¹ Most of the literary and epigraphical sources cited here are listed by Forni, , Dizionario Epigrafico IV, 2, s.v. limes, 1074 ffGoogle Scholar. (1959) and in TLL VII, 2, fasc. IX, p. 1415; see also Fabricius, RE XIII, s.v. It is possible that I have missed inscriptions published after 1959, after which perusal of L'Année épigraphique had to suffice; the term limes appears frequently there in the notes but only twice in the texts themselves: AE 1964, 197; 1967, 555, both boundary stones.

² Mommsen, Th., ‘Der Begriff des Limes’, Gesammelte Schriften V (1908), 456–64Google Scholar.

³ See below, p. 130.

⁴ Fabricius, RE XIII, 572–5.

⁵ Webster, G., The Roman Imperial Army (1979), 46Google Scholar see further Forni's definition, op. cit. (n. 1), 1080: 'Nel significato di “frontiera fortificata e stesa a difesa dell'impero romano”, in senso molto lato e per niente affatto corrispondente all'idea moderna di confine come line ideale contrassegnata da cippi o altro …’. Usually, however, it is considered unnecessary to explain what the term means, either in ancient sources or as used in modern studies.

⁶ R. Syme, CAH XI (1936), 182 f. Hence M. Speidel, P. writes in Studien zu den Militärgrenzen Roms III, 13Google Scholar. Internationaler Limeskongress, Aalen 1983, Vorträge (1986), 657Google Scholar, of the forts on the eastern shore of the Black Sea: ‘While one is certainly justified to call this well defined sector of the Roman frontier the limes Ponticus, from a strategic point of view it may be better to speak of it as the Caucasus frontier’. In a note Speidel refers to Lefkinadze, V.A., Limes, Pontijski, Drevnei Historij (1969), 75–93Google Scholar, though he points out that neither term is attested.

⁷ Cf. Florus II, 27: Drusus (also in Germany) ‘invisum atque inaccessum in id tempus Hercynium saltum patefecit’. ‘Aperuit et stravit’ is found on milestones; see below, p. 131.

⁸ Cf. Goodyear, F. R. D., The Annals of Tacitus I (1972), pp. 315–17Google Scholar. Cf. Statius, Silvae IV, 3, 41: ‘rescindere limites’ (of the Via Domitiana); Lucretius II, 406: ‘rescindere vias’.

⁹ See the remarks by A. Oxé, BJb 114 (1906), 128.

¹⁰ Ann. I, 61, 2. See also the ‘aggeres et pontes’ constructed during the campaign against the Frisians, IV, 73. 2.

¹¹ For the interpretation of this passage see below, n. 14, with Syme, CAH XI, 162 f. Schönberger, H., JRS 59 (1969), 159Google Scholar, strangely reverted to the assumption that Frontinus meant to say that lateral barriers were constructed. See further Millar, F., Britannia 13 (1982), 14CrossRefGoogle Scholar.

¹² Cf. Anderson, J. D. C., Tacitus, Germania (1938), ad loc., p. 149Google Scholar: ‘Limitem agere is one of the technical expressions for driving such a road …’. See also Virgil, Aen. X, 513: ‘proxima quaeque metit gladio latunique per agmen ardens limitem agit ferro’ (‘he made his way through the enemy ranks’).

¹³ See further Oxé, art. cit. (n. 9), 99–133, a study more often cited than read. See in particular the conclusions at 121 f.: ‘Nur zuweilen … übernimmt der Limes die Funktion der Grenze … der Limes selbst istnie ein gefestigter Weg, geschweige denn eine Befestigung mit Palisaden oder Wall mit Graben … Zum schluss mag nur noch betont werden, dass dem reinen Begriff des Limes die man ihm oft angedichtet hat, völlig fremd sind: Grenze, Befestigung, Querweg’. These passages are similarly interpreted by Fabricius, op. cit. (n. 4), 572–5. However, in the discussion which follows, Fabricius ignores the conclusions to be drawn from his own interpretation of the sources. Similarly Forni, op. cit. (n. 1), 1079; Piganiol, A., Quintus Congressus Internationalis Limitis Romani Studiosorum (1963), 119–22Google Scholar. Mommsen, op. cit. (n. 2), 459, misinterpreted all these passages in an effort to define the term on the basis of surveyors' vocabulary.

¹⁴ See n. 4 above. Note also Vegetius' observations on the dangers of marching on a narrow road: ‘melius est praecedere cum securibus ac dolabris milites et cum labore aperire vias’ (‘it is preferable that soldiers lead the way with hatches and pickaxes and laboriously open up roads’) (III, 6).

¹⁵ Cf. Siculus Flaccus, Grom. 163, 24: ‘Territoria inter civitates, id est municipia et colonias et praefecturas, alia fluminibus finiuntur, alia summis montium iugis ac devergiis aquarum alia etiam lapidibus positis praesignibus, qui a privatorum terminorum forma differunt: alia etiam inter binas colonias limitibus perpetuis derigentur’. Here a continuously demarcated boundary is meant as opposed to boundary stones set at intervals.

¹⁶ It is curious that Mommsen does not cite this passage (see n. 13).

¹⁷ Forni, op. cit. (n. 1).

¹⁸ For discussion see Rivet, A. L. F. and Smith, Colin, The Place-Names of Roman Britain (1979), 154–6.Google Scholar

¹⁹ See TLL, s.v., for a few additional cases where limes is clearly used to mean the boundary of the empire.

²⁰ G. W. Bowersock doubts whether in Latin crossing the border would be expressed by per and wonders whether limes could mean ‘borderland’ in the present passage. This is possible, but perhaps support for my rendering may be found in Caes., BG III, 26: ‘Hostes undique circumventi desperatis omnibus rebus se per munitiones deicereet fuga salutem petere intenderunt’.

²¹ For discussion see Matthews, J. F. in Goodburn, R. and Bartholomew, P. (eds), Aspects of the Notitia Dignitatum (1976), 157–86Google Scholar, esp. 170 f. For the meaning of centenarium see Duncan-Jones, R. P., Chiron 8 (1978) 548, 552–6Google Scholar.

²² Pelham, below, n. 25. See also references above, p. 125 and n. 3.

²³ As indeed they usually are in Britain.

²⁴ Mommsen, op. cit. (n. 2), 456–64: ‘Es scheint den Limes-forschern wenig zum Bewusstsein gekommen zu sein, dass der Limes seinem Wesen nach bei alien sonst möglichen Verschiedenheiten, eine irgendwie markierte zweifache Grenze, eine äussere und eine innere fordert’.

²⁵ Essays by H. F. Pelham, ed. F. Haverfield (1911), in the paper ‘The Roman Frontier System’, 164–78, esp. 168–9. Pelham, however, went on to say that ‘there is no doubt that “limes”, like “march”, was frequently used to include not only the frontier line with its defences, but also the territory stretching along both sides of it’.

²⁶ Forni, op. cit. (n. 1), passim; MacMullen, R., Soldier and Civilian in the Later Roman Empire (1963), 39 n. 53Google Scholar. Poidebard, A., La trace de Rome dans le désert de Syrie: Le limes de Trajan à la conquête arabe (1934)Google Scholar, thought he could recognize in the Strata Diocletiana an inner line of defence and farther eastward an outer line. Later Mouterde, R. and Poidebard, A., Le Limes de Chalcis, organisation de la steppe en haute Syrie romaine (1945)Google Scholar, claimed to have discovered two lines farther north, an inner limes east of Chalcis and an outer one on the Euphrates. In the Negev Gichon, M. traced two lines: Studien zu den Militärgrenzen Roms, Vorträge des 6. Limes-Kongresses (1967), 175 ff.Google Scholar; id., Roman Frontier Studies 1979 (1980), 852–5; see also, Gutwein, K. C., Third Palestine (1981), 309–11Google Scholar. In Africa: Baradez, J., Fossatum Africae (1949), 358–60Google Scholar; in Mesopotamia: Wheeler, R. E. M., Roman Frontier Studies 1949 (1952), 112–29Google Scholar; esp. 127. For criticism see Liebeschuetz, W., Studien zu den Militärgrenzen Roms II (1977), 487–99Google Scholar, esp. 488 f.; G. W. Bowersock, HSCP 80 (1976), 219–29; B. Isaac, HSCP 88 (1984), 191 with nn. 103 and 104. Most recently Parker, S. Thomas, Romans and Saracens (1986), 6Google Scholar, argued for the existence of ‘a broad, fortified zone, not a single fortified line’ in Arabia. Elsewhere, however, he speaks of a ‘main limes’ and a ‘secondary line of defence’ (p. 142).

²⁷ See Bartholomew, P., Britannia 15 (1984), 179CrossRefGoogle Scholar n. 45: ‘The text of Ammianus XXVIII. 5. 1 provides an illustration of the dubious authenticity of ‘limes’. According to Clark and the other modern editors, the object of the Saxon attack on north-east Gaul in 370 was ‘Romanum limitem’. But the reading of M (which survives at this point) is ‘Romanum militem’; and M is followed by V. ‘Limitem’ appears only as a correction in an inferior fifteenth-century manuscript, and in Ghelen. This indicates the readiness with which Renaissance scholars thought of ‘limites’ in the context of late Roman military operations. … The decision of modern editors to accept ‘limitem’ instead of the better attested ‘militem’ must appear distinctly questionable.’ For a similar case see Seeck, Not. Dig. Oc. V, 126: ‘Comites limitum infrascriptorum’, where all the MS read: ‘militum’; cf. Bartholomew, , Britannia 10 (1979), 370CrossRefGoogle Scholar. The search for fortresses sometimes leads to even more peculiar conclusions. Gutwein, op. cit. (n. 26), translates ‘ex divisione praesidium Palaestinae’ (Jerome, Quaestiones in Genesim 21:30, PL XXIII, 969) as ‘a recent division of fortifications’. Jerome, , Vita Hilarionis 18Google Scholar (PL XXIII, 35), tells the edifying story of ‘Orion vir primarius et ditissimus urbis Aelae, quae mari Rubro imminet, a legione possessus demonum’; this then is taken as a reference to the Legion X Fretensis based at Aela.

²⁸ For further references see also Mócsy, A., Pannonia and Upper Moesia (1974), 196 f.Google Scholar

²⁹ Cf. DE I, s.v. burgus, 1053 f.; IV (1962), s.v. limes, 1089 f.; MacMullen, op. cit. (n. 26), 38 f., 57. See also the article on burgarii and the cursus publicus by Labrousse, M., Mélanges d'archéologie et d'histoire 55 (1938), 151–67Google Scholar.

³⁰ E.g. Pliny, Pan. 82, 4. It is used in the Antonine Itinerary in a list of stations along the Euphrates between Satala and Melitene, see Crow, J. in The Defence of the Roman and Byzantine East, Proceedings of a Colloquium held at the University of Sheffield in April 1986, ed. Freeman, P. and Kennedy, D. (1986), 81 f.Google Scholar

³¹ Tacitus, Hist. IV, 26: ‘dispositae per omnem ripam stationes quae Germanos vado arcerent’. The praefecti are discussed by J. F. Gilliam, TAPhA 72 (1941), 157–75. He compares them with other praefecti, those in command of the ora maritima in Mauretania (CIL XI, 5744); the ora Pontica (Pliny, Ep. X, 21; 86a) and the Baliorum insulae (ILS 9196). To the evidence in Gilliam's paper add AE 1968, 321: a praefectus ad ripam (sc. Rheni) under Claudius and Nero.

³² Gilliam points out that, unlike the later duces, this officer was subordinate to the governor of Syria. Of interest for the later meaning of the word limes is that the term ripa can also be used for a fiscal district, e.g. on an inscription from the agora of Palmyra, which mentions a curator ripae superioris et inferioris, see Gilliam, op. cit., 165 n. 35; 174 f. It is not clear what was the difference between this dux ripae and the praefecti attested elsewhere.

³³ See the catalogue of milestones by P. Thomsen, Zeitschr. D. Pal. Ver. 40 (1917), 1 ff. Aperuit is only partly applicable, for most of the road followed the alignment of an ancient caravan-route marked by Nabataean road-stations.

³⁴ For a similar use of the term see XI (III), 5, 4 (addressed to Maximianus in A.D. 291): ‘transeo limitern Raetiae repentina hostium clade promotum’.

³⁵ The passage goes on to state that the same happened in the East, where Syria had only the Euphrates for protection until the Persian kingdoms spontaneously surrendered themselves to Diocletian.

³⁶ Cf. Pliny, NH IV, 12, 58: ‘Creta inter ortum occasumque porrigitur’, and other examples.

³⁷ Cf. TLL, s.v. under the heading fines extremi imperil Romani.

³⁸ Similarly, TLL, 13, 3: ‘omnemque ilium limitem non equestribus neque pedestribus copiis sed praesentiae tuae terrore tutatus es: quantoslibet valebat exercitus Maximianus in ripa’ (‘you have protected that whole limes not with cavalry nor with infantry but with the terror inspired by your presence: Maximianus on the riverbank is worth an army ever so great’). The limes, then, is seen as distinct from the riverbank; cf. the similar argument in the case of the ‘limitis … custus’, below, p. 137.

³⁹ Cf. ibid, XVIII, 82: ‘tu Gratiane, tot imperii limites, tot flumina et lacus, veterum intersaepta regnorum … celeriore transcursu evoluis’.

⁴⁰ Contrast Forni, op. cit. (n. 1), 1081: ‘Nel basso impero, in seguito allo sdopiamento delle competenze militari e civili, e ancora in età Giustinianea il concetto di limes venne allargato fino a comprendere, in aggiunta alle strade alle fortificazioni e alle truppe vaste territori affidati all'amministrazione militare’. P. Mayerson, BASOR 262 (1986), 35–47, esp. 39, denies that the term has a formal, administrative content and refers to the U.S. concept of a ‘frontier’: ‘that part of a country which forms the border of its settled or inhabited regions’ (as defined in the Oxford English Dictionary). This suggests that it is an informal geographical notion. In later sources, such as Malalas, the term is indeed used informally, but there is no doubt that the term limes has a formal meaning in the fourth century and afterward.

⁴¹ See also Ammianus XV, 8, 6: ‘rupta limitum pace’ and other passages cited in the Thesaurus. For the city of Cercesium see Oppenheimer, et al. , Babylonia Judaica (1983), 377–82Google Scholar.

⁴² Equally anachronistic is the dux limitis mentioned in SHA, Tyr. Trig. 3, 9; 29, 1; Aurelian 13, 1.

⁴³ Cf. Eadie's comments on p. 148 with references to Petrus Patricius, fr. 14 (FHG IV, 188 f.) and Ammianus, XXV, 7, 9, where the same region with five peoples is mentioned. Needless to say, nowhere are military structures referred to. The river Indus mentioned in this source is the Tigris, cf. Aurelius Victor, de Caesaribus 13, 3; also, Eutropius VIII, 3, 2. It may be added that Malalas, Chron. XII (Dindorf, 307) refers to the ‘Indolimiton’, where Theophanes speaks of ‘Inner Persis8 (Chron. ad ann. 5793 (A.D. 293)) and Eutropius, Brev. 9, 25, of ‘ultimas regni solitudines’.

⁴⁴ Similarly: J. Rougé, REA 68 (1966), 284 f.

⁴⁵ Cf. the passages from Pan. Lat., cited above, pp. 132–3.

⁴⁶ In two other passages in the SHA limes apparently has the same meaning: Vita Probi 14, 5: ‘nisi si limes Romanus extenderetur et fieret Germania tota provincia’; Aurelian 10, 2: ‘limites restitueret’. In both cases, however, it is possible that ‘boundary’ is meant.

⁴⁷ For further discussion of this passage, see my forthcoming book, The Limits of Empire: The Roman Army in the East.

⁴⁸ See Eunapius, FHG IV, p. 14. van Berchem, D., L'Armée de Dioclétien et la réforme constantinienne (1952), 115Google Scholar.

⁴⁹ Teixidor, J., Syria 40 (1963), p. 33CrossRefGoogle Scholar, no. 1 (an inscription from Qasr Helqum on the road from Palmyra to Hit which commemorates those who were with Abgar the son of Hairan in the frontier-zone (BQST'). Starcky, J., Syria 40 (1963), 47–55CrossRefGoogle Scholar (an inscription from a road-station on the same route which commemorates ‘Abgar son of Shalman son of Zabdibol, who came to the limits of the frontier-zone (BRS QST')’). See also Matthews, op. cit. (n. 21), 168 f.

⁵⁰ See n. 26.

⁵¹ Van Berchem, , L'Armée, 5 fGoogle Scholar.

⁵² Fr. 21, FHG II, p. XVII. Tryphon concentrated his troops and encamped near the city of Chalcis (cited by Mouterde and Poidebard, op. cit. (n. 26), 4–5).

⁵³ Jerome, Vita Malchi 3 (PL XXIII, col. 56 f.); Theodoret, HE IV, 28, ed. Parmentier, 268, 8; for further references see Mouterde and Poidebard.

⁵⁴ See the collection in Mouterde and Poidebard. Note in particular the inscription recording work carried out by private individuals at the fort of el Bab, 71–3, 187 f., MUSJ 22 (1939), 65 n. 1; and the fortified horreum at et Touba where a similar inscription has been found, 197–201. There is one inscription in honour of Justinian, no. 39 on p. 209.

⁵⁵ Procopius, BP I, 17, 34.

⁵⁶ Cf. Dunand, M., Revue Biblique 40 (1931), 227–48Google Scholar.

⁵⁷ Abel: ‘Bien qu'il a quitte la divine terre de la frontière de Palestine D. a pourtant sa part au distinctions du Basileus.’ Alt: ‘D. wird schwerlich ein einfacher limitaneus gewesen sein. … Türstutz eines Grabgebäudes … dessen Errichtung auf Staatskosten [?] war vielleicht die letzte königlichen Ehrung für den im Dienst an der palästinensischen Limes gebliebenen Mann. Der Lobpreis des “göttlichen Landes” ist höffentlich mehr als eine poetische Floskel.’ Could the king of the inscription be God rather than the emperor? The phrase would then allude to the man's name, Dorotheus.

⁵⁸ It may be noted that it occurs in a chapter full of poetic and rhetorical expressions. The term limiton is also taken over into Syriac: see Michael the Syrian, Chron. IX, 16 and 26 on the region of the Balīkh and the Khabūr (pillaged by al-Mundhīr).

⁵⁹ Malalas, p. 425, clearly derived from Josephus, Ant. VIII, 6, 1 (154).

⁶⁰ Theophanes, Chron. ad ann. 6020 (A.D. 520), p. 267.

⁶¹ Mommsen, Th., ‘Das römische Militärwesen seit Diocletian’. Gesammelte Schriften VI, 209–11Google Scholar.

⁶² See for instance Grosse, R., Römische Militärgeschichte von Gallienus bis zum Beginn der byzantinischen Themenverfassung (1920), 63–70Google Scholar; Delbrück, H., Geschichte der Kriegskunst im Rahmen der politische Geschichte (1921), 231Google Scholar; Stein, E., Geschichte des spätrömischen Reiches I (1928), 90Google Scholar believes the process started under Septimius Severus; Rostovtzeff, M., The Social and Economic History of the Roman Empire, second ed. by Fraser, P. M. (1952), 426 with n. 50Google Scholar; Dillemann, L., Haute Mésopotamie orientate et pays adjacents (1962), 104Google Scholar. Van Berchem, , L'Armée, 19–24Google Scholar, did not accept the pre-Diocletianic reference.

⁶³ Seston, W., Historia 4 (1955), 286–91Google Scholar = Scripta Varia (1980), 483–90; Jones, A. H. M., The Later Roman Empire (1964), 649–53Google Scholar; see also Mazzarino, S., Aspetti sociali del quarto secolo (1951), 330–40Google Scholar, who argued that there is no evidence for the existence of farmer-soldiers in the fourth century.

⁶⁴ Clemente, G., La ‘Notitia Dignitatum’ (1968), 319 with n. 1Google Scholar; Gray, E. W., Proceedings of the African Classical Associations 12 (1973), 24Google Scholar; Cornell, T. and Matthews, J., Atlas of the Roman World (1982), 172Google Scholar.

⁶⁵ The entry on limitanei in RE, Suppl. XI (1968) contains no reference to Oates, Jones. D., Studies in the Ancient History of Northern Iraq (1968), 94Google Scholar; MacMullen, op. cit. (n. 26), 13 n. 34 and, in general, chapter 1; id., Constantine (1969), 43 f.; Piganiol, A., L'Empire Chrétien (second ed., 1972), 365Google Scholar; Chastagnol, A., L'Evolution politique, sociale et économique du monde romain, 284–363 (1982)Google Scholar; Johnson, J. S. in Hassall, M. W. C. and Ireland, R. I. (eds), De Rebus Bellicis (1979), 68Google Scholar; Williams, S., Diocletian and the Roman Recovery (1985), 97, 207, 213Google Scholar; Luttwak, E. N., The Grand Strategy of the Roman Empire (1976), 170–3Google Scholar; see also Ferrill, A., The Fall of the Roman Empire (1986), 49Google Scholar; Parker, op. cit. (above, n. 26), 149–52.

⁶⁶ See above, p. 134.

⁶⁷ Mommsen, Militärwesen, 200; Rostovtzeff, op. cit. (n. 62), 377; Stein, op. cit. (n. 62), 90. Rejected by Alföldi, A., Archaeologiai Értesitö 1 (1940), 234Google Scholar. Van Berchem, L'Armée, 21, 41 considered it prudent not to rely on the statement. A. R. Neumann, RE, Suppl. XI, s.v. limitanei, 876, admits that the reference to duces is anachronistic, but relies on the statement as partial support for his theory that the limitanei existed in some form since the second century. Grosse, op. cit. (n. 62), 63 assumed that the reference reflects the situation in the fourth century. Seston, op. cit. (n. 63), argued that the limitanei did not exist as farmer-soldiers in the time of Diocletian.

⁶⁸ Mann, J. C., Legionary Recruitment and Veteran Settlement during the Principate (1983), 67Google Scholar suggests that the passage refers in fact to veterans, not to serving troops.

⁶⁹ Pointed out by van Berchem, op. cit. (n. 48), 34, 101.

⁷⁰ For soldiers as farmers see MacMullen, op. cit. (n. 26), chapter 1.

⁷¹ Mann, J. C. in CBA Research Report No. 18: The Saxon Shore, ed. Johnson, D.E. (1977), 11Google Scholar.

⁷⁸ Van Berchem, , L'Armée, 17–18Google Scholar.

⁷³ See, for instance, Josephus, Ant. XV, 11,3 (401): ‘within this wall (ἐνδοτέρω δἐ [SC. τοῦ τείχους]) and on the very summit (of the Temple Mount) ran another wall of stone’. Cf. LSJ, s.v.; Stephanus, H., Thesaurus Linguae Graecae III, 1041 fGoogle Scholar.

⁷⁴ For criticism of van Berchem's interpretation see also Mann, op. cit. (n. 71), 12 with n. 8, observing that permanent ducates are attested under Diocletian only in Valeria, Scythica, and Augusta Euphratensis.

⁷⁵ John Lydus, De Mag. II, 11. Cf. Aurelius Victor, liber de Caesaribus 41, 12: ‘Quo excruciato, ut fas erat, servili aut latronum more, condenda urbe formandisque religionibus ingentem animum avocavit (sc. Constantinus), simul novando militiae ordine’.

⁷⁶ I am grateful to Professor Tony Honoré for clarification. J. C. Mann has discussed the term ripenses and the development of the various categories of troops in two articles: CBA Research Report No. 18 (above, n. 71), 11–15 and in R. Goodburn and P. Bartholomew (eds), Aspects of the Notitia Dignitatum (1976), 2.

⁷⁷ For discussion see van Berchem, , L'Armée, 83–7Google Scholar. SHA, Aurelian XXVI, 38, 4: ‘hi compressi sunt septem milibus Lemberiorum et ripariensium et castrianorum et Daciscorum interemptis’, is merely another of the SHA's anachronisms,

⁷⁸ For instance in RE, ibid.

⁷⁹ Grosse, op. cit. (above, n. 62), 66, already pointed out that part of the limitanei were not stationed in frontier zones, such as Isauria and Upper Egypt.

⁸⁰ Cyril of Scythopolis, Vita S. Sabae 73, ed. Schwartz, 178.

⁸¹ Similarly, Justinian's novel 102 regarding Arabia.

⁸² Alt, p. 5, would then have missed the point in his translation: ‘[Es sollen ihre Abgaben entrichten die … der] jeweiligen Duces, sowie die treuergebenen unter[-stellten] Grenzsoldaten [und die übrigen Steuerpflichti]gen Jahr für Jahr in folgender Weise’. It is not impossible that the [??] of the Duces and the soldiers were to receive payment instead of paying others. Similar texts from the reign of Anastasius were found in Arabia, see E. Littmann et al., Publications of the Princeton University Archaeological Expedition to Syria, in 1904–1905 III, Greek and Latin Inscriptions A2 (Leiden, 1910), p. 33, frs 15–19; IGLS XIII, I, no. 9046; J. Marcillet-Jaubert, ADAJ 24 (1980), 122 f.; Kennedy, D., Archaeological Explorations (1982), 44–8Google Scholar: a text concerned with the payment of penalties by officials, which makes reference to, among others, ‘those in charge in the limes of Palaestina and of Euphratensis’ (i.e. Commagene). For the Beer Sheva edict see also P. Mayerson, ZPE 64 (1986), 141–8.

⁸³ See Matthews, loc. cit. (above n. 21).

⁸⁴ For Military land in the earlier period in the western provinces see F. Vittinghoff, Ac. Naz. Lincei 194 (1974), 109–24.

⁸⁵ For references to supplies to the limes see above, with CTh VIII, 4, 6 (358); CJ XII, 8 (386).

⁸⁶ Jones, op. cit. (n. 63), 651.

⁸⁷ See Jones, op. cit., 663 for the suggestion that the limitanei and the comitatenses were equally poor soldiers.

⁸⁸ For reservations see also Luttwak, op. cit. (n. 65), 172 f., and for discussion of the quality of frontier soldiers in the late empire, de Ste. Croix, G. E. M., The Class Struggle in the Ancient Greek World (1981), 261 ffGoogle Scholar.

⁸⁹ This is to be inferred from Procopius II, 16, 17; see also 1, 13, 5; 11, 8, 2; 11, 19, 33. Liebeschuetz, W., Studien zu den Militärgrenzen Roms II (1977), 497Google Scholar, assumes that this force was separate from the limitanei, because these were intended for defence of their immediate locality. However, the troops from Phoenicia Libanensis used by Belisarius were intended for the defence of their own province, as emphasized by Procopius, who represents Belisarius' step as unusual. Moreover, the conclusion of this paper is that all troops under the command of a dux limitis were limitanei.