¹ Still a conceivable ambition at the beginning of Commodus's reign, see Herodian I 5.6 (L.C.L. vol. i p. 26 with note by C. R. Whittaker).

² For the Roman campaign against the Dacians, see Patsch, C., Der Kampf um den Donauraum unter Domitian und Trajan, Beiträge zur Völkerkunde von Südosteuropa V/2 (Vienna 1937)Google Scholar; also Longden, R. P., CAH xi (1938), 223–36.Google Scholar On Dacia before the Roman conquest see H. Daicoviciu, Dacia de la Burebista la cucerirea romanǎ (Cluj 1972).

³ For the impact of the conquest of Dacia on the other Danubian provinces and peoples bordering the empire see Patsch, C., Rev. int. Ét. Balkaniques (1934–1935), 426.Google Scholar

⁴ Monografia geografica a R.P. Române i (Bucureşti 1960), 95–6, 97, 187, 229–30.Google Scholar

⁵ The first systematic treatment of the defences of Roman Dacia is Domaszewski, A. v., Rhein. Mus. xlviii (1893), 240–3Google Scholar cf. Fabricius, E., RE xiii (1926), 640–2Google Scholar (Limes). Also noteworthy are Jung, J., Fasten der Provinz Dacien (Innsbruck 1894), 130 ff.Google Scholar; Christescu, V., Istoria militarǎ a Daciei romane (Bucureşti 1937)Google Scholar; Daicoviciu, C., La Transylvanie dans l'antiquité (Bucureşti 1945), 104 ff.Google Scholar; Szilagyi, J., A Daciai erödrendszer helyörségei és a katonai téglabelyegek (Budapest 1946) (Diss. Pann. ser 2, 21) 4–39Google Scholar; Ferenczi, St., Apulum xi (1973), 191 ff.Google Scholar

⁶ Construction of Roman forts depicted on Trajan's column probably represents those in the immediate vicinity of the Dacian citadels, C. Daicoviciu and Al. Ferenczi, Cetǎţile dacice din munţii Orǎstiei (Bucureşti 1951), 43 ff.

⁷ Macrea, M., Viaja in Dacia romanǎ (Bucureşti 1969), 222–3.Google Scholar

⁸ Macrea, M. in Istoria Romaniei i (1959), 370.Google Scholar

⁹ M. Macrea, op. cit. (note 8), 376.

¹⁰ For the use of the word limes in literary and epigraphic sources see Forni, G. in Dizionario Epigrafico iv (1959), 1076–84.Google Scholar

¹¹ Romanian scholars have in the past used the term limes only for an artificial frontier with rampart and ditch, for example Daicoviciu, C., AISC ii (1933–1935), 255–6Google Scholar and Dacica. Studii şi articole privind istoria veche a pǎmintului românesc (Cluj 1969), 249.Google Scholar Others have applied the term to an advanced line of towers in front of the forts, Torma, C., A limes dacicus felsö része (Budapest 1880).Google Scholar

¹² Tudor, D., Oltenia Romanǎ (4th ed., Bucureşti 1978), 251Google Scholar; Tab. Imp. Rom., sheet L 34, pp. 30, 43, 45, 87.

¹³ Tab. Peut. seg. VIII/I (ed. Miller, K., Stuttgart 1916).Google Scholar

¹⁴ D. Tudor, op. cit. (note 12), 250.

¹⁵ Ferenczi, St., A Koloszvâri, ‘Victor Babeş és Bolyai Farkas’ egyetamek közlemenyei, i, 1–2 (1956), 165–7Google Scholar also Macrea, , Istora Romaniei i, 351 ff.Google Scholar, Viața in dacia romanǎ, 108 ff.; C. Daicoviciu, La Transylvanie…, 107.

¹⁶ M. Macrea, op. cit. (note 7), 34–42 (with bibliography), Syme, R., JRS xlix (1959), 26Google Scholar ff. (= DanubianPapers (Bucureşti 1971), 122–34), Florescu, Gr., Omagiu C. Daicoviciu (Bucureşti 1960), 229Google Scholar, also SCIV i (1950), 169–74Google Scholar; iii (1954), 218–20.

¹⁷ Gerov, B., Klio 37 (1959), 195 ff.CrossRefGoogle Scholar; Daicoviciu, C. and Protase, D., Acta Mus. Napoc. i (1964), 172Google Scholar; Glodariu, I., Acta Mus. Napoc. iii (1966), 434Google Scholar; Protase, D., Acta Mus. Napoc. iv (1967), 47–50Google Scholar; Macrea, M., Acta Mus. Napoc. iii (1966), 134Google Scholar, also Viaţa in Dacia romanǎ, 37–8; Daicoviciu, C., Civilta romana in Romania (Roma 1960), 71.Google Scholar

¹⁸ Russu, I. I., Inscr. Doc. Rom. i (1976), 19Google Scholar (with bibliography).

¹⁹ Concerning the line of the limes south and west of Micia (Veţel) a number of hypotheses exist, all based on intuition rather than evidence: (1) the entire Banat was occupied, with boundaries along the Mureş, and Theiss, and that this belonged either to Dacia (C. Daicoviciu, La Transylvanie…, 96) or to Moesia Superior (Domaszewski, A. v., AEM xii (1894), 140–4)Google Scholar; (2) The Banat was not occupied and the western limit of Roman territory lay along the line Dierna-Tibiscum-Micia (Brandis, C., RE iv (1907), col. 1967–76Google Scholar (Dacia), Alföldi, A., Bericht VI Int. Kong, für Arch. (Berlin 1939), 528–38)Google Scholar; (3) Eastern Banat was included in Dacia; until 118 the boundary followed the line Lederata-Berzobis-Tibiscum (Protase, D., Acta Mus. Napoc. iv (1967), 66–7Google Scholar) or until 159 (A. Radnóti, Limes 3 (Basel), 145); (4) The Banat was conquered by the Romans only as far as the line Lederata-Berzobis-Tibiscum (Szilagyi, J., Közlemények iii (1943), 90Google Scholar).

²⁰ Dǎniiǎ, D. Protase-St., SCIV 19 (1968), 531–40Google Scholar; Horedt, K. (SCIV 25 (1974), 578Google Scholar) upholds the view that the fort at Bretçu and the advanced zone in south-east Transylvania were abandoned under Gallienus.

²¹ D. Tudor, op. cit. (note 12), 38.

²² Tudor, D., Historica i (Craiova 1970), 67–84Google Scholar. SHA Claudius 15,2 says of the future emperor (at that time dux totius Illyrici) ‘habet in potestatem… Dacos exercitus’ and this means that the limes was under attack.

²³ Tab. Peut. seg. VII and VIII.

²⁴ C. Daicoviciu, op. cit. (note 15), 107; M. Macrea., Istoria Romanieii, 350–1, and Viaţa în Daciaromanǎ, 108.

²⁵ Fabricius, E., RE xiii (1926), 641–2Google Scholar (limes); G. Forni, op. cit. (note 10), 1275 does not mention this problem.

²⁶ Research on the two legions is confined to the articles written by Ritterling, RE xii (1924), col. 1572–88, 1710–26.Google Scholar

²⁷ Macrea, M., Acta Mus. Napoc. ii (1965), 141–60Google Scholar; D. Tudor, op. cit. (note 12), 10 ff.; Russu, I. I., Inscr. Dac. Rom. i, 33–60.Google Scholar

⁸⁸ For the history of this see Ferenczi, St., Apulum xi (1973), 191–2.Google Scholar

²⁹ The ‘weaknesses’ of the Dacian limes is used to support notions that Dacia could not not have been fully Romanized, A. Alföldi, Daci e Romani…, 20 ff. and recently Balla, L., Act. Class. Univ. Scient. Debrecensis x–xi (1974–1975), 139.Google Scholar

³⁰ See literature cited in N. Gudea, Aufstieg u. Niedergang der römischen Welt, II Principat, 6, 858–60. The fullest study is Macrea, M., Istoria Romaniei i, 219–35.Google Scholar

³¹ Only for the south west and south east of Dacia do there remain major questions over the line of the frontier. For problems of the former see note 19. For that of the south east see D. Tudor, Oltenia Romanǎ ⁴, (Bucureşti 1978), 319. The line on the west between the Crişul Repede (Bologa) and the Mureş (Veţel) has not yet been identified. Dumitraşcu, S., Acta Mus. Napoc. vi (1969), 483–91Google Scholar, believes that some linear ramparts located and traced by him at the openings of valleys opening west from the Apuseni mountains form part of the Roman frontier. The forts of Sector IX raise problems of dating (Gr. Florescu, Omagiu C. Daicoviciu (1960), 227 attributes them to Moesia Inferior). Other problems include the chronological relationship between Sectors X and XI, in particular for the linear frontier-rampart. This has been dated (a) to Hadrian (Zangemeister, K., Neue Heidelberger Jahrbuch v (1895), 81–2Google Scholar, I. B. Cǎtǎniciu, In memoriam C. Daicoviciu (1974), 56); (b) to Antoninus Pius (Kornemann, E., Klio 7 (1907), 105)Google Scholar; (c) to Septimius Severus (Christescu, V., Istros i (1930), 73–5Google Scholar; Tudor, D., SCIV 6 (1955), 90Google Scholar; Macrea, M., SCIV 8 (1957), 221).Google Scholar

³² D. Tudor, Oltenia Romanǎ ⁴, 319 presents it as a fourth line of defence; see also M. Macrea, Viaţa in Dacia romanǎ, 233.

³³ Viaţa in Dacia romanǎ, 218.

³⁴ Gudea, N., Saalburg Jahrbuch 31 (1974), 41–9Google Scholar with some inversions to the list of forts (pp. 48–9) which do not match the text. See Aufstieg u. Niedergang der römischen Welt II Principat 6, 849–76; also Acta Mus. Porolissensis i (1977), 97–112.Google Scholar

³⁵ Birley, Eric, Research on Hadrian's Wall (Kendal 1961)Google Scholar; Schönberger, H., JRS lix (1969), 144–96Google Scholar.

³⁶ The ancient itineraries contain little on Dacia, Miller, K., Itineraria Romana (Stuttgart 1916)Google Scholar; V. Christescu, op. cit. (note 5), 106–9.

³⁷ M. Macrea, op. cit. (note 27), 153. The archaeological evidence is awaited.

³⁸ C. Torma, op. cit. (note 11), whose work was continued by Buday, A., Dolg. Cluj iii (1912), 107–18Google Scholar; A, Radnóti, Arch. Ért. (1944–5), 137–68; Daicoviciu, C., AISC ii (1933–1935), 254–6Google Scholar; St. Ferenczi, SCIV (1959), 337–50; Acta Mus. Napoc. v (1968), 75–98.Google Scholar The expression Limes Dacicus has been largely eliminated from specialist literature but this does not mean that it is not still employed erroneously by different historians either for a certain sector (the north west) or for the whole limes of Dacia.

³⁹ Lattyak, A., Dolg. Cluj viii (1917), 218–32Google Scholar; Teglás, G., Akademiai Értesitö vi (1885), 413–22Google Scholar; Archeologiai Közlemények 19 (1895), 5–54Google Scholar; Erdely Muzeum 13 (1896), 384–9, 416–27Google Scholar; 17 (1900), 261–9, 313–24.

⁴⁰ Gudea, N., Acta Mus. Napoc. vi (1969), 507–30.Google Scholar

⁴¹ St. Ferenczi, St. Com. Satu Mare (1969), 91–110; File de Istorie ii (1972), 37–46Google Scholar; iii (1974), 181–99; Limes IX Mamaia, 201–5; Sargetia x (1970), 79–104Google Scholar; xi–xii (1974–5), 295–99. I. Mitrofan of the Museum of History of Transylvania at Cluj-Napoca has excavated a tower on the northward-facing limes. The investigations begun by St. Ferenczi on the line of towers in the east remain unpublished.

⁴² As yet only for the advanced line of the north-west section, Ferenczi, St., Acta Mus. Napoc. v (1968), 75–98Google Scholar (with map); Gudea, N., Acta Mus. Napoc. vi (1969), 507–30.Google Scholar

⁴³ Gudea, N., AIIA 18 (1975), 71–87Google Scholar, has attempted the first systematic survey of evidence relating to the earth phase of the forts. With some revision through more recent discoveries the hypothesis of the writer remains firmly based. Roman forts in the vicinity of the Dacian citadels ought also to be added to the evidence for this phase, V. Christescu, op. cit. (note 5), 131; C. Daicoviciu and Al. Ferenczi, op. cit. (note 6), 43 ff.; Tab. Imp. Rom., Sheet L 34, pp. 107, 94, 81, 75, 68, 120, 49, 50.

⁴⁴ The archaeological evidence appears to be supported by a number of scenes on the column of Trajan, Cichorius, K., Die Reliefs der Trajanssäule i (Berlin 1896)Google Scholar; scene of fort-building in the first Dacian War (A.D. 101–2), XI–XIII, XVI–XVII, LX, LXV; and in the second War (A.D. 105–6), CXXVII, CXIII, CXXXV.

⁴⁵ Gudea, N., AIIA 18 (1975), 71 ff.Google Scholar

⁴⁶ Chirilǎ, E.–Gudea, N.–Lucǎcel, V.–Pop, C., Castrul roman de la Buciumi (Cluj 1972), 13 ff.Google Scholar; Gudea, N., Apulum 14 (1977).Google Scholar

⁴⁷ M. Macrea, op. cit. (note 7), 220, 223; D. Tudor, op. cit. (note 12), 259.

⁴⁸ From the evidence of a relatively small sample of these buildings, some resemblances are evident in those of forts built at the same date, in the northern part of the Olt (Sector XI) and in the north west part (Sector V). For these sectors some consistent typology in principia may be recognized.

⁴⁹ See note 46.

⁵⁰ M. Macrea, op. cit. (note 8), 443 and also Gudea, N. and Pop, I., Das Römerlager von Rosenau (Braşov 1970), 65.Google Scholar

⁵¹ M. Macrea, op. cit. (note 8), 443. Horedt, K., ‘Interpretǎri arheologice’, SC1VA 25 (1974), 555–8Google Scholar supports the view that the restoration of forts at Comalǎu and Boroşneu and the building of the vallum ‘Honarka’ occurred under Gallienus.

⁵² D. Tudor, op. cit. (1958), (note 12), 212; M. Macrea, op. cit. (note 8), 440.

⁵³ Ritterling, E., RE xii (1924), 1572–88 (legio).Google Scholar

⁵⁴ III Gallica: Daicoviciu, C., RE xxii, I (1953), 267Google Scholar (Porolissum), M. Macrea, op. cit. (note 7), 194; VII Gemina: Gudea, N., SCIVA 27(1976), 109–14Google Scholar; VII Claudia: D.Tudor, op.cit.(1978) (note 12), 269, M. Macrea, op. cit. (note 7), 194.

⁵⁵ Diplomas of A.D. 110, CIL xvi, 57Google Scholar and 163 (= Inscr. Dac. Rom. i, dipl. II, III).

⁵⁶ Among these may be noted the Ala Siliana, Ala Tungrorum Frontoniana, Coh. I Alpinorum and Coh. II Nervia Brittonum, still recorded in the army of Pannonia Inferior in 110 or 114, CIL xvi, 61 and 164.

⁵⁷ Notably some units not attested in other provinces, for example Coh. IIII Hispanorum, Coh. I Aelia Gaesatorum and others.

⁵⁸ Most recently Beneš, J., ‘Die römischen Auxiliarformationen im unteren Donau’, Zbornik Prače Filosofiski Fakulty Brnske Universitet 19 (1970), 159–210Google Scholar; Russu, I. I., SCIV 23 (1972), 63–77.Google Scholar

⁵⁹ Daicoviciu, C., AISC iii (1936–1940), 200 ff.Google Scholar; Protase, D., Problema continuitǎţii in lumina arheologiei şi numismaticii (Bucureşti 1966).Google Scholar For new finds, Gudea, N., SCIV 21 (1970), 299–311.Google Scholar

⁶⁰ G. Forni, op. cit. (note 10), 1094 ff.

⁶¹ G. Forni, op. cit. (note 10), 1103–32; 1139–50; 1151–63; 1181–95; 1196.

⁶² M. Macrea, op. cit. (note 8), 29–106; 436–45.