iii. 11, 6) are not military districts, but, as is apparent with

special clearness in Ptolemy, land-districts, which correspond with the tribes (στρατηγία Μαιδική, Βεσσική κ. τ. λ.) and form a contrast to the towns. The designation στρατηγός has, just like _praetor_, lost subsequently its original military value. Here perhaps the analogy of Egypt, which likewise was divided into urban domains under urban magistrates and into land-districts under _strategoi_, served primarily as a basis. A στρατηγὸς Ἀστικῆς περὶ Πέρινθον from the Roman period occurs in _Eph. epigr._ ii. p. 252.]

[206: In Deultus, the _colonia Flavia Pacis Deultensium_, veterans of the eighth legion, were provided for (_C. I. L._ vi. 3828). Flaviopolis on the Chersonese, the old Coela, was certainly not a colony (Plin. iv. 11, 47), but belongs to the peculiar settlement of the imperial menials on this domanial possession (_Eph. epigr._ v. p. 83).]

[207: This town Νικόπολις ἡ περὶ Αἷμον of Ptolem. iii. 11, 7, Νικόπολις πρὸς Ἴστρον of the coins, the modern Nikup on the Jantra, belongs to lower Moesia geographically, and, as the names of governors on the coins show, since Severus also administratively; but not merely does Ptolemy adduce it in Thrace, but the places where the Hadrianic terminal stones (_C. I. L._ iii. 749, comp. p. 992) are found, appear to assign it likewise to Thrace. As this Greek inland town fitted neither the Latin town-communities of lower Moesia nor the κοινόν of the Moesian Pontus, it was assigned at the first organising of the relations to the κοινόν of the Thracians. Subsequently it must, no doubt, have been attached to one or the other of those Moesian groups.]

[208: The κοινὸν τῆς Πενταπόλεως is found on an inscription of Odessus, _C. I. Gr._ 2056 _c._, which may fairly belong to the earlier imperial period, the Pontic Hexapolis, on two inscriptions of Tomis probably of the second century A.D. (Marquardt, _Staatsverw._ i.² p. 305; Hirschfeld, _Arch. epigr. Mitth._ vi. 22). The Hexapolis in any case, and in accordance therewith probably also the Pentapolis, must have been brought into harmony with the Roman provincial boundaries, that is, must have included in it the Greek towns of lower Moesia. These are also found, if we follow the surest guides,--the coins of the imperial period. There were six mints (apart from Nicopolis, p. 282, note) in lower Moesia: Istros, Tomis, Callatis, Dionysopolis, Odessus, and Marcianopolis, and, as the last town was founded by Trajan, the Pentapolis is thereby explained. Tyra and Olbia hardly belonged to it; at least the numerous and loquacious monuments of the latter town nowhere show any link of connection with this city-league. It is called κοινὸν τῶν Ἑλλήνων on an inscription of Tomis, printed in the Athenian _Pandora_ of 1st June 1868 [and in _Anc. Gr. Inscr. in the British Museum_, ii. n. 175]: Ἀγαθῆ τύχη. Κατὰ τὰ δόξαντα τῆ κρατήστη βουλῆ καὶ τῶ λαμπροτάτω δήμω τῆς λαμπροτάτης μητροπόλεως καὶ αʹ τοῦ εὐωνύμου πόντου Τόμεως τὸν Ποντάρχην Αὐρ. Πρείσκιον Ἀννιανὸν ἄρξαντα τοῦ κοινοῦ τῶν Ἑλλήνων καὶ τῆς μητρ[ο]πόλεως τήν αʹ ἀρχὴν ἁγνῶς, καὶ ἀρχιερασάμενον, τῆν δι' ὅπλων καὶ κυνηγεσίων ἐνδόξως φιλοτειμίαν μὴ διαλιπόντα, ἀλλὰ καὶ βουλευτὴν καὶ τῶν πρωτευόντων φλαβίας Νέας πόλεως, καὶ τὴν ἀρχιέρειαν σύμβιον αὐτοῦ Ἰουλίαν Ἀπολαύστην πάσης τειμῆς χάριν.]

[209: This is shown by the remarkable inscription in Allard (_La Bulgarie orientale_, Paris, 1863, p. 263): Θεῶ μεγάλω Σαράπ[ιδι καὶ] τοῖς συννάοις θεοῖς [καὶ τῶ αὐ]τοκράτορι Τ. Αἰλίω Ἀδριαν[ῶ Ἀ]ντωνείνω Σεβαστῶ Εὐσεβ[εῖ] καὶ Μ. Αὐρηλίω Οὐήρω Καίσαρι Καρπίων Ἀνουβίωνος τῶ οἴκω τῶν Ἀλεξανδρέων τὸν βωμὸν ἐκ τῶν ἰδίων ἀνέθηκεν ἔτους κγʹ [μηνὸς] φαρμουθὶ αʹ ἐπὶ ἱερέων [Κ]ορνούτου τοῦ καὶ Σαραπίωνος [Πολύ]μνου τοῦ καὶ Λον[γείνου]. The mariner's guild of Tomis meets us several times in the inscriptions of the town.]

[210: Olbia, constantly assailed in war and often destroyed, suffered, according to the statement of Dio (_Borysth._ p. 75, n.), about 150 years before his time, _i.e._ somewhat before the year 100 A.D., and so probably in the expedition of Burebista (iv. 305), its last and most severe conquest (τὴν τελευταίαν καὶ μεγίστην ἅλωσιν). Εἷλον δὲ, Dio continues, καὶ ταύτην Γέται καὶ τὰς ἄλλας τὰς ἐν τοῖς ἀριστεροῖς τοῦ Πόντου πόλεις μέχρι Ἀπολλωνίας (Sozopolis or Sizebolu, the last Greek town of note on the Pontic west coast): ὅθεν δὴ καὶ σφόδρα ταπεινὰ τὰ πράγματα κατέστη τῶν ταύτῃ Ελλήνων, τῶν μὲν οὐκέτι συνοικισθεισῶν πόλεων, τῶν δὲ φαύλως καὶ τῶν πλείστων βαρβάρων εἰς αὐτὰς συῤῥυέντων. The young citizen of rank with a marked Ionic physiognomy, with whom Dio then meets, who has slain or captured numerous Sarmatians, and though not acquainted with Phocylides, knows Homer by heart, wears mantle and trousers after the Scythian fashion, and a knife in his girdle. The townsmen all wear long hair and a long beard, and only one has shorn both, which is suspected in him as a token of servile attitude towards the Romans. Thus a century later matters there looked quite such as Ovid describes them at Tomis.]

[211: Quite commonly the father has a Scythian name and the son a Greek, or conversely; _e.g._ an inscription of Olbia set up under or after Trajan (_C. I. Gr._ 2074) records six _strategoi_, M. Ulpius Pyrrhus son of Arseuaches, Demetrios son of Xessagaros, Zoilos son of Arsakes, Badakes son of Radanpson, Epikrates son of Koxuros, Ariston son of Vargadakes.]

[212: As Asander reckoned his archonship probably from the very time of his revolt from Pharnaces, and so from the summer of 707 {47.}, and assumes the royal title already in the fourth year of his reign, this year may warrantably be put in the autumn 709-710 {45-44.}, and the confirmation have thus been the work of Caesar. Antonius cannot well have bestowed it, as he only came to Asia at the end of 712 { 42.}; still less can we think of Augustus, whom the pseudo-Lucian (_Macrob._ 15) names, interchanging father and son.]

[213: Mithradates, whom Claudius in the year 41 made king of Bosporus, traced back his descent to Eupator (Dio, lx. 8; Tacitus, _Ann._ xii. 18), and he was followed by his brother Cotys (Tacitus, _l.c._). Their father was called Aspurgus (_C. I. Gr._ ii. p. 95), but need not on that account have been an Aspurgian (Strabo, xi. 2, 19, p. 415). Of a subsequent change of dynasty there is no mention; king Eupator in the time of Pius (Lucian, _Alex._ 57; _vita Pii_, 9) points to the same house. Probably, we may add, these later Bosporan kings, as well as the immediate successors of Polemon not even known to us by name, stood in relations of affinity to the Polemonids, as indeed the first Polemon himself had as his wife a granddaughter of Eupator. The Thracian royal names, such as Cotys and Rhascuporis, which are common in the Bosporan royal house, connect themselves doubtless with the son-in-law of Polemon, the Thracian king Cotys. The appellation Sauromates, which frequently occurs after the end of the first century, has doubtless arisen through intermarriage with Sarmatian princely houses, but, of course, does not prove that those who bore it were themselves Sarmatians. If Zosimus, i. 31, blames the petty and unworthy princes who attained to government after the extinction of the old royal family, for the fact that the Goths under Valerian could carry out their piratical expeditions in Bosporan ships, this may be correct, and in the first instance Pharnaces may be meant, of whom there are coins from the years 254 and 255. But even these, too, are marked with the image of the Roman emperor, and later there are again found the old family names (all the Bosporan kings are Tiberii Julii), and the old surnames, such as Sauromates and Rhascuporis. Taken as a whole, the old traditions as well as the Roman protectorate were still at that time here retained.]

[214: The last Bosporan coin is of the year 631, of the Achaemenid era, A.D. 335; this is certainly connected with the installation, which falls in this very year, of Hanniballianus, the nephew of Constantine I., as "king," although this kingdom embraced chiefly the east of Asia Minor and had as its capital Caesarea in Cappadocia. After this king and his kingdom had perished in the bloody catastrophe after Constantine's death, the Bosporus was placed directly under Constantinople.]

[215: The Bosporus was still in Roman possession in the year 366 (Ammianus, xxvi. 10, 6); soon afterwards the Greeks on the north shore of the Black Sea must have been left to themselves, until Justinian reoccupied the peninsula (Procopius, _Bell. Goth._ iv. 5). In the interval Panticapaeum perished under the assaults of the Huns.]

[216: The coins of the town Chersonesus from the imperial period have the legend Χερσονήσου ἐλευθέρας, once even βασιλευούσης, and neither name nor head of king or emperor (A. v. Sallet, _Zeitschrift für Num._ i. 27; iv. 273). The independence of the town evidences itself also in the fact that it coins in gold no less than the kings of the Bosporus. As the era of the town appears correctly fixed at the year 36 B.C. (_C. I. Gr._ n. 8621), in which freedom was conferred upon it presumably by Antonius, the gold coin of the "ruling city" dated from the year 109 was struck in 75 A.D.]

[217: According to Strabo's representation (xi. 2, 11, p. 495) the rulers of Tanais stand independently by the side of those of Panticapaeum, and the tribes to the south of the Don depend sometimes on the latter, sometimes on the former; when he adds that several of the Panticapaean princes ruled as far as Tanais, and particularly the last, Pharnaces, Asander, Polemon, this seems more exception than rule. In the inscription quoted in the next note the Tanaites stand among the subject stocks, and a series of Tanaitic inscriptions confirms this for the time from Marcus to Gordian; but the Ἕλληνες καὶ Ταναεῖται alongside of the ἄρχοντες Ταναειτῶν and of the frequently mentioned Ἑλληνάρχαι confirm the view that the town even then remained non-Greek.]

[218: In the only vivid narrative from the Bosporan history which we possess, that of Tacitus, _Ann._ xii. 15-31, concerning the two rival brothers, Mithradates and Cotys, the neighbouring tribes, the Dandaridae, Siracae, Aorsi, are under rulers of their own not legally dependent on the Roman prince of Panticapaeum.--As to titles, the older Panticapaean princes are wont to call themselves archons of the Bosporus, that is, of Panticapaeum, and of Theudosia, and kings of the Sindi and of all the Maitae and other non-Greek tribes. In like manner what is, so far as I know, the oldest among the royal inscriptions of the Roman epoch names Aspurgos, son of Asandrochos (Stephani, _Comptes rendus de la comm. pour 1866_, p. 128), as βασιλεύοντα παντὸς Βοοσπόρου, Θεοδοσίης καὶ Σίνδων καί Μαϊτῶν καί Τορετῶν Ψησῶν τε καὶ Ταναειτῶν, ὑποτάσαντα Σκύθας καὶ Ταύρους. No inference as to the extent of the territory may be drawn from the simplified title.--In the inscriptions of the later period there is found once under Trajan the doubtless adulatory title βασιλεὺς βασιλέων μέγας τοῦ παντὸς Βοοσπόρου (_C. I. Gr._ 2123). The coins generally, from Asander onward, know no title but βασιλεύς, while yet Pharnaces calls himself βασιλεὺς βασιλέων μέγας. Beyond doubt this was the effect of the Roman sovereignty, with which a vassal-prince placed over other princes was not very compatible.]

[219: This was the king Mithradates, installed by Claudius in the year 41, who some years afterwards was deposed and replaced by his brother Cotys; he lived afterwards in Rome, and perished in the confusions of the four-emperor-year (Plutarch, _Galba_, 13, 15). The state of the matter, however, is not clear either from the hints in Tacitus, _Ann._ xii. 15 (comp. Plin. _H. N._ vi. 5, 17), or from the report (confused by the interchange of the two, Mithradates of Bosporus, and Mithradates of Iberia) in Petrus Patricius _fr._ 3. The Chersonese tales in the late Constantinus Porphyrogenitus, _de adm. imp._ c. 53, do not, of course, come into account. The bad Bosporan king Sauromates, Κρισκωνόρου (not Ῥησκοπόρου) υἱός, who with the Sarmatians wages war against the emperors Diocletian and Constantius, as well as against the Chersonese faithful to the empire, has evidently arisen from a confusion of names between the Bosporan king and people; and just as historical as the variation on the history of David and Goliath, is the despatch of the mighty king of the Bosporans, Sauromates, by the small Chersonesite Pharnaces. The kings' names alone, _e.g._ besides those named, the Asander, who comes in after the extinction of the family of the Sauromatae, suffice. The civic privileges and the localities of the city, for the explanation of which these _mirabilia_ are invented, certainly deserve attention.]

[220: There are no Bosporan gold or pseudo-gold coins without the head of the Roman emperor, and this is always that of the ruler recognised by the Roman senate. That in the years 263 and 265, when in the empire elsewhere after the captivity of Valerian Gallienus was officially regarded as sole ruler, two heads here appear on the coins, is perhaps due only to want of information; yet the Bosporans may at that time have made another choice amid the many pretenders. The names are at this time not appended, and the effigies are not to be certainly distinguished.]

[221: This we may be allowed to believe at the hands of the Scythian Toxaris in the dialogue placed among those of Lucian (c. 44); for the rest he narrates not merely μύθοις ὅμοια, but a very myth, of whose kings Leucanor and Eubiotes the coins, as may well be conceived, have no knowledge.]

[222: As respects the export of grain, the notice in the report of Plautius (p. 218), deserves attention.]

[223: From the offer of a township of the Siracae (on the Sea of Azoff) hard pressed by the Roman troops to deliver 10,000 slaves (Tacitus, _Ann._ xii. 17), it may be allowable to infer a lively import of slaves from these regions.]