1937), pp. 352 ff., do not materially alter this fact.

210. Script. Hist. Aug. Trig. Tyr. 9.

211. See NK 25 (1926) 72.

[Previous] [Next]
[Back to Index]