v. 7, at that time merely Lesser Armenia remained Roman, this may not

be incorrect, in so far as the dependence of the vassal-king of Great Armenia after the peace was doubtless merely nominal.

[83] The Biblical account (1 Kings, ix. 18) as to the building of the town Thamar in Idumaea by king Solomon has only been transferred to Tadmor by a misunderstanding doubtless old; at all events the erroneous reference of it to this town among the later Jews (2 Chron. viii. 4, and the Greek translation of 1 Kings, ix. 18) form the oldest testimony for its existence (Hitzig, _Zeitschr. der deutschen morgenl. Gesellschaft_, viii. 222).

[84] This is nowhere expressly stated; but all the circumstances tell in favour of it. That the Romano-Parthian frontier, before the Romans established themselves on the left bank of the Euphrates, was on the right a little below Sura, is most distinctly said by Pliny (_H. N._ v. 26, 89: _a Sura proxime est Philiscum_--comp. p. 95, note 1--_oppidum Parthorum ad Euphratem; ab eo Seleuciam dierum decem navigatio_), and there it remained till the erection of the province of Mesopotamia under Severus. The Palmyrene of Ptolemy (v. 15, 24, 25) is a district of Coele-Syria, which seems to embrace a good part of the territory to the south of Palmyra, but certainly reaches as far as the Euphrates and includes Sura; other urban centres besides Palmyra seem not to be mentioned, and there is nothing to stand in the way of our taking this large district as civic territory. So long in particular as Mesopotamia was Parthian, but subsequently also with reference to the adjoining desert, a permanent protection of the frontier could not here be dispensed with; as indeed in the fourth century, according to the tenor of the Notitia, Palmyrene was strongly occupied, the northern portion by the troops of the Dux of Syria, Palmyra itself and the southern half by those of the Dux of Phoenice. That in the earlier imperial period no Roman troops were stationed here, is vouched for by the silence of authors and the absence of inscriptions, which in Palmyra itself are numerous. If in the Tabula Peutingeriana it is remarked under Sura: _fines exercitus Syriatici et commercium barbarorum_, that is, “here end the Roman garrisons and here is the place of exchange for the traffic of the barbarians,” this is only saying, what at a later time is repeated by Ammianus (xxiii. 3, 7: _Callinicum munimentum robustum et commercandi opimitate gratissimum_) and further by the emperor Honorius (_Cod. Just._ iv. 63, 4), that Callinicon was one of the few entrepôts devoted to the Romano-barbarian frontier-traffic; but it does not at all follow from this as regards the time when the Tabula originated, that these imperial troops were stationed there, since in fact the Palmyrenes in general belonged to the Syrian army and might be thought of in using the expression _exercitus Syriaticus_. The city must have furnished a force of its own in a way similar to that of the princes of Numidia and of Panticapaeum. By this means alone we come to understand as well the rejection of the troops of Antonius as the attitude of the Palmyrenes in the troubles of the third century, and not less the emergence of the _numeri Palmyrenorum_ among the military novelties of this epoch.

[85] Ammianus, xxiii. 5, 2: _Cercusium ... Diocletianus exiguum ante hoc et suspectum muris turribusque circumdedit celsis, ... ne vagarentur per Syriam Persae ita ut paucis ante annis cum magnis provinciarum contigerat damnis_. Comp. Procopius _de aed._ ii. 6. Perhaps this place is not different from the Φάλγα or Φάλιγα of Isidorus of Charax (_mans. Parth._ 1; Stephanus Byz. _s. v._) and the _Philiscum_ of Pliny (p. 94, note).

[86] Of the seven dedications, hitherto found outside of Palmyra, to the Palmyrene Malach Belos the three brought to light in Rome (_C. I. L._ vi. 51, 710; _C. I. Gr._ 6015) have along with a Greek or Latin also a Palmyrene text, two African (_C. I. L._ viii. 2497, 8795 add.) and two Dacian (_Arch. epig. Mitth. aus Oesterreich_, vi. 109, 111) merely Latin. One of the latter was set up by P. Aelius Theimes, a _duoviralis_ of Sarmizegetusa, evidently a native of Palmyra, _diis patriis Malagbel et Bebellahamon et Benefal et Manavat_.

[87] Whence these names of the months come, is not clear; they first appear in the Assyrian cuneiform writing, but are not of Assyrian origin. In consequence of the Assyrian rule they then remained in use within the sphere of the Syrian language. Variations are found; the second month, the Dios of the Greek-speaking Syrians, our November, is called among the Jews Markeshvan, among the Palmyrenes Kanun (Waddington, n. 2574_b_). We may add that these names of the months, so far as they came to be applied within the Roman empire, are adapted, like the Macedonian, to the Julian calendar, so that only the designation of the month differs; the year-beginning (1 Oct.) of the Syro-Roman year finds uniformly application to the Greek as to the Aramaean appellations.

[88] _E.g._ Archon, Grammateus, Proedros, Syndikos, Dekaprotoi.

[89] This is shown by the inscription of Palmyra (_C. I. Gr._ 4491, 4492 = Waddington 2600 = Vogué, _Insc. sém. Palm._ 22) set up to this Hairanes in the year 251 by a soldier of the legion stationed in Arabia. His title is in Greek ὁ λαμπρότατος συγκλητικός, ἔξα[ρχος (= _princeps_) Παλμυ]ρηνῶν, in Palmyrene “illustrious senator, head of Tadmor.” The epitaph (_C. I. Gr._ 4507 = Waddington 2621 = Vogué, 21) of the father of Hairanes, Septimios Odaenathos, son of Hairanes, grandson of Vaballathos, great-grandson of Nassoros, gives to him also senatorial rank.

[90] Certainly the father of this Odaenathus is nowhere named; but it is as good as certain that he was the son of the Hairanes just named, and bore the name of his grandfather. Zosimus, too, i. 39, terms him a Palmyrene distinguished from the days of his forefathers by the government (ἄνδρα Παλμυρηνὸν καὶ ἐκ προγόνων τῆς παρὰ τῶν βασιλέων ἀξιωθέντα τίμης).

[91] In the inscription Waddington 2603 = Vogué 23, which the guild of gold and silver workers of Palmyra set up in the year 257 to Odaenathus he is called ὁ λαμπρότατος ὑπατικός, and so _vir consularis_, and in Greek δεσπότης, in Syriac _mâran_. The former designation is not a title of office, but a statement of the class in which he ranked; so _vir consularis_ stands not unfrequently after the name quite like _vir clarissimus_ (_C. I. L._ x. p. 1117 and elsewhere), and ὁ λαμπρότατος ὑπατικός is found alongside of and before official titles of various kinds, _e.g._ that of the proconsul of Africa (_C. I. Gr._ 2979, where λαμπρότατος is absent), of the imperial legate of Pontus and Bithynia (_C. I. Gr._ 3747, 3748, 3771) and of Palestine (_C. I. Gr._ 4151), of the governor of Lycia and Pamphylia (_C. I. Gr._ 4272); it is only in the age after Constantine that it is in combination with the name of the province employed as an official title (_e.g._ _C. I. Gr._ 2596, 4266_e_). From this, therefore, no inference is to be drawn as to the legal position of Odaenathus. Likewise, in the Syriac designation of “lord,” we may not find exactly the ruler; it is also given to a procurator (Waddington 2606 = Vogué 25).

[92] Syria in the imperial period formed an imperial customs-district of its own, and the imperial dues were levied not merely on the coast but also at the Euphrates-frontier, in particular at Zeugma. Hence it necessarily follows that farther to the south, where the Euphrates was no longer in the Roman power, similar dues were established on the Roman eastern frontier. Now a decree of the council of Palmyra of the year 137 informs us that the city and its territory formed a special customs-district, and the dues were levied for the benefit of the town upon all goods imported or exported. That this territory lay beyond the imperial dues, is probable--first, because, if there had existed an imperial customs-line enclosing the Palmyrene territory, the mention of it could not well be omitted in that detailed enactment; secondly, because a community of the empire enclosed by the imperial customs-lines would hardly have had the right of levying dues at the boundary of its territory to this extent. We shall thus have to discern in the levying of dues by the community of Palmyra the same distinctive position which must be attributed to it in a military point of view. Perhaps, on the other hand, there was an impost laid on it for the benefit of the imperial exchequer, possibly the delivering up of a quota of the produce of the dues or a heightened tribute. Arrangements similar to those for Palmyra may have existed also for Petra and Bostra; for goods were certainly not admitted here free of dues, and according to Pliny, _H. N._ xii. 14, 65, imperial dues from the Arabic frankincense exported by way of Gaza seem only to have been levied at Gaza on the coast. The indolence of Roman administration was stronger than its fiscal zeal; it may frequently have devolved the inconvenient tolls of the land-frontier away from itself on the communities.

[93] These caravans (συνοδίαι) appear on the Palmyrene inscriptions as fixed companies, which undertake the same journeys beyond doubt at definite intervals under their foreman (συνοδιάρχης, Waddington, 2589, 2590, 2596); thus a statue is erected to such a one by “the merchants who went down with him to Vologasias” (οἱ σὺν αὐτῷ κατελθόντες εἰς Ὀλογεσιάδα ἔνποροι, Waddington 2599 of the year 247), or “up from Forath (comp. Pliny, _H. N._ vi. 28, 145) and Vologasias” (οἱ συναναβάντες μετ’ αὐτοῦ ἔμποροι ἀπὸ Φοράθου κὲ Ὀλογασιάδος, Waddington, 2589 of the year 142), or “up from Spasinu Charax” (οἱ σὺν αὐτῷ ἀναβάντες ἀπὸ Σπασίνου Χάρακος, Waddington, 2596 of the year 193; similarly 2590 of the year 155). All these conductors are men of standing furnished with lists of ancestors; their honorary monuments stand in the great colonnade beside those of queen Zenobia and her family. Specially remarkable is one of them, Septimius Vorodes, of whom there exists a series of honorary monuments of the years 262-267 (Waddington, 2606-2610); he, too, was a caravan-head (ἀνακομίσαντα τὰς συνοδίας ἐκ τῶν ἰδίων καὶ μαρτυρηθέντα ὑπὸ τῶν ἀρχεμπόρων, Waddington, n. 2606 _a_; consequently he defrayed the costs of the journey back for the whole company, and was on account of this liberality publicly praised by the wholesale traders). But he filled not merely the civic offices of _strategos_ and _agoranomos_, he was even imperial procurator of the second class (_ducenarius_) and _argapetes_ (p. 104, note 1).

[94] According to the Greek account (Zonaras, xii. 21) king Tiridates takes refuge with the Romans, but his sons take the side of the Persians; according to the Armenian, king Chosro is murdered by his brethren, and Chosro’s son, Tiridates, fled to the Romans (Gutschmid, _Zeitschrift der deutschen morgenl. Gesellsch._ xxxi. 48). Perhaps the latter is to be preferred.

[95] The only fixed chronological basis is furnished by the Alexandrian coins, according to which Valerian was captured between 29th August 259 and 28th August 260. That after his capture he was no longer regarded as emperor, is easily explained, seeing that the Persians compelled him in their interest to issue orders to his former subjects (continuation of Dio, _fr._ 3).

[96] The better accounts simply know the fact that Valerian died in Persian captivity. That Sapor used him as a footstool in mounting his horse (Lactantius, _de Mort. persec._ 5; Orosius, vii. 22, 4; Victor, _Ep._ 33), and finally caused him to be flayed (Lactantius, _l. c._; Agathias, iv. 23; Cedrenus, p. 454) is a Christian invention--a requital for the persecution of the Christians ordered by Valerian.

[97] The tradition according to which Mareades (so Ammianus, xxiii. 5, 3; Mariades in Malalas, 12, p. 295; Mariadnes in contin. of Dio, _fr._ 1), or, as he is here called, Cyriades, had himself proclaimed as Augustus (_Vit. trig. tyr._ 1) is weakly attested; otherwise there might doubtless be found in it the occasion why Sapor caused him to be put to death.

[98] He is called Callistus in the one tradition, doubtless traceable to Dexippus, in Syncellus, p. 716, and Zonaras, xii. 23, on the other hand, Ballista in the biographies of the emperors and in Zonaras, xii. 24.

[99] He was, according to the most trustworthy account, _procurator summarum_ (ἐπὶ τῶν καθόλου λόγων βασιλέως: Dionysius in Eusebius, _H. E._ vii. 10, 5), and so finance-minister with equestrian rank; the continuator of Dio (_fr._ 3 Müll.) expresses this in the language of the later age by κόμης τῶν θησαυρῶν καὶ ἐφεστὼς τῇ ἀγορᾷ τοῦ σίτου.

[100] At least according to the report, which forms the basis of the imperial biographies (_vita Gallieni_, 3, and elsewhere). According to Zonaras, xii. 24, the only author who mentions besides the end of Callistus, Odaenathus caused him to be put to death.

[101] That Odaenathus, as well as after him his son Vaballathus (apart, of course, from the time after the rupture with Aurelian), were by no means Augusti (as the _vit. Gallieni_, 12, erroneously states), is shown both by the absence of the name of Augustus on the coins and by the title possible only for a subject, _v(ir) c(onsularis)_ = ὑ(πατικός), which, like the father (p. 97, note 3), the son still bears. The position of governor is designated on the coins of the son by _im(perator) d(ux) R(omanorum)_ = αὐτ(οκράτωρ) σ(τρατηγός); in agreement therewith Zonaras (xii. 23, and again xii. 24) and Syncellus (p. 716) state that Gallienus appointed Odaenathus, on account of his victory over the Persians and Ballista, as στρατηγὸς τῆς ἐῴας, or πάσης ἀνατολῆς; and the biographer of Gallienus, 10, that he _obtinuit totius Orientis imperium_. By this is meant all the Asiatic provinces and Egypt; the added _imperator_ = αὐτοκράτωρ (comp. _Trig. tyr._ 15, 6, _post reditum de Perside_--Herodes son of Odaenathus--_cum patre imperator est appellatus_) is intended beyond doubt to express the freer handling of power, different from the usual authority of the governor.--To this was added further the now formally assumed title of a king of Palmyra (_Trig. tyr._ 15, 2: _adsumpto nomine regali_), which also the son bears, not on the Egyptian, but on the Syrian coins. The circumstance that Odaenathus is probably called _melekh malke_, “king of kings,” on an inscription set up in August 271, and so after his death and during the war of his adherents with Aurelian (Vogué, n. 28), belongs to the revolutionary demonstrations of this period and forms no proof for the earlier time.

[102] The numerous inscriptions of Septimius Vorodes, set up in the years 262 to 267 (Waddington, 2606-2610), and so in the lifetime of Odaenathus, all designate him as imperial procurator of the second class (_ducenarius_), but at the same time partly by the title ἀργαπέτης, which Persian word, current also among the Jews, signifies “lord of a castle,” “viceroy” (Levy, _Zeitsch. der deutschen morgenl. Gesellschaft_, xviii. 90; Nöldeke, _ib._ xxiv. 107), partly as δικαιοδότης τῆς μητροκολωνίας, which, beyond doubt, is in substance at any rate, if not in language, the same office. Presumably we must understand by it that office on account of which the father of Odaenathus is called the “head of Tadmor” (p. 97, note 2); the one chief of Palmyra competent for martial law and for the administration of justice; only that, since extended powers were given to the position of Odaenathus, this post as a subordinate office is filled by a man of equestrian rank. The conjecture of Sachau (_Zeitschr. der d. morgenl. Gesellsch._ xxxv. 738) that this Vorodes is the “Wurud” of a copper coin of the Berlin cabinet, and that both are identical with the elder son of Odaenathus, Herodes, who was killed at the same time with his father, is liable to serious difficulties. Herodes and Orodes are different names (in the Palmyrene inscription, Waddington, 2610, the two stand side by side); the son of a senator cannot well fill an equestrian office; a procurator coining money with his image is not conceivable even for this exceptional state of things. Probably the coin is not Palmyrene at all. “It is,” von Sallet writes to me, “probably older than Odaenathus, and belongs perhaps to an Arsacid of the second century A.D.; it shows a head with a headdress similar to the Sassanid; the reverse, S C in a chaplet of laurel, appears imitated from the coins of Antioch.”--If subsequently, after the breach with Rome in 271, on an inscription of Palmyra (Waddington, 2611) two generals of the Palmyrenes are distinguished, ὁ μέγας στρατηλάτης, the historically known Zabdas, and ὁ ἐνθάδε στρατηλάτης, Zabbaeos, the latter is, it may be presumed, just the Argapetes.

[103] The state of the case speaks in favour of this; evidence is wanting. In the imperial biographies of this epoch the Armenians are wont to be adduced among the border peoples independent of Rome (_Valer._ 6; _Trig. tyr._ 30, 7, 18; _Aurel._ 11, 27, 28, 41); but this is one of their quite untrustworthy elements of embellishment.

[104] This more modest account (Eutropius, ix. 10; _vita Gallieni_, 10; _Trig. tyr._ 15, 4; Zos. i. 39, who alone attests the two expeditions) must be preferred to that which mentions the capture of the city (Syncellus, p. 716).

[105] This is shown by the accounts as to Carinus (cont. of Dio, p. 8) and as to Rufinus (p. 106, note 2). That after the death of Odaenathus Heraclianus, a general acting on Gallienus’s orders against the Persians, was attacked and conquered by Zenobia (_vita Gallieni_, 13, 5), is in itself not impossible, seeing that the princes of Palmyra possessed _de iure_ the chief command in all the East, and such an action, even if it were suggested by Gallienus, might be treated as offending against this right, and this would clearly indicate the strained relation; but the authority vouching it is so bad that little stress can be laid on it.

[106] This we learn from the characteristic narrative of Petrus, _fr._ 10, which is to be placed before _fr._ 11.

[107] The account of the continuator of Dio, _fr._ 7, that the old Odaenathus was put to death, as suspected of treason, by one (not elsewhere mentioned) Rufinus, and that the younger, when he had impeached this person at the bar of the emperor Gallienus, was dismissed on the declaration of Rufinus that the accuser deserved the same fate, cannot be correct as it stands. But Waddington’s proposal to substitute Gallus for Gallienus, and to recognise in the accuser the husband of Zenobia, is not admissible, since the father of this Odaenathus was Hairanes, in whose case there existed no ground at all for such an execution, and the excerpt in its whole character undoubtedly applies to Gallienus. Rather must the old Odaenathus have been the husband of Zenobia, and the author have erroneously assigned to Vaballathus, in whose name the charge was brought, his father’s name.

[108] All the details which are current in our accounts of Zenobia originate from the imperial biographies; and they will only be repeated by such as do not know this source.

[109] The name Vaballathus is given, in addition to the coins and inscriptions, by Polemius Silvius, p. 243 of my edition, and the biographer of Aurelian, c. 38, while he describes as incorrect the statement that Odaenathus had left two sons, Timolaus and Herennianus. In reality these two persons emerging simply in the imperial biographies appear along with all that is connected with them as invented by the writer, to whom the thorough falsification of these biographies is to be referred. Zosimus too, i. 59, knows only of one son, who went into captivity with his mother.

[110] Whether Zenobia claimed for herself formal joint-rule, cannot be certainly determined. In Palmyra she names herself still after the rupture with Rome merely βασιλίσση (Waddington, 2611, 2628), in the rest of the empire she may have laid claim to the title _Augusta_, Σεβαστή; for, though there are no coins of Zenobia from the period prior to the breach with Rome, yet on the one hand the Alexandrian inscription with βασιλισσης καὶ βασιλέως προσταξάντων (_Eph. epigr._ iv. p. 25, p. 33) cannot lay any claim to official redaction, and on the other hand the inscription of Byblos, _C. I. Gr._ 4503 b = Waddington, n. 2611, gives in fact to Zenobia the title Σεβαστή alongside of Claudius or Aurelian, while it refuses it to Vaballathus. This is so far intelligible, as Augusta was an honorary designation, Augustus an official one, and thus that might well be conceded to the woman which was refused to the man.

[111] So Zosimus, i. 44, narrates the course of events with which Zonaras, xii. 27 and Syncellus, p. 721, in the main agree. The report in the life of Claudius, c. 11, is more displaced than properly contradictory; the first half is only indicated by the naming of Saba; the narrative begins with the successful attempt of Timagenes to ward off the attack of Probus (here Probatus). The view taken of this by me in Sallet (_Palmyra_, p. 44) is not tenable.

[112] The determination of the date depends on the fact that the usurpation-coins of Vaballathus cease already in the fifth year of his Egyptian reign, _i.e._ 29th August 270-71; the fact that they are very rare speaks for the beginning of the year. With this essentially agrees the circumstance that the storming of the Prucheion (which, we may add, was no part of the city, but a locality close by the city on the side of the great oasis; Hieronymus, _vit. Hilarionis_, c. 33, 34, vol. ii. p. 32 Vall.) is put by Eusebius in his Chronicle in the first year of Claudius, by Ammianus, xxii. 16, 15, under Aurelian; the most exact report in Eusebius, _H. Eccl._ vii. 32, is not dated. The reconquest of Egypt by Probus stands only in his biography, c. 9; it may have happened as it is told, but it is possible also that in this thoroughly falsified source the history of Timagenes has been _mutatis mutandis_ transferred to the emperor.

[113] This is perhaps what the report on the battle of Hemesa, extracted by Zosimus, i. 52, wished to bring out, when it enumerates among the troops of Aurelian the Dalmatians, Moesians, Pannonians, Noricans, Raetians, Mauretanians, and the guard. When he associates with these the troops of Tyana and some divisions from Mesopotamia, Syria, Phoenice, Palestine, this applies beyond doubt to the Cappadocian garrisons, which had joined after the capture of Tyana, and to some divisions of the armies of the East favourably disposed to Rome, who went over to Aurelian upon his marching into Syria.

[114] By mistake Eutropius, ix. 13, places the decisive battle _haud longe ab Antiochia_: the mistake is heightened in Rufius, c. 24 (on whom Hieronymus, _chron. a. Abr._ 2289 depends), and in Syncellus, p. 721, by the addition _apud Immas_, ἐν Ἴμμαις, which place, lying 33 Roman miles from Antioch on the road to Chalcis, is far away from Hemesa. The two chief accounts, in Zosimus and the biographer of Aurelian, agree in all essentials.

[115] This is the name given by Zosimus, i. 60, and Polemius Silvius, p. 243; the Achilleus of the biographer of Aurelian, c. 31, seems a confusion with the usurper of the time of Diocletian.--That at the same time in Egypt a partisan of Zenobia and at the same time robber-chief, by name Firmus, rose against the government, is doubtless possible, but the statement rests only on the imperial biographies, and the details added sound very suspicious.

[116] The chronology of these events is not quite settled. The rarity of the Syrian coins of Vaballathus as Augustus shows that the rupture with Aurelian (end of 270) was soon followed by the conquest. According to the dated inscriptions of Odaenathus and Zenobia of August 271 (Waddington, 2611), the rule of the queen was at that time still intact. As an expedition of this sort, from the conditions of the climate, could not well take place otherwise than in spring, the first capture of Palmyra must have ensued in the spring of 272. The most recent (merely Palmyrene) inscription which we know from that quarter (Vogué, n. 116) is of August 272. The insurrection probably falls at this time; the second capture and the destruction somewhere in the spring of the year 273 (in accordance with which, I. 166, note 1, is to be corrected).

[117] It throws no light on the position of the Armenians, that in descriptions otherwise thoroughly apocryphal (_vita Valer._ 6; _vita Aurel._ 37, 28) the Armenians after the catastrophe of Valerian keep to the Persians, and appear in the last crisis of the Palmyrenes as allies of Zenobia by the side of the Persians; both are obvious consequences from the general position of things. That Aurelian did not subdue Armenia any more than Mesopotamia, is supported in this case partly by the silence of the authorities, partly by the account of Synesius (_de regno_, p. 17) that the emperor Carinus (rather Carus) had in Armenia, close to the frontier of the Persian territory, summarily dismissed a Persian embassy, and that the young Persian king, alarmed by its report, had declared himself ready for any concession. I do not see how this narrative can be referred to Probus, as von Gutschmid thinks (_Zeitschr. d. deutsch. morgenl. Gesell._ xxxi. 50); on the other hand it suits very well the Persian expedition of Carus.

[118] The reconquest of Mesopotamia is reported only by the biographer,