^{page 97 note 1} A History of Cyprus (1940), p. 52. A preliminary discussion of Kafizin, may be found in the Report of the Department of Antiquities, Cyprus, 1937/1939 (1949), pp. 126 ffGoogle Scholar. Excavation and the valuable criticism of Dr. M. N. Tod have, between them, invalidated much of what I said there.

^{page 97 note 2} Thus Meister, R., Gr. Dial, ii (1889), p. 193Google Scholar, considered that dialect and syllabary became extinct at the outset of the Ptolemaic period. Hoffmann, O., Gr. Dial, i (1890), p. 42Google Scholar, is more cautious: ‘das Jahr 300 v. Chr. ist jedenfalls eine viel zu hohe Grenze’. Myres, J. L., Handbook of the Cesnola Collection (1912), p. 301Google Scholar, shares this diffidence, suggesting a survival at a ‘few great Sanctuaries’. On coins, the latest date is given by Hill, as 312–306 B.C.: B.M. Coins, Cyprus (1904), cxiiiGoogle Scholar; History of Cyprus, i, p. 143. Cf. also Larfeld, W., Gr. Epig. ³, p. 202Google Scholar.

^{page 97 note 3} On the relation of syllabary to Cretan Linear A and B, cf. especially Casson, S., Ancient Cyprus (1937), pp. 72 ffGoogle Scholar. A more recent and more critical estimate is that of Kober, A. E., A.J.A. lii (1948), 99 ffGoogle Scholar. Miss Kober's reluctance to see in the classical syllabary a key to the Cypro-Minoan script and, a fortiori, to the scripts of the Aegean is justified by the non-existence of a syllabic palaeography. Deecke's famous table of signs in S.G.D.I. is based, not on a chronological, but on a geographical basis.

^{page 97 note 4} A.J.A. xxx (1926), pp. 249 ff.Google Scholar; Roussel, P., R.E.G. xxxix (1926), p. 285Google Scholar; Tod, M. N., J.H.S. xlvii (1927), p. 213Google Scholar; S.E.G. vi (1932), 838–40Google Scholar; Kasten, H., Burs, Jahresb. cclxi (1938), p. 136Google Scholar. Cf. also Arch. Pap. xiii. 1 (1938), p. 15¹Google Scholar; J.E.A. xxvi (1940), p. 61⁵Google Scholar.

^{page 97 note 5} This text of the two inscriptions on the bowl I give from my own reading of the editors' photographs, A.J.A., I.c., pp. 250–1. The editors are at fault in ascribing the lacuna to the ‘outer’ inscription, and in reading δεκατηϕ⋯ρος instead of δηκατηϕ⋯ρος in the ‘inner’.

^{page 98 note 1} These thirteen fragments, illustrated on p. 253, are poorly rendered on pp. 250, 251. Many broken letters are disregarded. For example, No. 5 should read Ὀνησαγ[⋯ρας] and not Ὀνησα[γ⋯ρας]. In No. 4, for ⋯γαθῇ τ⋯[χῃ] read ⋯γαθτ⋯[χε]; in No. 17, for Ὀνησ[αγ⋯ρας] read Ὠνησ[γ⋯ρας]; in No. 18, for ⋯γαθ⋯[ι] τ⋯χ[ηι] read ⋯γαθτ⋯χ[ε].

^{page 98 note 2} These are Nos. 3, 10, 11, 13, 14, and 19. The last we shall find to be a date. In Nos. 3 and 14 we may recognize an invocation to the Nymph as Ἀδελϕ⋯.

^{page 98 note 3} But of the nine, two are not syllabic. No. 5 is a date. Of No. 6 Professor Kent says, ‘one line of six characters, that at the left fragmentary’, He finds that the O is ‘the ordinary Greek letter’ and proceeds to read:

SA · ne· o· ki· ro· to· = - σ]τρ⋯(ϕι)(γ)γι Ὀνησα[γ⋯ρας -

The inscription is in fact alphabetic, with upsilon and kappa, omikron, and upsilon written in ligature. The correct reading is - Φιλου]ου Κορ[ε⋯ς (rather than the - ]δρου κουρ[ε⋯ς which I previously offered). I may add that there is a further instance of a ligatured inscription among the unpublished texts.

^{page 98 note 4} For Nos. 5 and 6 of his ‘syllabic’ inscriptions I refer to the preceding note. No. 7 is too worn for a photograph to be fully legible. No. 8 is correctly rendered Φιλων⋯υ. For the rest we should read: in No. 3, not Φιλων⋯[ου] but σ]τρ⋯ϕι[(γ)γι; in No. 4, not Ὀ]<να>νη(σα)γ⋯ρα[ς (sic) but Ὀ]να⋯γ⋯ρα[ς; in No. 9, not Ὀ<να>νη[σαγ⋯ρας but Ὀνα⋯[γ⋯ρας. Of 26 signs—I exclude those of No. 7—no less than 14 are misinterpreted.

^{page 98 note 5} p. 257, no. 1. The presence of these fragmentary letters is noted by Mrs. Dohan.

^{page 98 note 6} Cf. inter alios, Néroutsos, , R.A. ix (1887), pp. 290 ff.Google Scholar; Pagenstecher, , A.J.A. xiii (1909), pp. 387 ffGoogle Scholar. For the dating, Pomtow, , B.ph. Woch. 1910, col. 1087 ffGoogle Scholar.

^{page 98 note 7} Arch. Pap. xiii. 1 (1938), p. 15, n. 1Google Scholar. A bronze ring, recently discovered at Curium by Dr. B. Hill, carries an inscription in cursive letters of (I think) the late third century B.C. The style occurs, though perhaps less freely, in both the second and first centuries. Of frequent occurrence in a private or religious context, it is virtually unknown in public documents.

^{page 99 note 1} Sigma with widely angled hastae, alpha with unbroken cross-bar are characteristic of the third century: they are not found after the first two decades of the second.

^{page 99 note 2} The sign L = ἔτος is used in Cyprus to the exclusion of all others throughout the Ptolemaic and Roman periods. It has been thought to be of Egyptian origin. The earliest instance cited by Avi-Yonah, , Abbreviations in Greek Inscriptions, p. 114Google Scholar, is Cypriot but wrongly ascribed to the year 299 B.C.: an inscription of Old Paphos, dated to the 24th year of—I believe—Euergetes I. It should be noted that at Kafizin the horizontal stroke is curved downwards and backwards; and this may militate against the theory that L is in in origin a truncated epsilon.

^{page 99 note 3} A.J.A.: τυηδ On the photograph I see no convincing trace of tau.

^{page 99 note 4} A Phoenician inscription from Idalium, C. I. Sem. 93: Cooke, N. Sem. Inscr. 27Google Scholar, is dated by the 31st year of Philadelphus and the 57th of Citium. An interesting statue-base now in the Cyprus Museum, Colonna-Ceccaldi, , R.A. xxvii (1874), p. 89Google Scholar: Sittig, E., Φων⋯τ⋯ς Κ⋯πρου 7/25 02 1914Google Scholar, was excavated at Idalium and yet is dated by the 47th year ὡς Κιτιεῖ ἄγουσιν. The spread of Phoenician influence from Citium across central Cyprus to the north coast at Lapethus was checked temporarily by Alexander after Issus, and much more radically by Soter in and immediately after 312 B.C. The institution of eras by Citium in 311–310 and by Lapethus in 307 B.C. is generally connected with this emancipation of these cities from Semitic despotism. But Pumiathon, the last king of Citium, was executed in 312 B.C., while at Lapethus an even greater interval elapsed before the era began. Both cities, in short, were somewhat tardy in celebrating their freedom; and this, together with the fact that in each Phoenician cults were maintained, Phoenicians continued to hold civic office, and Phoenician remained an official language, suggests that the explanation is inadequate. Idalium, we may note, is prepared to commemorate her subordination to Citium 47 and (as we shall see) even 90 years after the dissolution of the dominion of that city.

^{page 99 note 5} Dating at Kafizin in general follows the normal Cypriot usage. There is, however, one notable anomaly. Where two figures occur in a date, the larger is invariably preceded by the smaller, thus inverting the order which obtains elsewhere in Hellenistic Cyprus. Out of fifteen instances I find one exception only—and that in a Seleucid document, O.G.I. 257.

^{page 100 note 1} I am grateful to the Trustees of the Craven Fund and to the Leverhulme Foundation for their generous support of this venture; and I must thank also the Travel Fund of the University of St. Andrews. Without this help the excavation could never have taken place.

^{page 100 note 2} The finds of 1949 still await the expert attention of the Cyprus Museum staff. I do not, however, think that this figure of 310 separate inscriptions can be appreciably reduced.

^{page 100 note 3} On a fragment of an amphora which is very much perished. After kappa, however, an uncut space, which indicates the end of the inscription and, accordingly, a date. Before kappa, a horizontal stroke on a level with the top of the letter: since epsilon here is cursive, gamma seems inevitable.

^{page 100 note 4} This date is puzzling. On the sherd I: either zeta, therefore, or the syllabic ve. A year ζκ' cannot, however, fall within the reign of Euergetes; while ζι' would be widely separated from the other dates. If, on the other hand, we have here ve, we must suppose that it is substituted for its phonetic equivalent digamma, precisely as the sign se is written for sigma in an inscription from the neighbouring Golgoi: Myres, , Handbook of the Cesnola Collection, No. 1897Google Scholar. If this be the case, the year 26 will refer to the few months between 18 October 222 B.C. and the death of Euergetes early in February of the following year.

^{page 100 note 5} ḳα⋯ Ϟ', though cramped and poorly written, can be read with confidence. For kappa the potter may have written chi; but the confusion of these two letters is characteristic of Kafizin. Cf. below, p. 101, n. 5.

^{page 101 note 1} Or, allowing for a possible L κ', at the earliest 228 B.C.

^{page 101 note 2} Cf. Meister, , Gr. Dial. ii, p. 250Google Scholar: the lost sigma was doubtless represented in pronunciation by an aspirate. It is of interest to note that the famous Tablet of Idalium, , S.G.D.I. 60Google Scholar, although some two and a half centuries earlier, preserves this sigma. Being a public document, it shows a greater concern with etymology than with pronunciation.

^{page 101 note 3} For αἶλος cf. S.G.D.I. 60¹⁴. Elsewhere in these inscriptions I note: the presence of the definite article with a patronymic, as in ⋯ Φιλων⋯υ: ⋯π⋯ τ⋯ι στρ⋯ϕι(γ)γι for the ⋯ντ⋯ι στρ⋯ϕιγγι of the κοιν⋯: the formula ἰ(ν) τ⋯χαι. I would add that the syllabary of Kafizin is remarkable for its numerous unique forms. The following occur repeatedly and are well established: , , = pi; , = po; = mi; = mu; = ni. Kafizin demonstrates that the syllabary, even in its final phase, had not become static, and emphasizes the need for a syllabic palaeography.

^{page 101 note 4} Thus τ⋯δε invariably for το⋯.

^{page 101 note 5} Thus epsilon and eta, omikron and omega are used indiscriminately. I note also omikron for omikron upsilon, kappa for chi, theta for tau; the omission of a nasal before xi; the failure to render a double consonant. Iota occurs freely in place of epsilon, once in place of epsilon iota.

^{page 101 note 6} From early Hellenistic Idalium two inscriptions only are known. Both are alphabetic and both are in some measure public documents: R.A. xxvii (1874), p. 89Google Scholar, No. 1, a large block which carried an ex-voto to Apollo of Amyclae; R.A. xxvii, p. 89, No. 2, a record of the dedication of an Arsinoeion. From the more official atmosphere of Old Paphos not a single syllabic inscription of the Hellenistic period is forthcoming. Indeed, it is doubtful if the syllabary survived into the third century in certain of the larger coastal cities. The fact that some sixteen alphabetic texts can be ascribed to the fourth century is proof that the alphabet was becoming fashionable long before it had the sanction of the conqueror.

^{page 102 note 1} There is no reason to suspect that any of the 21 ‘black-slip’ vessels dedicated to Ν⋯μϕηἈδελϕ⋯ are earlier than Onesagoras: none is dated, but on six of them his name occurs.

^{page 102 note 2} Roussel, I.c., p. 285, suggested diffidently that ΚΟΥΡΕΥΣ stands for ΚΟΥΡΙΕΥΣ. In this he is followed by Kasten. And Dr. Tod writes that he believes (as does Sir John Beazley) Κουρε⋯ς to be either ethnic or deme-name. In this connexion it is noteworthy that in the dedication from Idalium cited in n. 6, p. 101, we find a Μνασ⋯ας ἌΨητος ΜΕΤΕΙΡ ΑΣ; while in S.G.D.I. 60²¹ a property in the territory of this city is described as belonging to Διρε⋯θεμις ⋯ Ἀρμανε⋯ς. Both Μετειρας and Ἀρμανε⋯ς are otherwise unknown, and are presumably deme-names, derived perhaps from villages within the city territory. The syllabic spelling ΚΟΡΟis attested by one inscription only: - Ὀνα⋯]γ⋯ρας ΚΟΡΟ⋯ δεκα[τα]ϕ⋯ρος -.

^{page 102 note 3} Names formed from the root ονησι are characteristically Cypriot, while -αγορας is the commonest termination in the prosopographia of the island. It is significant that the name Onesagoras occurs twice in the Idalium Tablet, S.G.D.I. 60: as the father of the eponymous magistrate and again as the father of a landowner. (Cf. further the Assyrian lists of the seventh century, for one Onesagoras, king of Lidir-Ledrae.) In the third century we meet it twice in the neighbouring Chytroi: Myres, and Richter, , Cyprus Museum Catalogue, 5142Google Scholar; Richter, , Ath. Mitt. ix (1884), p. 135, no. 1Google Scholar. It is clearly typical of the central Mesoria.

^{page 102 note 4} Δεκατηϕ⋯ρος, the tithe-receiving, known as a title of Apollo, is found here only as a noun.

^{page 102 note 5} Two constructions are confused in this inscription, π⋯ρειμι being substituted ungram-matically for ⋯μ⋯. We find elsewhere: ⋯γ⋯ π⋯ριμι (sic) κα⋯ ἄλλα πο[λλ⋯ -.

^{page 103 note 1} The name occurs but once in a case other than the genitive, and there its termination is uncertain. If we are correct in restoring Ἀνδροκλῄ[ν—lambda is followed by part of an upright —Ἀνδροκλο⋯ (Ἀνδροκλ⋯ου) will stand for Ἀνδροκλο⋯ς.

^{page 103 note 2} I.G. xi. 2. 135.

^{page 103 note 3} In a Zenon, papyrus, P.S.I. 505Google Scholar: το⋯⋯κ Κ⋯πρου οἰκον⋯μου, an official not mentioned in the epigraphy of the island. Οἰκον⋯μοι are to be found in the various Egyptian dominions, subordinate in rank to the local military commander, but answerable either to the διεικητ⋯ς of Alexandria or, as in Syria, to a local διεικητ⋯ς. In the second and first centuries B.C. the functions of the οἰκον⋯μος in Cyprus were assumed by the οτρατηγ⋯ς.

^{page 103 note 4} For such contractors at Cos cf. Rostovtzeff, , Hellenistic World, p. 241Google Scholar; in Judaea, ibid., p. 349. On the farming of certain taxes in Syria and Palestine cf. Préaux, C., L'Économie royale des Lagides, p. 420Google Scholar. According to Josephus, , A.J. 13Google Scholar. 169, the richest men of Syria assembled in Alexandria as candidates for the contracts of their province.

^{page 103 note 5} In an interesting but mutilated inscription we have: ἔδοξεν ⋯γαπ⋯σαι Ἀνσροκλῂ[ν - -] Ὀνεσαγ⋯ραι Φιλουν⋯ου - -.

^{page 103 note 6} Cf. in S.G.D.I. 60⁵ of Idalium a gift of lands ⋯κ τ⋯ι ρο⋯κωι τ⋯ βασιλ⋯ρος.

^{page 103 note 7} Δεκατισμ⋯ς, in the sense of tithe or payment of tithe, is ἂπαξ λεγ⋯μενον. The word is cited once only by L.S.⁹, from a late author and with the meaning ‘formation of decuriae’.

^{page 103 note 8} Here the syllabic signs, transliterated, give: a. po. ta. i/a. pa. i/re. i. etc. The alternative, ⋯π⋯ τ⋯ι, is no more helpful.

^{page 103 note 9} O.G.I.S. 55.

^{page 104 note 1} Following Heichelheim, ProfessorRostovtzeff, argues, Hellenistic World, p. 332Google Scholar, that Cyprus, as a province where ‘city life was undeveloped’, was very probably treated in a manner similar to ‘those provinces whose cities developed on Oriental lines’. I prefer to group Cyprus with Lycia and Cilicia rather than with the Semitic dependencies. My reasons for so doing I shall detail elsewhere.

^{page 104 note 2} That Tamassus was at this date a π⋯λις in the full sense is proved by its mention in the Delphic catalogue of theorodokoi: B.C.H. xlv (1921), p. 4, col. iGoogle Scholar. This ethnic also occurs in syllabic inscription, as yet unpublished, of (I think) the early third century. Jones, A. H. M., Cities of the Eastern Roman Provinces, p. 374Google Scholar, observes that there is not sufficient evidence to determine the status of Tamassus under the Ptolemies: this evidence is now available.

^{page 104 note 3} On a sherd ]!Δ Α Λ | [, perhaps to be restored [⋯π⋯ τ⋯ς] σαλ⋯[ς ϕορ⋯ς?]. For the unusual form of the ethnic, cf. S.G.D.I. 60¹⁶, ²⁶. Throughout the entire fourth century Idalium was subject to Citium; and it is argued by Jones, A. H. M., op. cit., p. 372Google Scholar, that at the outset of the Ptolemaic period it was so thoroughly semitized as to be an integral part of the Citian republic. The two inscriptions cited above in n. 6, p. 101, do not prove that Idalium enjoyed full civic status; while the fact that the era of Citium is twice employed may denote sub-ordination. But the present inscriptions show no trace of this Semitic influence.

^{page 104 note 4} Lidir is mentioned in the Assyrian lists of the seventh century B.C.; a bishop of Ledrae is found in the fourth century of our era. It is assumed that both refer to the same place, and that this is the precursor of the modem Nicosia (Leucosia). If this be so, Ledrae is lost to view for a thousand years.

^{page 104 note 5} A close parallel is offered by the cave at Vari, , A.J.A. vii (1903), p. 263Google Scholar, transformed into a sanctuary of the Nymph by a faithful devotee, Archidemus.

^{page 104 note 6} Dohan, E. H., A.J.A., I.c., p. 252Google Scholar: ‘an eddying pool in the foothills of Tremithus, fit abode for a nymph’. This view is quoted by S.E.G. and by Herter, H. in R.E. xvii (1936), col. 1, 534Google Scholar.

^{page 104 note 7} Such as may be seen on many summer days in the Mesoria: so Herter, H., R.E. xvii (1936), col. 1538Google Scholar.

^{page 104 note 8} Zingerle ad S.E.G. and Herter, H., R.E. xvii (1936), col. 1, 545Google Scholar, citing from Hesychius ⋯μπ⋯λαι · αἱ ν⋯μϕαι.

^{page 104 note 9} So L.S. ⁹, s.v. στρ⋯ϕιγξ.

^{page 105 note 1} so. to. ro. ΤΙ. ΚΙ = στ⋯ρ(γ)γι? and στ⋯ρ]ιγγι: neither instance, therefore, can be considered certain.

^{page 105 note 2} The ruins of the ancient Chytri lie some 9 miles to the north-north-east, those of Tamassus twelve to the west-south-west. Idalium, therefore, if we exclude Ledrae, is the nearest city.

^{page 105 note 3} Ἐϕημ. Ἀρχ. 1905, pp. 99 ff.; 1906, pp. 89 ff.

^{page 105 note 4} It occurs twice at Kafizin: Ν⋯[μ[ϕηι Ἐπηκ⋯ω[ι] and Ν⋯[μ]ϕ[η]ι Ἐπη[κ⋯]ωι [ελϕ⋯ι?]. As the epithet of a nymph it is rare. But cf. S.E.G. ii, 692; B.C.H. xxi, p. 136, No. 26; Weinreich, , Ath. Mitt, xxxvii, pp. 16 fGoogle Scholar.

^{page 105 note 5} On this date, Tarn, W. W., Cl. Rev. xl (1926), p. 86CrossRefGoogle Scholar.

^{page 105 note 6} Myres, J. L., Handbook of the Cesnola Collection, No. 1900Google Scholar.

^{page 105 note 7} Arch. Pap. xiii. 1 (1938), pp. 28 ff.Google Scholar; Hill, , History of Cyprus, pp. 184 fGoogle Scholar.

^{page 106 note 1} Mr. Kattis spoke of ‘bearded faces’ on five of his vessels. This detail cannot have been the product of his imagination, since the existence of such faces—portraying presumably the votary—was established only after excavation.

^{page 106 note 2} I need hardly add that Mr. Kattis disposed of his finds to his profit. One plaque, he tells me, with ‘five lines of writing in the syllabary’, he sold to a German tourist. Some pieces were smuggled from Cyprus by a Lefkara lacemerchant who is now dead; and from him, I have no doubt, the Philadelphia Museum purchased their acquisitions. Others, again, were sold locally, in Tricomo and Lamaca; but of these, beyond the assertion that some have made their way to France, there is now no trace.