[124] That the regnum Vannianum (Plin. H. N. iv. 12, 81), the Suebian state (Tacitus, Ann. xii. 29; Hist. iii. 5, 21), must be referred, not merely, as might appear from Tacitus, Ann. ii. 63, to the dwellings of the people that went over with Maroboduus and Catualda, but to the whole territory of the Marcomani and Quadi, is shown clearly by the second report, Ann. xii. 29, 30, since here, as opponents of Vannius alongside of his own insurgent subjects, there appear the peoples bordering on Bohemia to the west and north, the Hermunduri and Lugii. As boundary towards the east Pliny l.c. designates the region of Carnuntum (Germanorum ibi confinium) more exactly the river Marus or Duria, which separates the Suebi and the regnum Vannianum from their eastern neighbours, whether we may refer the dirimens eos with Müllenhoff (Sitzungsberichte der Berliner Akademie 1883, p. 871) to the Jazyges, or, as is more natural, to the Bastarnae. In reality both doubtless bordered, the Jazyges on the south, the Bastarnae on the north, with the Quadi of the March valley. Accordingly the Marus is the March, and the demarcation is formed by the small Carpathians that stretch between the March and the Waag. If thus those retainers were settled inter flumen Marum et Cusum, then the Cusus not elsewhere mentioned is, provided the statement is correct, not the Waag, or even, as Müllenhoff supposed, the Eipel falling into the Danube below Gran, but an affluent of the Danube westward of the March, perhaps the Gusen near Linz. The narrative in Tacitus xii. 29, 30, also requires the territory of Vannius to have reached to the west even beyond the March. The subscription to the first book of the Meditations of the emperor Marcus ἐν Κουάδοις πρὸς τῷ Γρανούᾳ, proves doubtless that then the state of the Quadi stretched as far as the river Gran; but this state is not coincident with the regnum Vannianum.
[125] Regibus Bastarnarum et Roxolanorum filios, Dacorum fratrum captos aut hostibus ereptos remisit (Orelli, 750) is miswritten; it must run fratres, or at any rate fratrum filios. In like manner afterwards per quae is to be read for per quem and rege instead of regem.
[126] In Pannonia there were stationed about the year 70 two legions, the 13th Gemina and the 15th Apollinaris, in room of which latter during its participation in the Armenian war for some time the 7th Gemina came in (C. I. L. iii. p. 482). Of the two legions added later, 1st Adiutrix and 2d Adiutrix, the first still at the beginning of the reign of Trajan lay in upper Germany ([p. 159], note 1), and can only have come to Pannonia under Trajan; the second stationed under Vespasian in Britain can only have come to Pannonia under Domitian ([p. 174], note 4). The Moesian army numbered after the union with the Dalmatian under Vespasian probably but four legions, consequently as many as the two armies together previously—the later upper–Moesian, 4th Flavia and 7th Claudia, and the later lower–Moesian, 1st Italica and 5th Macedonica. The positions shifted by the marching to and fro of the year of the four emperors (Marquardt, Staatsverw. ii. 435), which temporarily brought these legions to Moesia, need not deceive us. The subsequent third lower–Moesian legion, the Eleventh, was still under Trajan stationed in upper Germany.
[127] Josephus, Bell. Iud. vii. 4, 3: πλείοσι καὶ μείζοσι φυλακαῖς τὸν τόπον διέλαβεν ὡς εἶναι τοῖς βαρβάροις τὴν διάβασιν τελέως ἀδύνατον. By this seems meant the transference of the two Dalmatian legions to Moesia. Whither they were transferred we do not know. According to the Roman custom elsewhere it is more probable that they were stationed in the environs of the previous headquarters Viminacium than in the remote region of the mouths of the Danube. The camp there probably originated only at the division of the Moesian command and at the erection of the independent province of lower Moesia under Domitian.
[128] The chronology of the Dacian war is involved in much uncertainty. That it had begun already before the war with the Chatti (83), we learn from the Carthaginian inscription (C. I. L. viii. 1082) of a soldier decorated three times by Domitian, in the Dacian, in the German, and again in the Dacian war. Eusebius puts the outbreak of the war, or rather the first great conflict, in the year Abr. 2101 or 2102 = A.D. 85 (more exactly 1 Oct. 84–30 Sept. 85) or 86, the triumph in the year 2106 = 90; these numbers indeed have no claim to complete trustworthiness. With some probability the triumph is placed in the year 89 (Henzen, Acta Arval. p. 116).
[129] The fragment, Dio, lxvii. 7, 1, Dind., stands in the sequence of the Ursinian excerpts before lxvii. 5, 1, 2, 3, and belongs also in the order of events to a time before the negotiation with the Lugii. Comp. Hermes, iii. 115.
[130] Arrian, Tact. 44, mentions among the changes which Hadrian introduced into the cavalry, that he allowed to the several divisions their national battle–cries: Κελτικοὺς μὲν τοῖς Κελτοῖς ἱππεῦσιν, Γετικοὺς δὲ τοῖς Γέταις, Ῥαιτικοὺς δὲ ὅσοι ἐκ Ῥαίτων.
[131] The walls, which, three mètres in height and two mètres in thickness, with broad outer fosse and many remains of forts, stretch in two almost parallel lines, partly—to the length of ninety–four miles—from the left bank of the Pruth by way of Tabak and Tatarbunar to Dniester–Liman, between Akerman and the Black Sea; partly—to the length of sixty–two miles—from Leowa on the Pruth to the Dniester below Bendery (Petermann, Geograph. Mittheilungen, 1857, p. 129), may perhaps be also Roman; but there has not been as yet any exact settlement of this point.
[132] According to von Vincke’s estimate (Monatsberichte über die Verhandlungen der Gesellschaft für Erdkunde in Berlin in the years 1839–40, p. 197 f.; comp. in von Moltke’s Briefe über Zustände in der Turkei, the letter of 2d Nov. 1837), as well as according to the delineations and plans of Dr. C. Schuchhardt communicated to me, three barriers were here constructed. The south–most and probably oldest is a simple earthen wall with (singularly) a fosse in front of it towards the south; whether of Roman origin may be doubtful. The two other lines are an earthen wall, even now at many places as high as three mètres, and a lower wall, once lined with stones, which often run close beside each other and elsewhere again are miles apart. We might hold them as the two lines of defence of a fortified road, though in the eastern half the earthen wall, in the more southern half the stone–wall, is the more northerly, and they cross in the middle. At one spot the earthen wall (here more southerly) forms the rear of a fort constructed behind the stone–wall. The earthen wall is covered on the north side by a deep, on the south side by a shallow, fosse; each fosse is closed off by a bank. A fosse lies also in front of the stone–wall to the north. Behind the earthen wall, and mostly resting on it, are found forts distant from each other seven hundred and fifty mètres; others at irregular distances of the like kind behind the stone–wall. All the lines keep behind the Karasu–lakes as the natural basis of defence; from the point where this ceases, they are carried as far as the sea with slight regard to the character of the ground. The town Tomis lies outside of the wall and to the north of it; but its fortress–walls are put in connection with the barrier–fortification by a special wall.
[133] Vita Hadriani 6: cum rege Roxolanorum qui de imminutis stipendiis querebatur cognito negotio pacem composuit.