² CAH x, Preface, p. ix.

³ Cyprus is not mentioned, paceHondius, Saxa Loquuntur, p. 81, in the prospectus of IG; and Mr. Tod assures me that he knows nothing of an IG xv to be devoted to Cyprus. And yet Cypriot epigraphy of the Classical and Hellenistic periods has an interest which is more than local. Again, despite the enthusiasm which the deciphering of the Syllabary in 1871 excited, leading to the inclusion by Deecke of all the syllabic texts then known in SGDI i (1883), these inscriptions have during the last sixty years, save for the researches of R. Meister (interrupted however by his untimely death in 1912) received but casual attention. Indeed, it is quite characteristic that a recent and valuable work by Avi-Yonah, Abbreviations in Greek Inscriptions ( The Near East, 200 b.c. to A.D. 1100), though otherwise exhaustive, does not and perhaps cannot make any use of Cypriot material.

⁴ Cf. ‘The Roman Province,’; pp. 226–56, in Sir George Hill's History of Cyprus, a work which is already a classic. This chapter supersedes V. Chapot, Les Romains et Chypre, in Mélanges Cagnat, 1912, 59–83. For the bibliography, cf. Hill, o.c., 226, n. 1.

⁵ But Hogarth, o.c., 115, regretted that the ‘third and most important letter in the date was hopelessly lost.’ He adds, ‘If this was, as I believe, P, then reckoning from the establishment of the province we get A.D. 55 for the date of the inscription.’ There is no third letter in the date. Cf. further, note 16 below. In addition to Hogarth's version of the present inscription, three other inscriptions in Devia Cypria could with profit have been included in IGR. These are: p. 8, no. 1; p. 63, no. 15 and p. 109, no. 28 (below, note 118). Cagnat's neglect of Devia Cypria is the more surprising when we find in IGR iii, 936 a milestone originally published by Hogarth (D.C., p. 112, no. 34) cited from CIL (iii, 12111).

⁶ Notably in xi, which in the main text is Ξ, in the appendix

⁷ Whether the ‘circular edifice’ in which Cesnola states he found the inscription is the of the text, is open to question. Hogarth makes no mention of it. The Swedish excavators could discover no trace of this building, but the reason for their failure is not far to seek. In the Cyprus Museum files (CM. Files 36, pp. 24, 25) is a MS. report by the late G. Jeffery, the well-known architect, dated September 21, 1912, in which he describes the looting of the Soli site for stone by contract. The Public Works Department ‘also looted for stone to build a bridge.’ In 1910 one temple was ravaged ‘for the culverts of a road.’ Jeffery adds that ‘on the right of the road from Morphou there was still a dim trace of the circular temple with adytum or oracular cave beneath it, referred to by Cesnola.’ On p. 24 he gives a plan of the site, showing the ‘circular temple’ lying to the N. of the main road, opposite the harbour and below the theatre. If this ‘circular temple’ is the the ‘adytum or oracular cave beneath it’ is probably the family tomb of Apollonios.

⁸ What is essential is, not the but the object it encloses. This, as the privacy which the implies would itself suggest, is regularly a tomb: MAMA iv, no. 85: MAMA vi, no. 358: SEG vi, 1932) no. 676:— is here preferred, though new to Cyprus, since it can denote the vaulted, rockcut tomb that is characteristic of Cypriot burial.

That a man should construct a tomb for his parents and then specify that it is a family tomb, for the use of himself and his children, is a proceeding too common to require comment. What is remarkable in the present case is the employment of the second person in which I understand, with Hogarth, to refer to the parents.

⁹ On the civic police of the Roman period, L. Robert, Études anatoliennes, 96–110. Further, A. H. M. Jones, The Greek City, 212–13. The originally military, survived under the Empire in various Asiatic cities; where no doubt, as at Smyrna, it sank to the captaincy of the city police (Jones). At Cyzicus, however, throughout much of the history of that city the supplied the eponym of the year (Hasluck, F. W., Cyzicus (1910), 304–5).Google Scholar The Cypriot instances are: JHS xii (1891), 332 (from Larnaca, now lost) and two unpublished inscriptions of similar provenance. In one of these (where, however, the reading is conjectural) the may occupy a position of remarkable prominence.

¹⁰ For used without further qualification of the imperial cult, cf. IGR iii, 981 from Citium: (where is clearly to be supplied). Possibly also LB W 2759 from Salamis: That individual cities also had such High Priests is demonstrated for Lapethus by IGR iii, 933; for Salamis, by ib. 961 (not Palaepaphus as IGR, the provenance being the Kouklia in the Mesoria); for Citium, by Trans. Soc. Bibl. Arch, iv (1876), 42 (where I restore

¹¹ Not, pace Hill, o.c., 240 a ‘public library.’ While occurs freely in the sense of record office (e.g., Sammelbuch gr. Urkunden aus Aegypten iv, 7404), is nearly always so used (for refs. SB ii, P. 335). A priori, it is highly unlikely that a city so obscure as Soli should boast a public library, even as an adjunct to its gymnasium; cf. Poland, Öffentliche Bibliotheken in Griechenland u. Kleinasien.

¹² Cf. the interesting remarks of A. H. M. Jones, The Greek City, 171, on municipal censors in the Eastern provinces. Jones, however, argues that the Romanisation of the local councils in Cyprus—a deduction from the presence of in the present inscription (which he wrongly ascribes to Citium)—is the handiwork of Cato; who for his model took the constitutional changes which Pompey had instituted in Bithynia. This argument, which would be the better for a little evidence, does not take into account the fact that Cyprus under the Republic was not a province but an annexe of Cilicia. Cato's activities, from all we hear of them, were rather financial than constitutional; and for any con stitutional changes in the status of the Cypriot I would prefer to look to the lost edict of the un known governor of Cilicia in 58 b.c. Furthermore, from 48 to 30 b.c. Cyprus was restored to a Ptolemaic rule which, from the existence of coinage struck by Cleopatra for the island (Hill, o.c., 210), seems to have been effective enough. Finally, Augustus sent Paquius Scaeva to Cyprus for a second and extraordinary proconsulship expressly ad componendum statum in reliquum provinciae Cypri (CIL ix, 2845); and whatever else this phrase may mean, it can presumably cover such changes as are here in question.

¹³ Plut. Pomp. 13: Ann. Epigr., 1905, 120: = adlectum inter aedilicios.

¹⁴ Acts xiii, 4–13. Hogarth stoutly supports the identification; but no other authority is prepared to commit himself.

¹⁵ These are the extreme limits in time for St. Paul's celebrated encounter at Paphos. Hill, o.c., 247 sets the landing of the Apostle at Salamis in a.d. 45; the proconsulship of Sergius Paulus ‘about 46–48’ (o.c., 255, no. 12).

¹⁶ In a forthcoming article I show that the only system of dating, in the proper sense of this term, current in Cyprus between 30 b.c. and the reforms of Diocletian, was regnal. One exception only is known to me, the ‘new sacred year,’ a short-lived era, attested by coins for the years 76–7 to 78–9 (Hill, o.c., 234–5). Cf. Dittenberger, OGIS 582, n. 2; Hill, o.e., 235.

¹⁷ In datable inscriptions of the Principate, the status of iota adscript is as follows:

Thus in the first century iota adscript was freely but capriciously used. There is a tendency for it to become rarer as the century advances; but it shows a certain persistence with religious formulae (such as the dedications to Aphrodite of Paphos).

For the second century, when it is becoming an anachronism, I note 5 instances: under Trajan; under Hadrian or the Antonines (twice, but once with only); under Pius (with Aphrodite, but omitted with the titles of the emperor which follow) and under Marcus. In the third, it is an archaism, found once in the reign of Alexander Severas (?).

¹⁸ At no period during the first two centuries was the civitas common with native Cypriots outside Paphos (notably among the priestly families), Citium (magistrates and priests of the emperor-cult) and Salamis. At Soli under Marcus are not cives (IGR iii, 929). At Lapethus, neither Adrastos, son of Adrastos, who dedicated to Tiberius a statue and a shrine (IGR iii, 933) nor his grandson (or great grandson) under Trajan (unpublished) were Roman citizens. So, too, with Sosos, son of Sosos, of Salamis (CIG 2639; from its lettering, of the second century) and Sodamos, son of Sodamos, of Citium (LBW 2735), although both men had distinguished municipal careers.

¹⁹ For this alphabet, cf. Bradford Welles, Garasa, City of the Decapolis: Inscriptions, pp. 358–60. It was in common use at Jerash in the first century, is apparently of Syrian origin, and is found in its purest form at Dura-Europos and at Susa. In Cyprus its status is uncertain. Taking sigma formed from three sides of a rectangle as the essential characteristic (although this form occurs once in what is apparently a late Ptolemaic context), I find some seven datable examples. These range from the last years of Augustus (1) and the reign of Gaius (1) to Hadrian or the Antonines (2), Pius (1) and Septimius Severas (2). Of these seven, however, only two have three or more square letters, and both are Septimian. The present inscription, an excellent example of this alphabet, outdoes all the above in having a square inverted omega; while omicron is transitional, being in certain cases distinctly angular. For the lunar and classical alphabets, cf. note 85 below.

²⁰ In the early imperial inscriptions of the E. never appear. IGR iii, 1376 from Gerasa, now dated to A.D. 66 (McCown, , TAPA 64 (1933), 79) supplies the earliest instanceGoogle Scholar; though it may be doubted whether here bears any further significance than Josephus’ (apparently honorific) of Tiberias (Vita 13, 68 fF.). For the first occurrences in a true Greek environment, we must turn to Lycia during and after the reign of Hadrian; while the bulk of the epigraphic testimony belongs to the end of the second and to the third centuries. In origin the are generally admitted to have been a finance committee of the local council; but their chief function soon became the collection of taxes for the central government. The definitive study is that of Turner, E. G., JEA xxii (1936), 7–19Google Scholar, Cf. further Rostovtzeff, , Social and Economic History of the Roman Empire, 358, 594⁴⁵, 601¹⁹.Google Scholar Other Cypriot instances are: CIG 2639 from Salamis (from its lettering, of the second century) and an unpublished inscription of unknown provenance, of which the relatively late date is apparently betrayed by its use of ligatures.

²¹ IGR iii, 935 (Cesnola, o.c., 417, no. 11) = Myres, , Handbook of the Cesnola Collection (1914), 319Google Scholar, no. 1903 and p. 548. Myres* version, which consigns the IGR text to an epigraphic limbo, for the last three lines reads: It may be noted that of the emperor's name enough survives to make it certain that we are here concerned with one of the Julio-Claudians; while at the end the restored cognomen followed by is at least plausible.

The corrected version of IGR iii, 970 from Curium is also given by the Handbook (p. 548, no. 1904). Of this the first line should read:

²² The Sergii Pauli retained their nobilitas late into the second century, as is demonstrated by the Sergius Paulus, L. (PIR¹ iii, p. 221Google Scholar, no. 377) who under Marcus was proconsul of Asia. That the cognomen Paulus, however, was common to various gentes of senatorial rank, such as the Aemilii, Servilii, Statii, Vettii, etc., hardly calls for mention. In passing, it may be noted that it is far from certain that the Sergius Paulus of the Acts bore the praenomen Lucius. One L. Sergius Paulus was under Claudius curator alvei Tiberis and therefore an ex-consul (Dessau, 5926). That he is to be identified with the proconsul of Cyprus is at best a conjecture, now somewhat weakened by Myres' reading of IGR iii, 935.

²³ Myres, Handbook, pp. 322, 550, nos. 1914 and 1915—found but not published by Cesnola. During a brief visit to the Kastro site, Cesnola claims (Cyprus, p. 285) to have discovered on the summit the ‘ruins of an elliptical structure measuring 27 feet by 16.’ To-day of this only a heap of boulders is to be seen, though below it the slopes are littered with broken statues of life size, large stone basins, stone eagles, etc. For the possible equation of Zeus Labranios with the Ba'al of Lebanon of the eighth-century bronze bowls of Mouti Sinoas (some 7 miles to the E. of Phassoula) cf. Hill, o.c., 107; but it may be noted that the epigraphic evidence for Zeus Labranios in Cyprus is thoroughly late.

²⁴ Plutarch, Mor., 292 A.

²⁵ Hogarth, o.c., 117; Kadlec, in RE viii, 2469, no. 9 (1913).

²⁶ Dessau, , PIR ii, p. 148Google Scholar, no. 149; Cagnat on IGR iii, 944; Hill, o.c., 255, no. 18.

²⁷ So Dessau, I.c.

²⁸ So Kadlec, I.c.

²⁹ Of two inscriptions to Domitian which survive in Cyprus, one (unpublished) is a graffito which has been saved from erasure by its ephemeral character; the other, IGR iii, 945, an altar from Palaepaphus, on which Domitian's name can only be read with such difficulty that it has probably suffered deliberate damage. For this second inscription my squeeze gives the following:

³⁰ Cf.IGR iv, 1393.

³¹ An inscription from Lapethus (IGR iii, 933) tells of a shrine and statue, dedicated to Tiberius, being consecrated by the municipal high priest of the imperial cult. Later, with imperial statues, the consecration was regularly performed by the local representative of the emperor; but, save for the present instance, the practice finds no mention on Cypriot statue-bases before the Septimian era; to which belong SEG vi, nos. 810 and 811 (below, no. 4), and IGR iii, 977 (Julia Domna). Inscriptions under the statues of Nerva (IGR iii, 976—I examined this inscription at Munich in 1936 and found and not to be the correct reading), Trajan, (IGR iii, 987Google Scholar and an unpublished text from Lapethus), Hadrian (IGR iii, 934) and Marcus (IGR iii, 929) have nothing to say of consecration. With shrines, etc., the usage was apparently different; here, a verb was freely included, and we accordingly find consecration by proconsuls under Titus of a (unpublished), under Claudius (IGR iii, 971) of a building (altar?) of unspecified character. For IGR iii, 971, though cut on a statue-base which carries a Hellenistic inscription (OGI 152), from the use of the dative case and the fact that it is inscribed on a narrow face of the stone, can hardly be concerned with a statue. IGR iii, 971 has not been examined since the original copy by Vidua's correspondent; but (though in the interval it has perished considerably) may still be seen in the yard of the police station at Episkopi. The following text is based on a squeeze: The letters underscored are from Vidua's text. I see no good reason for ascribing this inscription to Nero rather than Claudius (Ritterling in RE xii, 1701, followed by PIR ² i, p. 108, no. 637 and Hill, o.c., 255); the margin between line 2 and the top of the stone is too narrow (0·05 to 0·055 m.) to admit of more than one line (the letters averaging 0·03 m. in height) Nor is it disputed that the of IGR iii, 974 from Amathus refers unambiguously to Claudius.

³² Tacitus, Ann. ii, 37 and 38; Suetonius, Tib. 47.

³³ IGR iii, 942. That it is to be ascribed to Tiberius‘ accession is an inference from the absence of date and motive formula.

³⁴ IGR iii, 941. Tiberius is here described as the city's saviour and benefactor. In 1. 6, before the name of the month, a regnal date is faintly legible: this I shall discuss in a forthcoming article on the chronology of Roman Cyprus.

³⁵ Cf. Hill, o.c., 245. A very serious disaster in A.D. 76 or 77, when three or more cities were de stroyed. Two of these, if the Sibylline Oracles may be trusted, were Salamis and Paphos (Hill).

³⁶ The rule is first broken by Pupienus and Balbinus: cf. Cagnat, Épigr. lat. ⁴, p. 161.

³⁷ The initial letter of 1.4 is most difficult. The top of omicron in the original inscription is visible; and this, it would seem, has been re-used: rho, with eta and iota as alternatives, are offered in the order of their probability. The second letter, however, is an undoubted kappa.

³⁸ The first two letters of 1.5 are most uncertain. At the beginning are traces of the Hellenistic epsilon, which does not seem to have been recut; but tau or gamma is not impossible here. The second letter has been mostly flaked away; but over an earlier upsilon, whatever else may have been cut, a vertical stroke immediately before sigma is certain. This is most probably part of eta; nu however cannot be excluded, while iota preceded by either alpha or lambda is a possibility. For the name as a whole, cf. Dessau, 976:

³⁹ For Hortensia, , daughter of the orator, RE viii (1913), 2481, no. 16.Google Scholar It is thought that she married Q. Servilius Caepio, father by adoption of M. Brutus; but of any issue to such a marriage nothing is known.

⁴⁰ From this it may fairly be deduced that dedications made by Paphos, in which that city bears no title, are earlier than 15 B.c. This will affect the dating of IGR iii, 952, 953 and 964, of which the first two already have 23 B.C. as their probable terminus post quern. Nor can the lost fragment from Paphos, CIG 2628, be referred (pace Cagnat IGR iii, 938) to Herodes Atticus: it is safer to ascribe it to Herod the Great, who may have had interests in Cyprus before acquiring in 12 B.C. his contract for the copper mines of the island.

⁴¹ Below, no. II.

⁴² IGR iii, 937 (of which the last two lines should read: [δι]ὰ τῶν τὸ ἔκτον [ἔτος ἀρξάν]∣ των(?)); SEG vi, nos. 810 and 811; and four milestones (IGR iii, 967; JRS xxix (1939), 193, no. 5 and p. 194, no. 6; and one as yet unpublished).

⁴³ It is not known why or when Paphos received the title Claudia. It has been sensibly suggested by Hill (o.c., 233) that Flavia was a return for the favourable reception which Titus was accorded at Old Paphos in A.D. 69.

⁴⁴ IGR iii, 962.

⁴⁵ But this alteration in the order of titles is understandable. A title derived from a reigning emperor or dynasty heads the list, which is then given in reverse (cf. IGR iii, 879 for the titulary of Tarsus). Paphos, by the Septimian era, had for over a century won no new title; and her titulary is accordingly presented in its proper chronological sequence.

⁴⁶ A rare alphabet at Old Paphos, the only dated example being SEG vi, no. 811 (sigma only). Square sigma is also found in JHS ix, 245, no. 77 (undated), and, according to edd., in IGR iii, 943 (No. 12 below).

⁴⁷ IGR iii, 989 (JHS xii (1891), 180, no. 15; IBM iv (1916), no. 983). The last two lines of this inscription should read: My squeeze shows that the lacuna in 1.5 is some twenty-one to twenty-three letters in length, while there is no justification for the ungrammatical hitherto accepted. It will be noted that though a restoration, hardly admits of an alternative. In the third year of Trajan Salamis still describes herself simply as (IGR iii, 987). How long she may have kept the title there is insufficient evidence to determine; but an inscription, ascribed chiefly on its lettering to the Antonine period (CIG 2639) speaks only of For a plurality of in one province (apart from the case of Egypt), Cilicia is instructive. Here, Tarsus became under Augustus (Strabo, xiv, 5, 13); but when at the beginning of the third century Anazarbus arrogated the title, Tarsus could only retaliate by styling herself (IGR iii, 879).

⁴⁸ BCH li (1927), 139, no. 3.

⁴⁹ Cagnat, Épigr. lat. ⁴, 209.

⁵⁰ For the curator civitatis and a brief but admirable history of his office, A. H. M. Jones, The Greek City, 136–8. A list of these curatores is given by Tod, M. N. in JHS xlii (1922), 172 ff.Google Scholar

⁵¹ PIR ² ii, p. 35, no. 181; p. 36, nos. 182 and 183. Seyng's suggestion (BCH li (1927), 141) that the Caesernius of our inscription is T. Caesernius Statius Quinctius Macedo Quinctianus (PIR ii, p. 36, no. 182) is untenable (as is his proposed restoration of CIL v, 865), because the latter was praetor under Hadrian, probably before the year A.D. 132.

⁵² IGR iii, 6; PIR ² ii, p. 35, no. 179.

⁵³ Campanius Marcellus, M. was proc. provin. Cypri (Dessau, 1398)Google Scholar not earlier than the reign of Hadrian. An inscription from the Phrygian Heraclia (MAMA vi (1939), no. 97), concerning a native of that city, T. Statilius Apollinarius, shows by a convincing restoration that From Apollinarius' career it may be deduced that he held this office under Pius.

⁵⁴ Keyes, C. W., The Rise of the Equites in the Third Century of the Roman Empire (1915), 3 ff.Google Scholar; Rostovtzeff, Social and Economic History of the Roman Empire, 354.

⁵⁵ Such Caesernius Statianus clearly was by birth, while a city so important as Nicomedia can hardly, even at this date, have had any but a senatorial curator. At the beginning of the third century, however, distinguished provincials and equites had in general long broken down the senatorial monopoly to the office; and by the middle of that century the curator was regularly a citizen of the city he administers.

⁵⁶ An analysis of the inscriptions of Kouklia, Hellenistic and Roman, shows some 118 statues, eighteen stelae and tablets, seven altars, etc., a herm and a few such miscellaneous objects as an inscribed tortoise, a dish, a ring. Of porticoes, shrines and buildings generally there is no record whatsoever. I have little doubt that the rarity of inscriptions for the second century A.D. is due solely to the overcrowding of that portion of the general temple complex that has been excavated.

⁵⁷ Cf. Avi-Yonah, Abbreviations in Greek Inscriptions, 37. His earliest instance of this sign, in its upright form and written after the word abbreviated, is from the middle of the second century. In Cyprus it occurs from the Antonine period (BMI iv, 986, dated by its lettering) to the Byzantine (when it is common), the latest instance, so far as I am aware, being from the reign of Heraclius (LBW 2763).

⁵⁸ At the end of 1.1 there appears a lower apex, indicating a vertical stroke. This, unless it be accidental, is decisive against omicron. At the end of 1. 2 alpha is definitely preferable to iota. The remainder of the text is without difficulty.

⁵⁹ Edd. describe the lost JHS ix, no. 97 as cut on a ‘white,’ JHS ix, no. 114 on a ‘pink’ stone. Further, they give for the former letter-forms (such as alpha with the bar broken; and cursive alpha, mu and nu) not shown in JHS ix, no. 114. JHS ix, no. 114 is in fact inscribed on a greyish-white marble; while JHS letter-forms are generally untrustworthy.

⁶⁰ IGR iii, 978, now immured into the Church of the Soteira at the Metropolis of Larnaca; it is not easy to see why should have been omitted by Vidua's source or by Angelati, unless indeed it was then mortared over. From Salamis, IGR iii, 992.

⁶¹ Cited by Waddington, Fastes des provinces asiatiques, under Asinius Gallus (no. 58).

⁶² IGR iii, 875.

⁶³ CIL v, 4319 (Dessau, 579) from Brixia; CIL iii, Suppl., 12736 (Dalmatia); CIL iii, Suppl., 7586, from Callatis. There is a possible fourth instance from Smyrna, CIL iii, 472. Homo, L., Essai sur le règne de l'empereur Aurélien (1904), 326Google Scholar, can offer no explanation.

⁶⁴ The 15th milestone from Paphos on the road to Curium, as yet unpublished. Homo notes, I.c., that in a second inscription from Brixia, CIL v, 4320, Aurelian's name shows no sign of damage; and, it may be added, the Moesian and Dalmatian instances are equally isolated. The explanation is doubtless to be found in local and uncoordinated malice.

⁶⁵ CIL iii, Suppl., 7586; CIL xi, 2635; CIL v, 4319; CIL vi, 1112.

⁶⁶ The only official title taken from this conquest; Arabicus and Palmyrenicus, since they refer to peoples legally within the Empire, being improperly used. For Parthicus, Persicus is once substituted (CIL xii, 5561), perhaps in error. Formidable problems are raised by every attempt to correlate Aurelian's titulary with our scanty knowledge of his actual campaigns. Here I can only refer to the reconstruction of Homo, o.c.; and for a more recent treatment, Mattingly in CAH xii (1939), 297–311.

⁶⁷ Homo, o.c., 340; Cagnat, Épigr. lat. ⁴, 227. Here too, the monuments show great irregularities, as yet unexplained. As an example, coins give a seventh (sic) tribunician count, coupled with the second consulship which falls in the year A.D. 274. So also the inscription, Dessau, 581.

⁶⁸ CIL viii, 9040; Orelli-Henzen, 5551 (Germanico Gothico maximo).

⁶⁹ Orelli-Henzen, 5551.

⁷⁰ CIL vi, 1112 (from Rome), dated to the end of 274. Aurelian is given in the text four titles derived from victories; and yet is described as imp. III (and of these one must be the initial acclamation on his assuming the purple). In CIL xii, 5548, of the pre ceding year, Imp. Ill is likewise found. Further, Homo, o.c., 141.³.

⁷¹ Cagnat, Épigr. lot. ⁴, 232. But this view is opposed by Seeck, O. in Rh.M xlviii, 196 ff.Google Scholar

⁷² While the dates of Aurelian's three consulships are not open to question, irregularities again occur in their presentation on the monuments: Cagnat, o.c., p. 227⁵. The proconsulship is almost invariably included by Septimius and his successors, whenever a full titulary is given. I note some nineteen instances with Aurelian, usually as the final title, immediately preceded by pater patriac.

⁷³ Dimensions: 1.0·61; h. 0·24 m. There is no trace of the first two lines. At the beginning of 1.3 are clearly visible the markings taken by edd. as ΑΤ (better ΔΤ). The stone here has been badly scratched; but I think it probable that these markings conceal an original Σι. Thereafter, a lacuna of c. 3 letters. The remainder of the line can be read with certainty.

⁷⁴ In the Acts of the Arval Brotherhood for A.D. 59 (Dessau, 229⁴¹), under the rubric January 3, Nero is styled cos. III, design. IIII. This is disputed by Constans, CRAcIns. 1912, 388, who points out that 1. 67 of the same inscription records the offering of sacrifices in March ob comitia consularia. But, as Dessau observes (note 20 on p. 61), the comitia in question were probably those of the year 51.

⁷⁵ Towards the end of 58: Tacitus, Ann. xiii, 41. So Stuart-Jones, , RA iii (1904), 266Google Scholar, followed by Anderson in CAH x, 762.

⁷⁶ Nero was imp. VII by January 1, 60 (Acts of the Arval Brotherhood for the year 60: CIL vi, 2042 d, 17). Tigranocerta fell before the close of the campaigning season of 59; the news would reach Rome in the early autumn: Stuart-Jones, I.c., 269; Anderson, I.c., 764².

⁷⁷ Mattingly, , JRS xx (1930), 79 ff.Google Scholar (after Stobbe, , Philologus xxxii (1873), 23)Google Scholar; more recently, Momi gliano, CAH x, 702.

⁷⁸ Mommsen, , Hermes ii (1867), 56Google Scholar; Cagnat, Épigr. lat. ⁴, 186; Sandys, Latin Epigraphy ¹, 237.

⁷⁹ Stuart-Jones, l.c., 269.

⁸⁰ Tacitus, Ann. xiv, 25.

⁸¹ Noted, however, by Constans, l.c., 385^s, as ‘incomplète, attribuée à Néron.’

⁸² V. Chapot, Mélanges Cagnat (1912), p. 80, rightly objects to the conjunction of with but can offer no remedy. with its variants and simply are, it would seem, properly used of individuals. Cities prefer a more stately phrase: I note two exceptions, with Salamis, (IGR iii, 992)Google Scholar and with the Κοινόν of Cyprus, (IGR iii, 993).Google Scholar In this connection, minor corrections may be made in two published inscriptions, both from Citium: LBW 2735, lines 6 and IGR iii, 982, lines 5 and 6 (for )

⁸³ Cf. IGR iii, 933.

⁸⁴ Note 31 above.

⁸⁵ The inscription is a remarkable example of the lunar alphabet, the more striking when contrasted with its contemporary, IGR iii, 985. Save for iota and the uprights in tau and pi, there is hardly a straight line in it, even the upper strokes in zeta and tau being up-curved. Datable specimens of this alphabet— none, however, so extreme, the normal criteria being cursive epsilon, mu, sigma and omega— show for the first century five instances: Tiberius, Nero, Domitian (2) and Nerva. Second century, eleven instances: Trajan, Hadrian, Marcus (2), Commodus and Septimius Severus (6). Third century, three instances: Macrinus and Aurelian (2).

From the middle of the second century to the Arab invasion, the lunar alphabet is in almost exclusive use, the square being virtually its only rival. Certain inscriptions combine lunar and square forms (e.g., IGR iii, 934) and lunar and classical (e.g., IGR iii, 950). Some again use lunar and square forms in different parts of one text (SEG vi, no. 811 ). The Septimian milestones from the W. of Cyprus show one good example of square lettering, while the remainder are lunar.

The classical alphabet, with sigma and omega as its criteria, is thoroughly characteristic of the first century, though almost invariably it uses the square-topped rho (which is foreign to the Hellenistic period). Rare in the second century after the reign of Trajan, from the third there is an isolated example of the classical sigma provided by an archaising inscription (IGR iii, 958), and an unpublished instance from the fourth, For the second century I note three occurrences under Trajan, one under Hadrian, one under Marcus, and, if the various edd. are to be trusted, there is a further instance under Marcus (IGR iii, 946) and another under Septimius (IGR iii, 937).

⁸⁶ CIL vi, 2042 d.

⁸⁷ Hermes ii (1867), 56.

⁸⁸ Droit public romain v, 62².

⁸⁹ Philologus xxxii (1873), 23–30.

⁹⁰ It may be noted that Stobbe confines himself to pointing out that the consules suffecti of the diploma may belong, not to 60, but to 61 A.D. Mommsen's answer is to be found in Droit public romain v, 62^a. However, the inscription Dessau, 8902 would appear to prove that by 63 Nero had returned to the original system of tribunician reckoning; supported as it is by the numismatic evidence cited by Mattingly. And, if we are not to give both the Arval record and the diploma the lie, some such explanation as that of Constans may prove to be necessary. Constans suggested, l.c. 391, that the comet recorded during the year 60 (Tacitus, Ann. xiv, 22), with its portent of a new reign, terrified the emperor into a superstitious tampering with the tribunician count to avert the omen. This suggestion would be attractive, could it be shown that the comet was first seen towards the end of 59.

⁹¹ Dessau, 231; Eph. Epigr. viii, 365.

⁹² Rostovtzeff, , Rev. Num. ii (1898), 82 and 100Google Scholar: a tessera stamped with the name Paulinus; cited by Stuart-Jones, l.c., p. 270.

⁹³ So Collingwood, in CAH x, 802, who thus accepts the chronology of Tacitus (Ann. xiv, 31 ff.). The argument for placing the revolt of the Iceni in the year 60 rather than 61 is well set forth by Henderson, Life and Principate of the Emperor Nero (1903), 477.

⁹⁴ The dative implies dedication to a god; by extension, to a deified mortal; and so, freely to emperors. There is, however, a Cypriot example of dedication to a proconsul (Hogarth, Devia Cypria, 8, no. 1) and to Marcia, wife of Paulus Fabius Maximus (IGR iii, 939). The object dedicated can naturally vary from a precious article to an altar or a temple. In the present case, while I find it incredible that a shrine or some similar structure dedicated to the Divus Augustus should be rededicated to Tiberius, I do not feel the same difficulty with an altar in some gymnasium.

⁹⁵ CIG and IGR:

⁹⁶ regularly follows the noun it qualifies; so in six further occurrences of the phrase in Cyprus. But cf. IGR iv, 257, and passim. Of the Cypriot instances it may be noted that three concern priests (Hellenistic), two high priests (Roman), one a (Hellenistic). It is perhaps remarkable that our high priest should be alive some twenty-five years after the original dedication; but such a priesthood, being acquired by purchase, is an indication of the priest's wealth and pro-Roman sentiment and not of his age.

⁹⁷ Tacitus, , Ann. iii, 62–3.Google Scholar

⁹⁸ It is presumed that the Donations of Antony to Cleopatra were officially annulled on her death in August 30. But Cyprus in effect must have passed from the hands of Egypt, if not in the months immediately following Actium, certainly by the time of Augustus' advance through Syria upon Egypt in the early summer of 30. It is not known what provision was made for Cyprus before 27, when the island became an imperial province. Cf. also note 16 above.

⁹⁹ Both Pompey (Valerius Maximus i, 5, 6) and Titus (Tacitus, , Hist, ii, 2–4)Google Scholar called at Paphos on a precisely similar voyage; and Gaius, unlike them, was admittedly on a tour of inspection. A fragmentary inscription from Soli (not included, I think, in CIL) is restored by its editor (Tubbs, , JHS xi (1890), 75, no. 22)Google Scholar to Gaius: [CAI]O. CAESARI. D. A[VG. F]| - - \SOLIORVM.

¹⁰⁰ In all Cypriot inscriptions which refer to Augustus during his lifetime, he is a god: IGR iii, 932, 939, 940, 973, 984 (below, note 115) 994; REG xvii (1904), 212. Not so with Tiberius or any of his successors. Cyprus, which down to Actium had for centuries worshipped ‘the King, Ptolemy, the God,’ took easily to the Divus Augustus. The difference between a princeps and their former rulers (which may not have been immediately obvious to the Cypriot cities) was a lesson to be learnt. If is to be retained in the altered text, however, it may be paralleled by IGR i, 659; iv, 257.

¹⁰¹ No account was taken of his co-regency with Augustus in computing his regnal years: IGR iii, 933.

¹⁰² There is no trace of the of edd.; which, as they show it, spoils the symmetry of the inscription.

¹⁰³ The inscription is very faint, but, apart from this remarkable name, without difficulty. Here the spacing makes it clear that no letter is missing before the initial iota; the second letter is damaged, so that it may be either eta or sigma; the fifth and sixth letters, as there is no trace of a cross-stroke, are both lambda rather than alpha; the seventh is an apparently intact nu, but a horizontal stroke, which appears above it but is probably casual, makes it impossible to rule out pi; the eighth, apparently alpha, but vague upright lines make eta a possibility. I can find nothing to parallel the name Isoullna, which Prof. Rose suggests to me may be a pet-name derived from Isis. Mr. Tod writes, however, ‘ seems to me much more likely, whether derived from Isis or from the Latin insula.’ For not recognised by edd. as a name, cf. MAMA iv, no. 358, For in epigraphy, SEG vi (1932), no. 666.

¹⁰⁴ But for the names and the square-topped rho (note 85 above), JHS ix, 259, no. 3 might be Hellenistic. In the second inscription the middle stroke in epsilon is not contiguous—a sure sign of later date.

¹⁰⁵ BCH li (1927), 143, no. 4, now in the police station at Ktima. I believe this inscription to be Republican. My reasons for so thinking I shall give elsewhere.

¹⁰⁶ An Italian trading community was established at Alexandria as early as the reign of Euergetes II (OGI, 133, 135). As for Cyprus, the bilingual inscription, JHS ix (1888), 234, no. 28 (IGR iii, 965; CIL iii, 12101; Dessau, 7208) from Palaepaphus, may well be late Hellenistic. The Greek lettering seems to me definitely of that period; while the fact that the stone has been re-used—it also carries JHS ix (1888), 234, no. 25 of (it would seem) the early second century—is significant. I have noted sixteen instances of a Roman and a Hellenistic inscription cut on the same stone, but none where both are Roman. These figures do not include several cases where the earlier text has been too thoroughly defaced to be legible. Edd. find ‘a considerable space between the Latin and the Greek’; but this in fact is no more than the normal spacing of the lines. While they consider the Greek ‘regular and well cut,’ the Latin ‘has a later appearance.’ I am inclined to think that the Latin and the Greek are contemporaneous and definitely Ptolemaic rather than Republican. My squeeze gives me:

¹⁰⁷ OGI 164, 165.

¹⁰⁸ Edd.: 1. 2, at end: ΔΙ; 1. 3: Between the words of 1. 2 they show uncut spaces of two or three letters length, but 1. 3 with no such gaps. If they are correct in this and the inscription consistent, there may have been sufficient room in 1. 3 for the inclusion of

¹⁰⁹ For the omission of the dedicator in short inscriptions, JHS ix (1888) 232, no. 18; JHS ix (1888), 249, no. 99 (Archiv. f. Pap. xiii, 1938, 31⁴); in longer inscriptions, OGI 174; IGR iii, 940; possibly IGR iii, 984 (note 115 below). All are inscriptions on statue bases.

¹¹⁰ Under her statue at Salamis, Livia is simply Cf. note 115 below. The bulk of the datable inscriptions in which Livia is deified are later than A.D. 14.

¹¹¹ IGR iv, 257. In the E. she was often worshipped under the name of the local goddess (Charlesworth in CAH x, 611).

¹¹² Since Livia became Augusta officially only after the death of Augustus.

^112a But IGR has

¹¹³ Cagnat, by the inclusion of this fragment in IGR, presumably considers it to be on a ‘high level.’ The presence of in 1. 2 seems to me to make this certain; for I do not see how this cognomen could find its way to a recent recipient of the civitas. For the omission of the dedicator, cf. note 109 above; but the inscription as here restored may very well be incomplete.

¹¹⁴ No. 11 above; IGR iii, 939 and 940.

¹¹⁵ IGR iii, 984, now lost. JHS xii, 176, no. 5 shows that there is sufficient space in the 4th line to restore, quite normally, Cf. note 100 on p. 225.

¹¹⁶ REG xvii (1904), 212 = SEG vi (1932), no. 837. To the above may be added IGR iii, 973 from Amathus, now lost (perhaps an early milestone) and JHS xi (1890), 75, no. 22 from Soli. A High Priest of Augustus for all Cyprus is known from IGR iii, 994 from Salamis.

¹¹⁷ Excluding milestones and those inscriptions where the emperor's name is used solely as a date, the figures are: Tiberius, 4; Gaius, 1 (unpub.); Claudius, 1; Nero, 2; Titus, 2 (one unpub.); Domitian, 1; Nerva, 2; Trajan, 4 (two unpub.); Hadrian, 2; Pius, 1 (unpub.); Marcus and Commodus, 2; Septimius Severus and Julia Domna, 3 (one unpub.). Add to this, a month Tiberieios at Paphos and a mention of Domitian's birthday festival (unpub.) and the evidence is complete.

¹¹⁸ Devia Cypria, 109, no. 28: [ἡ πὁλις ή Κιτιἑων(?) τὸν δεῖνα τοῦ δεῖνος ] ἁρΧερἑα τῆς ῾Ρὡ[μης ] τὸν ἀγωνοθέτη[ν τῆι] πινταετηρ[δι κα[τὰ] τὸ ή L ἐφ᾿ σῦ πρῶ[τον] τὀ ἄλιμμα (sic) ἐτέθ[η] We need waste no time on Hogarth's explanation of the date (‘the eighth period of five years since the games were estab lished’), since the inscription belongs to an eighth regnal year, if we may trust Hogarth's lettering, rather in the earlier than the later half of the first century. These quinquennial games, like those of Salamis, were known outside Cyprus, on the evidence of IGR iii, 1012 (dated to A.D. 221). The Actaean games of Lapethus have already been noted: REG xvii (1904), 212 = SEG vi (1932), no. 837. Cf. further, Cesnola, Cyprus, 431, no. 44 = Myres, Handbook of the Cesnola Collection, 320, no. 1905 and p. 548, for a late agonistic inscription from Citium, the only one of this type which Cyprus can show. Hogarth's inscription has a further point of interest. Domaszewski, Abhand. z. röm. Religion, 234 ff., comments upon the non-inclusion of Roma in the emperor-worship of Cyprus. Here is his answer. Cf. also Trans. Soc. Bibl. Arch, iv (1876), p. 42 as restored in note 10 above.

¹¹⁹ L. 2: after final sigma, an uncut space; but whether the line is complete or not, cannot be decided. The phrase is found in an unpublished inscription from Larnaca. For the form of the present inscription, cf. the fragment LBW 2804 which may still be seen at Kouklia: and which, save to the left, seems to me complete.

¹²⁰ For this name: a variant of (Tod), for which cf. CIG Index, p. 71.

¹²¹ unqualified at Palaepaphus is to be referred, not to the imperial cult (note 10 above), but to the local worship of Aphrodite. And, with few exceptions, during the Roman period the word (and ) at Palaepaphus is always unqualified. Whether Claudia Rhodokleia, however, was High Priestess of Aphrodite is a very different matter, on which I suspend judgement: no priestess of Aphrodite of Paphos is known to me either from literature or from epigraphy. But it is quite possible that the Rhodokleia of the present inscription was also a High Priestess. An exception may be given by an enigmatic inscription from Palaepaphus, now lost, of which the lettering points to the second century (cf. note 85 above). This is JHS ix (1888), 249, no. 101. Here, if the edd. are to be trusted, the Hellenistic phrase seems to recur; and, if this be so, the inscription as a whole may have read:

The date which I give to this is supported by the late orthography of Possible implications and the title I shall discuss elsewhere.

¹²² IGR iii, 929; probably JHS xxviii (1908), 198, no. 31 = SEG vi, no. 812, from Paphos, a fragment of an architrave, now in the police station at Ktima; IGR iii, 946; 931 = No. 2 above; and an unpublished dedication to Pius from Soli. For the title Armeniacus, cf. Cagnat, Épigr. lat. ⁴, 199, 200, 202.

¹²³ JRS xxix (1939), 190; PIR ² ii, 172, no. 795.

¹²⁴ PIR ² ii, 173, no. 800; probably a brother of our proconsul. My original failure to recognise the name Paterculianus and still more so my misreading of the last three letters of 1. 11 must be ascribed to carelessness.