² Cf. footnote 85 in my ‘Notes on some Published Inscriptions from Roman Cyprus’ in BSA XLIIGoogle Scholar.

³ The names Τρύφων and Τρύφαινα a occur at Amathus (3), Curium, Citium and Old Paphos; (Φίλων at Amathus (4), Curium, Old Paphos (2) and Khandria. Both names therefore appear to be characteristic of the south of the island. Against this should be set the great preponderance of funerary inscriptions from this area, and accordingly a much larger prosopographia.

⁴ The figure on a silver stater of Nicocles, King of Paphos (Head, , Historia Numorum,² 741Google Scholar) is thought by some to be Nemesis, by others Aphrodite, by others again Aphrodite–Nemesis: Herter, H. in RE XXXII Halbband (1935)Google Scholars.v. ‘Nemesis’.

⁵ Published by myself in JHS LVII (1937)Google Scholar no. 8. I soon saw that [Τύχ]ηι was to be restored for my original and inept [Βουλ]ήι. It was too much to expect that this would pass Professor Robert: Études anatoliennes, 1937, p. 175⁶.

⁶ IGR iii. 962Google Scholar.

⁷ For the identification of Nemesis with Dike and Tyche, cf. the hymn of Mesomedes, a contemporary of Hadrian; Dessau, 3737 from Capua (Iustitiae Nemesi Fatis); Amm. Marc., xiv. 11, 25 f.; Herter, H., RE XXXII Halbband (1935) 2375 ffGoogle Scholar. For the worship of this goddess in the Greek East, cf. Seyrig, , Syria, XIII (1932) 50 ff.CrossRefGoogle Scholar; Volkmann, , Arch. Rel., XXXI (1934) 57 ff.Google Scholar; Visser, E., die Götter u. Kulte im Ptol. Alexandrien (Amsterdam, 1938) 42, 43Google Scholar.

⁸ In my ‘Notes on some Published Inscriptions from Roman Cyprus’ in BSA XLIIGoogle Scholar.

⁹ Cf. note 75 below.

¹⁰ A ἄπαξ λεγόμενον, but clearly derived from the Homeric κνίση. Professor W. M. Calder writes to me that he does not think in the circumstances the word can be a neologism. I agree. Phoenician seems to have died out both at Lapethus and at Citium by the middle of the third century B.C., so that by the time of Commodus the island had been exclusively Greek-speaking for four and a half centuries. Furthermore, in the majority of the coastal cities Greek must have been at least the official language from Homeric times. This is not the environment in which linguistic experiments are to be expected.

¹¹ For νεώτερος, cf. MAMA I 422Google Scholar; where it is noted by Calder that ‘probably indicates that Albus was the younger son of Hermon … whereas would normally mean Albus son of Albus, son of Hermon.’ Also MAMA V 204Google Scholar; VI 74, where νεώτερος after the Roman tria nomina indicates that a son bore the exact name of his father.

¹² C. B. Welles, Gerasa, City of the Decapolis: the Inscriptions, no. 27.

¹³ The three instances cited by Avi-Yonah in his Abbreviations in Greek Inscriptions are from the second, the third a the third (or early fourth) centuries.

¹⁴ Some milestones of the Septimian period from Western Cyprus (JRS XXXIX (1939) 184 ff.Google Scholar) suggest that the island as a whole was divided up into city territories and that these divisions were recognised by the government as administrative areas. But the mines, both at Soli and at Tamassus, were the property of the State; while there may have been imperial estates—former possessions of the Ptolemies or unclaimed forest land—of which we know nothing. If we are to restore ἅρχων or some such word in the present inscription, it may indicate that the central massif was invaded at least at this point by the territory of a city.

¹⁵ Notable forms are: alpha and epsilon, with the horizontal stroke disconnected; delta with one side vertical and the apex pointing left. The first delta is very long and narrow.

¹⁶ Omega is formed from two disconnected portions, which resemble the capital letters V and U.

¹⁷ Delta with a vertical side is well illustrated in Fig. 5. The horizontal stroke in epsilon is disconnected. In phi the upright does not bisect the circle. Omega is a sprawling letter, tilted over sharply to the right.

¹⁸ Two interesting forms, xi and omega, are clearly shown in Fig. 6.

¹⁹ Xi has the form of an early Hellenistic sigma.

²⁰ As seen by the observer.

²¹ Two remarkable forms, well illustrated in Fig. 8, are xi and omega. The first is like a Hellenistic sigma reversed; the second resembles phi with the upright not continued beneath the circle. Upsilon is V-shaped.

²² JHS IX (1888) 261Google Scholar, no. 3; Myres, HB 321, 549Google Scholar, no. 1912. Cf. note 50 below.

²³ Preisigke, Namenbuch s.v.

²⁴ Bechtel, H.P. 415, 417.

²⁵ Myres, HB 322, 550Google Scholar, nos. 1914 and 1915; KBH 21.

²⁶ Cesnola 285 f.

²⁷ Welles remarks, op. cit. 360, that ‘the second century shows a marked preference for rounded forms.’ This is true of Cyprus also, but there are some remarkable exceptions: e.g., IGR iii. 986Google Scholar, which can be dated with certainty to the seventh year of Nero.

²⁸ Cook, , Zeus II 598 ffGoogle Scholar.

²⁹ Thus Zeus Meilichios at Amathus itself, JHS LVII (1937) 29Google Scholar, was doubtless a Ptolemaic importation from the Aegean (so Hill 80¹). In general the more remarkable cult-names are to be found in the interior of the island.

³⁰ KBH 21; Hill 107.

³¹ Dialectic survivals into the Hellenistic period are: στρόφιγξ = whirlpool?, SEG VI 838, 839Google Scholar; ‘Ορομπάτας = Ορειβάτης, Hermes, L (1915) 158Google Scholar (note 63 below); άρχός, LBW 2798; ἀρχεύω, OGI 166; ήγήτωρ, BMI 975, OGI 164: κάρπωσις, BMI 975, with Hill 78¹, Bechtel, , Gr. Dial. I 449Google Scholar. From the Roman period perhaps only ὀπάων, for which cf. No. 14 below. The vocabulary of the Cypriot dialect is discussed by Bowra, in JHS LIV (1934) 54 ff.CrossRefGoogle Scholar

³² Acta S. Auxibii, 6 in Acta Sanctorum: templum Jovis dei. Further, the inscription SCE III 623Google Scholar, no. 8 (Robert, L., REG LV (1942) 363)Google Scholar, now in the Statens historiska Museum at Stockholm. To the kindness of Mr. O. Wessberg I owe a squeeze, on the strength of which a version can be offered which differs in certain details from that of Robert:

L. 2: after the second Σ, an undoubted Ε; thereafter a 33, 34 Notes on p. 34.

lacuna of two or three letters. L. 3 begins with a lacuna of two or three letters, according to whether the first surviving sign is Λ or Μ, both being equally possible. After Ε, the segment of a curved letter which is almost certainly Ο. This is followed by a lacuna of two or three letters. At the end of l. 4 are two broken characters which from the squeeze are certainly ΟΣ. Before these a lacuna of two or, with a narrow letter, three. L. 5 has lost at the beginning two letters of average width.

Ethnics of suitable length beginning with ΣΕ are not numerous. I find only Σελγεύς (from the Pisidian Selge) and Σερίφιος, and of these the former is geographically the more likely, while Seriphus at this period was little more than a penal colony. The meaning of this cryptic inscription seems then to be as follows. One Philotimus (?), who is presumably a citizen of Soli since he is given no ethnic, is instructed by an oracle to dedicate on behalf of the people of Selge (?) and a certain Leon a statue of Zeus Olympios at the shrine of a deity whose Beiname is concealed in the letters.. ΠΡΕΙ. Here no suitable verb suggests itself; while the verb is, furthermore, regularly omitted with the dedication of statues. It is unprofitable to speculate on Philotimus' relationship with a foreign city— he may for all we know have been πρόξενος—or with the mysterious Leon; but it may be noted that the inscription as thus restored does not substantiate the templum Jovis dei of the Acta S. Auxibii.

³³ CIG 2641; LBW 2739; de Vogué, , Journal asiatique, Series VI, Vol. X 103, 165Google Scholar.

³⁴ Hermes L (1915) 158Google Scholar (note 63 below). For Zeus Meilichios, cf. note 29 above.

³⁵ E.g., Ἀρίστιον (3), (2), (2), , all from the Hellenistic period. Roman instances are equally numerous.

³⁶ The following forms are notable: delta, with the left diagonal much longer than the right; mu, like eta, but with right hasta and horizontal curved; upsilon in l. 1 is like iota, but has a short stroke branching off to the right near its top.

³⁷ SEG VI (1932) no. 672Google Scholar; cf. also θεὸς ἀσυλαῑος in Plutarch, , Rom., 9Google Scholar.

³⁸ The name Ἀσία was common in Egypt: Preisigke, Namenbuch s.v. Bechtel H.P. 85 cites Fασίας from Lebadeia for the second century B.C.

³⁹ Perdrizet, , BCH XX (1896) 361Google Scholar, nos. 1 and 2: and . Both are now lost.

⁴⁰ Dain, A., Inscr. grecques du Musée du Louvre, textes inédits (1933) 83Google Scholar, no. 71: .

⁴¹ At Athens, both Zeus Hypsistos and Theos Hypsistos: Hesperia I (1932) 193 ffGoogle Scholar. The definitive study of this cult is that of Roberts, Skeat and Nock, in the Harvard Theological Review XXIX (1936) 39 ff.CrossRefGoogle Scholar

⁴² Harvard Theological Review, loc. cit. 63: ‘in Cyprus and at Athens Theos Hypsistos is a god of healing, perhaps some form of Esmun.’ Ibid. 63⁵⁹ is cited a Delian inscription, almost certainly Jewish, in which the god is thanked for a healing.

⁴³ Menardos, S., Ἀθηνᾱ XXII (1910) 444 fGoogle Scholar. The date of this inscription, which I discuss in a forthcoming number of Byzantion, is not Hadrianic or Constantinian; it belongs to the sixth rather than to the fifth century and records the complete restoration of a synagogue.

⁴⁴ Beaudouin, and Pottier, , BCH III (1879) 167Google Scholar, no. 12: , inscribed on a cippus.

⁴⁵ LBW 2740 and Perdrizet, , BCH XX (1896) 363³Google Scholar = , ‘sur un autel rond, orné de guirlandes’—presumably a typical funerary cippus. Examined in a house at Larnaca; but as this town has long been noted as a market for antiquities, its attribution to Citium is questionable.

⁴⁶ If Imperial at all, the inscription must belong to the very outset of the Empire, for rho does not have the rectangular top characteristic of Roman times. Pi in l. 3, with its right hasta represented by a reversed comma, suggests the first half of the third century B.C., but the same letter in l. 1 does not seem to have this distinctive shape. Whether alpha has a straight or broken cross-bar cannot be determined.

⁴⁷ Bechtel, H.P.; Preisigke, Namenbuch.

⁴⁸ JHS IX (1888) 152 ff.Google Scholar

⁴⁹ These are:

(1) JHS IX (1888) 261Google Scholar; no. 2. Without doubt one of the two inscribed statuettes from Amargetti in the Ashmolean Museum. These were, however, packed away during the war and are not yet available again for study. . L. 3. JHS: . For the Cypriot name Ὀνα, cf. SCE III 622Google Scholar, no. 4 (from Soli; early Hellenistic); while it may occur at Amargetti itself: JHS IX (1888) 261Google Scholar, no. 5, discussed below.

(2) JHS IX (1888) 261Google Scholar, no. 4. In the Museum of Classical Archaeology at Cambridge (together with JHS IX (1888) 261Google Scholar, no. 5 and p. 262, no. 13; they are thus not included by Heichelheim, F. in his ‘Greek Inscriptions of the Fitzwilliam Museum’ in JHS LXII (1942) 14 ff.CrossRefGoogle ScholarJHS text confirmed.

.

(3) JHS IX (1888) 261Google Scholar, no. 5. Museum of Classical Archaeology at Cambridge. The following text is taken from my squeeze:

.

L. 1: part of the right diagonal of Λ is visible on the squeeze. L. 2: one arm of Υ is clear. L. 3: first Ο is preceded by an uncut space; before this a fragment of a diagonal as in A. L. 4: Θ is preceded by the tip of either Ε, Σ or Υ; Hogarth's A seems to me excluded.

For ὰλιεύς, cf. MAMA III 521Google Scholar from Corycus; no suitable ethnic suggests itself as an alternative. The lacuna in l. 3 was doubtless occupied by two names, of which the second was possibly Μαχάων (Tod). But perhaps the uncut space before Ο is significant, and in that case ΟΝΑ may be the genitive case of the name Ὀνᾶς (cf. JHS IX (1888) 261Google Scholar, no. 2 as restored above). L. 3 might then read ; but against this is tne fact that we would normally expect υἱοὺς Ὀνᾶ. The participle in the last line was perhaps attracted into the nominative through the dedicator sharing with his sons in the same deliverance, doubtless from death at sea. For ῥύομαι in the late κοινή, cf. Ev. Matth. 6, 13; Ev. Luke, 1, 74. I can see no reason to accept Hogarth's statement that the stone is a foreign marble.

(4) JHS IX (1888) 261Google Scholar, no. 6; CCM 164, no. 5922. Now in the Cyprus Museum. Text revised.

.

L. 1: the second Α is no longer legible.

(5) JHS IX (1888) 261Google Scholar, no. 7; CCM 163, no. 5921. In the Cyprus Museum until 1894, the effective date of the CCM; I could find no trace of this inscription and presumably it has now perished.

.

JHS shows nothing after ΥΠΕΡ, at which point it is indicated that the stone was broken away. CCM: , which, however, cannot be correct.

(6) JHS IX (1888) 262Google Scholar, no. 8. Lost without trace.

.

(7) JHS IX (1888) 262Google Scholar, no. 9. Lost without trace.

.

JHS: . The two signs at the beginning may represent 200 drachmai (cf. Avi-Yonah, op. cit. 114 for the sign ŀ. Alternatively, they may have been misread for Lε′, a fifth regnal year.

(8) JHS IX (1888) 262Google Scholar, no. 10; CCM 164, no. 5924. In the Cyprus Museum until 1894, since when it has apparently perished.

.

JHS:? Φί]λαιος; but the name Adlos is known in Egypt: Preisigke, Namenbuch s.v. CCM: . Since CCM mistakes the fifth line for the first, it is probable that by 1894 only the last three lines survived.

(9) JHS IX (1888) 262Google Scholar, no. 12. I have failed to trace this inscription.

.

In l. 3 JHS has only ll.

⁵⁰ JHS IX (1888) 261Google Scholar, no. 3: . Without doubt the second of the two inscribed statuettes, which are known to have been in the Ashmolean Museum unti l 1940. The inscription is dated to a thirteenth regnal year in (it would seem) the second century A.D.

⁵¹ These are:

(1) JHS IX (1888) 260Google Scholar, no. 1. Apparently left in situ. I have not seen this inscription.

.

JHS: καθιερώσει.

(2) JHS IX (1888) 262Google Scholar, no. 11. A terracotta tablet, lost without trace.

.

JHS: , however, is the normal abbreviation for φιλοτιμία or φιλοτιμος (Avi-Yonah, op. cit., p. 108) and it would seem that φιλοτειμία is here used for the normal ἐκ φιλοτειμίας.

(3) JHS IX (1888) 262Google Scholar, no. 13. In the Museum of Classical Archaeology at Cambridge. My text is based on a squeeze and photograph.

.

Hogarth read ειτ′. Ἡρα(κλείου)? κ′ and explained this as the ‘20th day of the month Heraclius (or Heraion?) of the year 257 A.D.’ The implications of this new reading, which presents no difficulty beyond the omission of ο, I shall discuss elsewhere.

⁵² (1) Cesnola, 414, no. 3; Colonna-Ceccaldi, , RA XXVII (1874) 69 f.Google Scholar; Myres, HB 321, 549Google Scholar, no. 1912. Now in New York:

.

(2) Cesnola, 414, no. 4; Colonna-Ceccaldi, loc. cit. There has been no record of this inscription for more than eventy years, and presumably by now it has perished.

.

On the provenance of these two inscriptions, ascribed by Cesnola to Old Paphos, cf. JHS IX (1888) 262Google Scholar. In l. 2 of Cesnola, no. 3 iota mutum is written adscript by Cesnola, subscript by Myres: whether in fact it appears on the stones another matter.

⁵³ Cesnola, 416, no. 8; Colonna-Ceccaldi, , RA XXVII (1874) 194Google Scholar; Myres, HB 321, 550Google Scholar, no. 1913; inscribed on the breast of a statuette, now in New York, . Ascribed by Colonna-Ceccaldio the Salinae at Larnaca, the provenance given by Cesnola is New Paphos. Myres points out that the statuette is very reminiscent of those from Amargetti, and I have little doubt that this in fact is where it came from.

⁵⁴ JHS IX (1888) 261Google Scholar, nos. 2 and 3; cf. also Cesnola, 416, no. 8 of note 53 above.

⁵⁵ JHS IX (1888) 261Google Scholar, no. 4 can be dated with confidence to the first half of the third century B.C. Nos. 1 and 12 belong rather to the first than to the second half of the following century. No. 11 is probably the earliest representative of the Roman period; the monumental Ω given by Hogarth is unlikely to occur in a site such as Amargetti after the first century, while the sign σ does not antedate the Christian era. No. 2, with its mixed square and oval alphabet, may be Hadrianic; the remainder seem to me to be Antonine or Septimian. Myres finds the lettering of Cesnola nos. 3 and 8 ‘late.’

⁵⁶ Cf. Hogarth, , Athenaeum no. 3164 (1888) 769Google Scholar; JHS IX (1888) 171 ffGoogle Scholar; Devia Cypria 24 (two separate interpretations offered); Reinach, S., REG II (1889) 225 ff.CrossRefGoogle Scholar; Usener, , Götternamen 145 ff.Google Scholar; Kruse, , RE XXIX Halbband (1932) 427Google Scholar; Herter, , RE XXXVIII Halbband (1938) 1696Google Scholar; Hill, 80³.

⁵⁷ For the single consonant in Ἀπόλωνι, cf. MAMA V no. 173.

⁵⁸ The earliest instance cited for the use of the sign L = ἔτος is usually the Cypriot inscription JHS IX (1888) 231Google Scholar, no. 15 (so Avi-Yonah, op. cit., 114), dated by its editors to the year 299 B.C. But from the character of the lettering the L κδ′ of the inscription should be referred to the reign of Euergetes rather than Philadelphus, to 224–3 rather than 262–1 B.C.; for there can be no question of a Ptolemaic era. So far as I am aware this is the first occurrence of the sign L in Cyprus.

⁵⁹ Cf. note 31 above.

⁶⁰ Westholm, , Temples of Soli (1936) 150Google Scholar and Pl. XIX.

⁶¹ Cf. MAMA III no. 28.

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

⁶³ Sittig, , Hermes L (1915) 158Google Scholar (Cook, , Zeus II 869Google Scholar; Hill, 80¹), now in the Cyprus Museum. The fracture in l. 1 is of such a length that a maximum of seven letters have been lost, while the balance of the inscription suggests either three or four. Either Κυπρ[ίαι] therefore or Κυπρ[ίαι].

⁶⁴ Smith, A. H. in Excavations in Cyprus (1900) 97Google Scholar, no. 7; BMI 975; Nilsson, Gr. Feste, 369Google Scholar; Hill, 78. This inscription, hitherto considered Roman, is from its lettering certainly Hellenistic and probably of the reign of Philometor. The phrase supports this dating, or at the outset of Epiphanes' reign the Ptolemaic governor of Cyprus became officially high-priest of all the cults of the island, and thereafter to the end of the Hellenistic period no other high-priests are found save him. Cf. a recently published inscription from Citium (?), Arch. Pap. XIII i (1938)Google Scholar no. 12. Before this and throughout the Roman period high-priests are common.

⁶⁵ Once only, of Athena : LBW 2788.

⁶⁶ Sittig, E., ZVS LII (1924) 194 ff.Google Scholar; SEG VI (1932)Google Scholar no. 818; Bork, , ‘Die Sprache von Alashiya’ in Mitt. Altorient. Ges. V, i (1930)Google Scholar; Hill, 53³.

⁶⁷ E.g., Macrobias, , Sat. iii. 8Google Scholar on the ‘bearded Venus’; Plutarch, , Thes. 20Google Scholar on the cult of Ariadne. For the Phoenician element, Hill in Mél. Boisacq I (1937) 485 ff.Google Scholar

⁶⁸ It is generally agreed that few if any of the megalithic monuments of Cyprus are prehistoric. They are discussed in Hill, p. 21. It might otherwise be suggested that a group of standing stones within a circle of orthostats became the holy place of the Seven within the Stelai.

⁶⁹ Hill, 71 f.

⁷⁰ Seyrig, notes, BCH LI (1927) 142⁵Google Scholar, that imperial statues are regularly dedicated by a representative of the Empire. Since our shrine is dedicated to Vespasian, the intervention of the proconsul is normal.

⁷¹ Hill, 80²; Ovid, , Met. x. 221, 238 ff.Google Scholar

⁷² Hill, loc. cit.

⁷³ As in SEG VI (1932)Google Scholar no. 810.

⁷⁴ Two unpublished inscriptions are evidence that the Salaminian calendar was in use during the first century at Tamassus and at Tremithus. But the majority of the coastal cities followed the Paphiot system. I hope to discuss the chronology of Roman Cyprus, a subject on which there is still much that can be said, as soon as the more pressing claims of the unpublished texts have been met.

⁷⁵ Cf. Hill, 235 f., where the current views on these calendars are briefly but admirably set out.

⁷⁶ The latest list, that of Hill, p. 254 f., includes L. Bruttius Maximus. (Brutius) Praesens, the friend of the younger Pliny (Ep. 7, 3; PIR² I 370Google Scholar, no. 161), was presumably of senatorial rank and therefore the first known member of the gens Bruttia to be ennobled. For his distinguished descendants, PIR² I 370 f.Google Scholar, nos. 164, 165, 166 and 167).

⁷⁷ Cf. my ‘Notes on some Published Ihscriptions from Roman Cyprus’ BSA XLIIGoogle Scholar.

⁷⁸ IGR iii. 944Google Scholar lays special emphasis on the Divus Vespasianus.