^{page 2 note 1} J. H. S. ix. pp. 151, 174.

^{page 2 note 2} Berlin, 1888, where references are given to the previous literature, and to the principal objects in the museums. The abundant illustrations are an admirable feature of the publication. As regards the Pasiades alabastron it may be interesting to add that a similar vase is said to have been found in a fragmentary condition. I do not know how far the description of it given me is accurate, but it might be worth while to track it down.

^{page 2 note 3} For an explanation of ‘rights of excavation’ v. J. H. S. ix. p. 161 note.

^{page 3 note 1} So far as Arsinoe is concerned the case is proved. Strabo, 683, places Arsinoe between the Acamas and Soli: The Stadiasmus Maris Magni 309 is more precise: cf. Ptolemaeus V. 14. The learned editor of the Geographi Graeci Minores thinks the distance 70 stades is inaccurate, but according to the Government survey map Poli is, as the crow flies and as a ship would sail, almost exactly nine miles from the point of the Acamas. The evidence is clenched by the inscription 2781 in Le Bas and Waddington.

The claim for Marium is less irresistible, but very strong. The city was destroyed by Ptolemy Lagi (Diod. xix. 79). It seems to have been refounded as Arsinoe, probably by Ptolemy Philadelphus (cf. Le Bas and Waddington, 2782), for Steph. Byz. remarks: and: There were two or three cities of the name Arsinoe in the island, but that near Poli best suits Scylax 103, where Marium is named after Soli and before Amathus. Moreover, the Stad. Mar. Magn. 233 reads as emended: where the MS. has The objection to the emendation is that Marium was destroyed nearly three centuries before the date to be assigned to the source of this part of the Stadiasmus, but the paragraph 233 bears every mark of having been inserted by the compiler from some other authority. Again, the archaeological evidence shows that there was a settlement here considerably older than Arsinoe, and thoroughly bears out the character attributed to Marium by Scylax's epithet Ἑλληνίς. Dr. Herrmann goes farther, and attempts to identify a separate site for the earlier and later foundations, but his ingenious argument is based on untrustworthy information and erroneous preconceptions, v. J. H. S. x. pp. 281–2.

^{page 4 note 1} Cf. Strabo, loc. cit. and Scylax 103: καὶ αὕτη (Soli) The Stad. Mar. Magn. 311 (here three centuries later than Strabo) speaks of Soli as πόλις ἀλίμενος.

^{page 5 note 1} The General's operations at Poli seem to have been of the slightest, v. Cyprus, pp. 226–7. He contrives, however, to set the village on the wrong bank of the river. I can find no mention of any inscription taken from the neighbourhood.

^{page 5 note 2} Wandering over the site on Feb. 13 I picked up near a remnant of wall foundation just south of the temple-site the inscribed fragment No. 18 below. There is plenty of the same red and white streaky-bacon stone lying around. Mr. Tubbs places the inscription in the early part of the 6th century B.C., but I cannot rid myself of the impression—shared by others who have seen the stone or a squeeze from it—that it is more probably of late Roman date. The circumstances of its discovery certainly point in this direction, for the stone lay loose on the surface among miscellaneous rubbish, which included a fragment of Roman mosaic pavement and other congruous objects.

^{page 6 note 1} Cf. the accompanying plan.

^{page 6 note 2} V. Das Gräberfeld von Marion, pp. 7 and 12.

^{page 6 note 3} The figures quoted by Dr. Herrmann show 261 tombs opened in the eastern, to 180 in the western necropolis. He accordingly sets down the former as the more extensive, rather rashly in view of his admission in the next sentence that the limits of neither had been reached. Our figures, added to the above, leave a balance of about 50 on the side of the western, but still neither necropolis is completely worked out.

^{page 7 note 1} Not, of course, the subterranean aqueduct above referred to, but the modern water-course which comes from Chrysochou.

^{page 8 note 1} Individual objects may of course be earlier, e. g. the very incomplete Cypriote capital, found in dispersed fragments in the shaft of tomb N, and much of the pottery might be of almost any date. I know no cogent reason for separating the two inscriptions from the other contents of the tombs in which they were found. B, a virgin tomb, and apparently Ptolemaic, produced a cup and a lamp, with two symbols from the syllabary scratched upon it. Cypriote inscriptions with fully-developed apices are, I believe, known.

^{page 10 note 1} To find both in the same tomb is not uncommon. Sometimes the inscription is bigraphic, e.g. ΘΕ and te· (7), ΔΙ and ti· (K. 45).

^{page 16 note 1} Strabo, 683, quoted above, p. 3 note. The notion is not borne out by the results of Mr. Tubbs' excavation, v. Section V. below.

^{page 19 note 1} As Dr. Herrmann does, op. cit. p. 8; cf. Cesnola, , Cyprus, p. 226Google Scholar.

^{page 20 note 1} Cf. Herrmann, op. cit. pp. 8–11.

^{page 22 note 1} Hermann, op. cit., Fig. 3, is a sufficient illustration.

^{page 23 note 1} Cf. Cesnola, , Cyprus, p. 67Google Scholar.

^{page 23 note 2} Cf. J. H. S. ix. p. 269.

^{page 23 note 3} Cf. Ibid. p. 270.

^{page 24 note 1} Op. cit., Fig. 4.

^{page 24 note 2} Cf. J. H. S. ix p. 266.

^{page 24 note 3} Cf. Perrot, and Chipiez, , Hist. of Art in Phoenicia, ch. 3Google Scholar. The steps to the Amrit tombs and others are noticeable.

^{page 26 note 1} Op. cit. p. 9.

^{page 29 note 1} That from K and one of those from K. 58 are broken below, but seem to have been the same as the rest.

^{page 29 note 2} Cyprus, p. 114.

^{page 29 note 3} Ibid. p. 109.

^{page 29 note 4} Cf. the inscription on the block from B. 12, where I believe ἐνέστασα is to be read, cf. Deecke, Sammlung No. 71.

^{page 31 note 1} In writing this section I have here and there profited by suggestions or information from Mr. A. S. Murray, Mr. Cecil Smith, and Mr. A. H. Smith, to all of whom I desire to record my thanks.

^{page 34 note 1} Cf. J. H. S. ix. p. 269.

^{page 34 note 2} Cf. J. H. S. ix. p. 270.

^{page 38 note 1} Cf. a very full treatment of the class in Dr. Herrmann's work § IV.

^{page 38 note 2} For illustrations v. fig. 6; Herrmann, op. cit. taf. 3, and figs. 32, 34, 36 to 44; Cesnola, Cyprus, pl. xliii. cf. p. 101; Cesnola, A., Salaminia, figs. 284–5.Google Scholar

^{page 40 note 1} Dr. Herrmann complains of this lack in the case of one of the jugs figured on his taf. 3 and ascribes it wrongly (p. 59) to rude and superficial modelling.

^{page 40 note 2} One may observe how the wings are utterly ignored in laying on the coloured decoration.

^{page 40 note 3} It is doubtful whether the tombs queried ought not rather to be classed as containing only jugs with figurines—fragments of the (d) variety were found in them which, when the fragments are not from shallow basins, probably mean jugs with figurines.

^{page 46 note 1} Possibly also one in tomb 7.

^{page 48 note 1} Published J. H. S. viii. pl. lxxxi.

^{page 48 note 2} Perhaps of Italian manufacture; cf. half-a-dozen precisely similar cups in the Naples Museum. It is to be noted that the cup bears two symbols from the Cypriote syllabary scratched underneath it, the same which appear on the lamp with moulded lion's head spout from the same tomb, mentioned below.

^{page 48 note 3} One is figured, Compte Rendu, 1863, pl. ii. 1. cf. p. 145.

^{page 49 note 1} Cf. with bull, Brit. Mus. E. 494.

^{page 49 note 2} Not uncommon; one is figured in the Compte Rendu. One from the Cyrenaica, Brit. Mus.

^{page 50 note 1} p. 30.

^{page 51 note 1} Cf. Herrmann, op. cit., esp. § III.

^{page 52 note 1} Cf. Herrmann, p. 44.

^{page 52 note 2} I picked up a somewhat similar fragment of fairly good style in the hedge of the vineyard, no doubt a relic of the former excavations.

^{page 54 note 1} Cf. Salaminia, figs. 207–8–9.

^{page 56 note 1} Salaminia, p. 24.

^{page 56 note 2} A larger one of bronze is figured in Salaminia, fig. 69.

^{page 56 note 3} Cf. the Xylino tombs at Kuklia, , J. H. S. ix. p. 269Google Scholar. Salaminia, figs. 232–3.

^{page 60 note 1} As Cypriote inscriptions are very seldom regular, while many of the characters have ‘tails,’ the measurements given are the limits of variation of size. Wherever there is no indication to the contrary each inscription is to be understood as complete, and the limits of the stone are accordingly not drawn on the cut. [I may add here that of two sets of facsimile copies I had made, one has been unfortunately mislaid, the other not returned from the printer. I have therefore not been able to correct to my satisfaction the proofs of the cuts: so far as I can judge from the copies in my note-book they have however been carefully prepared. Salamis, March, 1890.]

^{page 60 note 2} In Collitz, Sammlung d. griech. Dial.-Insch. Heft 1. This, as the standard tract on Cypriote, is referred to here and subsequently simply as ‘Deecke.’

^{page 60 note 3} In Transactions Soc. Bibl. Arch. v. pl. A 3.

^{page 60 note 4} The right- hand sections of the double columns contain the forms given by the graffiti.

^{page 61 note 1} The name of the dedicator ‘Tychon’ suggests a foreign origin: it is not Cypriote. The tomb [Nekr. i. 67] is apparently of the 4th century.

^{page 61 note 2} The main evidence for a date is supplied by a bearded terra-cotta head of poor style: the tomb too belongs to an Arsinoite necropolis. In general where an approximate date is, in this section, assigned to a tomb, the evidence is that of the contents taken in conjunction with the position and circumstances of the grave.

^{page 62 note 1} It is found e.g. Et Mag. 145, 53 as the name of a Korinthian historian, and occurs also in several other places, v. Pape-Benseler, s.v.

^{page 62 note 2} The shading in the cuts of the inscriptions will, I trust, explain its own meaning. It is intended to indicate on the one hand the condition of the stone; on the other, the faintness or strength of tooling in the characters as they at present exist.

^{page 63 note 1} If vo· cf. perhaps of Steph. Byz. p. 319 ed. West; a district near Soli. If si· is preferred—the change depends upon the significance assigned to a stroke at the tail of the letter (see facsimile)—possibly Σι[δωνίου], though it is doubtful whether other than insular appellatives can be represented in Cypriote by a single character.

^{page 63 note 2} Attributives [Beinamen] of deities were used as proper names in Greece, but only, I believe, in compounded form.

^{page 65 note 1} See Scyl. Perip. 100. Dioscor. 5. 146. Hierod, . p. 683, Γαγαία πόλις.Google Scholar

^{page 65 note 2} Dion, . Hal. i. 72Google Scholar. 2:—The form has been generally corrected into Σιγειεύς.

^{page 65 note 3} Strabo, xiii. pp. 595 and 600.

^{page 67 note 1} If this could be established it would be an important result. Deecke's latest inscriptions are, he thinks, of the age of Alexander, : Sayce [Proc. Bib. Arch. 1884, pp. 209 foll.]Google Scholar comes to a similar conclusion from an examination of the graffiti at Abydos, compared with their scarcity at Thebes. Cf. inf. ‘Inscriptions on vases, no. 1, note.

^{page 67 note 2} The tomb, if it existed, was quite shallow, some 4 or 5 feet deep at the most. It probably fell in, and the hole thus made was found useful as a receptacle for waste-stones.

^{page 68 note 1} Prof. Sayce very kindly communicated to me a copy of the graffito in which the character occurs. The graffito itself is published by him in Proc. Soc. Bibl. Arch. 1884, pp. 209 foll., no. 7. He read the sign ka; which is certainly wrong; Deecke suggested le· which is possible, but not very probable, as it requires the name recorded to be read Melermos.

^{page 68 note 2} In Prof. Sayce's graffito the character in question has certainly no tail. If it is to be considered a ne· the name will be Menermos.

^{page 69 note 1} It may perhaps be supported by names like

^{page 69 note 2} Its introduction would have to be regarded as due to Greek influence. The family of Onasagoras may have been immigrant into Marion.

^{page 70 note 1} The tombs in the immediate neighbourhood of M. 1 are mostly of a late period, very often Roman: two, however, which were entered from M. 3 [see plan] were probably of the 4th century.

^{page 71 note 1} It would have been interesting to complete the table of forms in vogue by embodying those given by the inscriptions found in 1886. In the alphabet obtained from the graffiti—illustrated in the right-hand sections of the two columns—I was able to make use of a part of the 1886 Journal, thanks to the courtesy of the authorities at Berlin.

^{page 72 note 1} Sitz.-ber. d. Kgl. bay. Akad. d. Wiss. Hist.-phil. Classe. 5 Mai, 1888, p. 320.

^{page 72 note 2} Six does indeed assign an earlier series of coins to Marion, but on very in sufficient grounds. See the section on Marion in Head, Hist. Num. I came across a small silver coin of Stasioikos, at Xeróbouno, near Limniti: an interesting find as tending to support the view that Marion did stand on the northern and not the southern coast. Kimon's simultaneous attack on Kition and Marion, though seeming to be adverse to the hypothesis, really goes to confirm it.

^{page 72 note 3} This is the name used by the men, and it is convenient to retain it as a specialised term.

^{page 74 note 1} Since I wrote thus I find that J. A. R. Munro has submitted a squeeze of this inscription to a conclave of Oxford archaeologists, and that their verdict is in favour of a Roman origin for the stone. In deference to their authority I wish to modify the opinion expressed in the text so far as to make it less categoric: at the same time I cannot find that an alternative reading was proposed. The place in which the fragment was found tells neither way.

^{page 74 note 2} Sitzber. d. kgl. bay. Ak. d. Wiss. 5 Mai, 1888, p. 320.

^{page 75 note 1} For the title and office see the Rosetta Stone (C. I. G. ad init.]

^{page 75 note 2} Beside the black-ware they are found not seldom on red-figured askoi.

^{page 76 note 1} The vase has now been restored in the Brit. Mus., but some of the fragments do not quite certainly belong to the diota.

^{page 76 note 2} The o· is Paphian, and indeed the inscription as a whole might be so termed. The x-shape of lo· is important: Deecke when he drew up his table was not prepared to admit this variety. pa· is one of the most constant types in the syllabary, but it does vary at Papho (int. al.). Its form on the vase is hitherto unexampled.

^{page 76 note 3} Deecke, no. 82, has another but doubtful instance of Ὤπας used as a proper name in Cyprus. He would read the two symbols o·pa·, on a certain relief discovered by Cesnola, as the genitive of the name Ὤπας. The relief is a curious one: Gen. Cesnola speaks of it as a ‘tablet representing a religious ceremony’; Dr. Hall says: ‘upon the stone is carved a long procession of people. The idea of the sculpture it is hard to trace further.’—I presume this relief is identical with that figured roughly in Col. Ceccaldi, , Mons. de Chypre, p. 75Google Scholar, though Deecke gives no reference to Ceccaldi. M. A. Dumont, in a letter appended to Ceccaldi's quite vague remarks, thinks the plaque represents a sacrifice to Apollo, a dance in his honour, and the subsequent banquet of the who had dedicated the stone. In that case what interpretation is to be put on the two Cypriote characters o· pa·? It is not altogether impossible that they may stand for Ὀπά(ονι), a title under which Apollo was worshipped in Cyprus. In last year's excavations a cult of Apollo Opaon was discovered by Mr. Hogarth at Amargetti [v. J. H. S. 1888: report on Amargetti]. Mr. Hogarth there expresses the opinion that Apollo Opaon was a purely local divinity, and that his second appellative Melanthius conceals the ancient name of Amargetti. If Apollo Opaon was only the deity of a small village it would probably be a mistake to look for his name on a votive tablet from Golgoi. There are however two inscriptions to this god which were found near the Salt Lake at Larnaca. So at any rate Ceccaldi states [Rev. Arch. xxvii. pp. 86–88, Larnaca. nos. 2 and 3]. They are obviously identical with the two inscriptions in Gen. Cesnola, Cyprus (Appendix nos. 3 and 4), where they are characteristically ascribed to Palae-paphos. Mr. Hogarth not having seen Ceccaldi's paper naturally assumed that the General had merely substituted Palae-paphos for Amargetti; but the evidence of Ceccaldi, who was frequently in Cyprus at the time when Cesnola was busy excavating, and often visited the diggings, speaks strongly for Larnaca as the find-spot. Gen. Cesnola's inability to command a wayward imagination would be restrained in the presence of an eye-witness.

If then Apollo Opaon was worshipped at Larnaca, he becomes at once a divinity of greater importance. Though he may have been merely a rustic power at Amargetti, at Kition he seems from the inscriptions to have been a true god of healing (Opaon therefore). If Golgoi may be added as a home of this cult, Opaon Melanthios will become one of the distinct Cypriote types of the god.

The relief is important also in another respect; it carries the use of the Cypriote syllabary down to the 1st century B.C. Dumont dates the work from the 2nd century; but if any reliance can be placed on the sketch in Ceccaldi, this date is considerably too high [and cf. cut Cesnola, , Cyprus, p. 149]Google Scholar.

^{page 78 note 1} If as a number, I would refer to a whole series of signs, which seems to start from a simple form, such as V, and, by the addition of single strokes, to be thence raised in power. was in Deecke's table set down as a presumptive numeral sign, and it will be seen that the series illustrated at the bottom of the table practically passes through

^{page 78 note 2} For Ὤπας, v. sup. under 1 and n. . If this be the right reading of the monogram here it will support the reading in 1; but the monogram may be intended only for the familiar Ὀνα—, an illegitimate stroke having been accidentally introduced.—The two graffiti referred to are from tombs 11, and 11. 78 (1886) respectively.

^{page 79 note 1} The letters are carefully cut, and have been burnt in. The signature is on the shoulder of the vase. Cf. no. 1 supra.

^{page 79 note 2} Graffiti as a rule would not be of much use for determining epigraphic forms; but in Cypriote there is not that decided severance between the monumental and cursive styles. v. sup. p. 73.

^{page 79 note 3} The I is closed at one end, but this is probably accidental. For an alternative view see later.

^{page 79 note 4} For graffiti indicating prices on Greek vases v. R. Schöne, Comm. in hon. Momms., who how ever does not introduce much fresh material.

^{page 80 note 1} This fragment is from tomb S. On a similar saucer-tray from A. 2 are the characters ke· ra· i.e. κέρα, a form of the imperative known in comedy. However, in this instance these may be merely the first part of a name Κεράμων (Xen., Mem. ii. 7. 3)Google Scholar.

The ‘saucer-tray’ seems to have taken the place of the cylix; at any rate the latter is generally absent where the former is found.

^{page 80 note 2} ad Il. iv. 412.

^{page 80 note 3} To go no further, cf. nos. 2866, 69, 73, 75, in the Berlin Antiquarium, which have the painted inscription φιλίας.

^{page 80 note 4} If it were an abbreviation it could only represent τεττα(ρα), and apart from the unlikeliness of finding a numeral written and not symbolized, τέτταρα as a purely Attic form would scarcely be used in Doric Marion.

^{page 80 note 5} pi· lo· on a saucer-tray from β [K. 2], an askos from II. 60 [1886:—the Journal has (by a mistake) on two trays from 1, and a third from 17. pi· lo· I take to be the Cypriote genitive.

^{page 80 note 6} The right-hand halves of the two columns are compiled from the graffiti: signs indicated in brackets are only known from the Journal of the 1886 excavations, and as the Journal, though complete, is by no means scientific, and has in several instances palpably confused an inscription, too great value must not be set upon its evidence for varieties of form. I have however, wherever possible, verified these forms from the fragments and vases purchased by the Berlin Antiquarium at the Paris sale. In the table will be found a suggested new form for tu·. This rests on a graffito which apparently is to be read tu· no·; cf. the name Τυννώνδας Plut., Sol. 14. 1.Google Scholar Two vases from B. 11 have the signs and respectively. The two groups have obviously the same meaning. The typical symbol for tu· is formed from that for to· by addition of an apex, generally applied to the second horizontal bar of If the present graffiti are to be read as I suggests something like a principle in the variations of secondary symbols in Cypriote makes its appearance.

^{page 81 note 1} Thus is found both at Poli and Kahun. The ornate form of te· at Poli may be represented by Mr. Petrie's and the of a vase from 17 may be connected rather with one of the signs on Mr. Petrie's one continuous inscription than with Greek Z [see however sup. p. 73]. A curious compound occurring on a vase obtained at Poli in 1886, compares with in Mr. Petrie's collection. It is not possible here to dwell at length on these coincidences, especially as it is understood Prof. Sayce is at work on a paper dealing with the results of the Egyptian finds in their relation to the Cypriote problem.

^{page 81 note 2} Brackets indicate ‘found in 1886’; asterisk ‘Occurs on vase only.’

^{page 82 note 1} I made inquiry for and purchased this arm on my arrival. It proved to be some 3 in. long, the forearm of a statuette of Cypriote-Greek workmanship. No further portions were discovered, and the fragment itself is quite probably not from Limniti at all. No reliance can be placed on the tales of the villagers, at any rate in the Limniti district.

^{page 84 note 1} So the inhabitants state themselves. There is little or no trace of buildings, which however need cause no surprise, as a few years suffice to turn a deserted house into the mud of which it was originally made.

^{page 84 note 2} Only a single letter was preserved; it had formed the end of a line.

^{page 84 note 3} The Fund had only secured the owner's rights for Mersinéri. By an arrangement with the owner of Ai Nicola I was enabled to test the character of the site, and found that to clear it would have required more time and money than were at my command. It is more than doubtful whether the site would repay the expense of clearing.

^{page 86 note 1} The type exhibited for example in a great number of objects from Kameiros and found spread over the whole Mediterranean, from Syria to Sardinia.

^{page 87 note 1} Attempts to drain off the water produced no appreciable result: the fall into the stream from the fountain is too slight, and had the excavating trenches at this point been dug anything like the proper depth they would have received the whole of the water from this very considerable spring (see plan).

^{page 87 note 2} Some of the smaller objects were more fortunate, though few even of them are left entire. The numerous fragments of arms, legs and feet need not all have originally formed part of statues: bodily members as ἀναθήματα are not necessarily, though certainly in most cases, rounded and finished off so as to be complete in themselves. At Limniti however not one of the fragments but showed a broken edge. It is singular that scarcely anything remains of the torsos: the loss of those in terra-cotta may perhaps be explainable.

^{page 87 note 3} Thanks to the wet loam in which they have been imbedded.

^{page 87 note 4} Head of trench H [see plan].

^{page 88 note 1} The difference is curious and rather instructive: while the Phoenician grove was a ‘high place,’ the Cypriote shrine was regularly down in the valley.

^{page 88 note 2} I came across no indication of the altar itself, although a chance coincidence produced some animal vertebrae [sheep and ox] from a spot near by. It was probably of rough stones, or even of earth only: as generally with a ritual borrowed from the Phoenicians [cf. e.g. Gen. xxii. 9, Ex. xx. 24 and 25, Judges vi. 25, 1 Kings xviii. 30—in fact O.T. passim].

^{page 88 note 3} Two θυμιατήρια were found—one in D, not far from which near D the altar, as I suppose, stood; the other in I. (It should be mentioned that the trenches are lettered consecutively in the order in which they were begun.) Other and similar objects, intended probably for the same use, were turned up in F and on the surface. Cesnola, Atlas, lxvi. 434 and 433, illustrates two ‘altar-shaped offerings’ of stone from ‘the temple of Golgoi.’ They bear, he says, ‘traces of fire’: in those I recovered there were none.

For the practice v. Ex. xxx. 1–10, Levit. v. 1–5.

^{page 88 note 4} The ‘images,’ which in the O. T. are so often mentioned in connection with the ‘groves and high places,’ need not be understood always of the god worshipped, but rather as representing his worshippers, and corresponding to the numerous figures of men found in Cypriote τεμίνη. Cf. Isai. xxvii. 9.

^{page 88 note 5} There are two heads in Berlin, and I found one or two fragments from similar colossi in D and F. Large-sized statues of terra-cotta were sometimes given the requisite strength by a rough core of stone-baked clay.

^{page 88 note 6} The instruments represented are the double flute, cymbals, tambourine.

^{page 88 note 7} For the dress of the professional musician cf. Hdt. i. 24. 6. It is as a god of music that the full dress is proper to Apollo.

^{page 89 note 1} In stone: Statuettes.—The body treated in the lazy, flat style habitual in Cyprus, the heads rendered with more care, and showing the gradual adoption and final supremacy of Greek art-teaching. The stone statuettes are generally female, whereas those in terra-cotta are male.

Miscellaneous: lamps, θυμιατήρια, animals, cone, clay-mill stone.

In terra cotta: Statues above life-size.—Only fragments found.

Statues life-size.—Heads and fragments.

Statues small.—Heads and fragments.

In one case only was part of the torso discovered.

Masks: generally life-size.

Figurines: [a] Cypriote in style and conception.

[b] Greek: probably imported.

[c] Rude ἀγάλματα of native fabric.

[d] Rough genre figures of mnsicians, having reference to the ritual; also mounted figures.

[e] Horned centaur. Several were found in the previous digging by the villagers, and are now in Berlin.

Miscellaneous: Fruits, flowers, animals; part of chariot; φυμιατήριον; object somewhat resembling a strigil; discs (weights, cf. similar objects in Berlin, 6082, 6743–4, 6789, 8162, which however are stamped with names or moulded with figures. See also previous note); lamps; fragment of black-glazed stamped and fluted pottery.

In metal: iron, head of dart; nails. bronze, three statuettes [two imperfect]; animals; coin: leaves of bay; nails; fragments.

In addition I need only mention a blue paste scarabaeoid, with device of a lion and goat (?).

^{page 90 note 1} Perrot takes these figures to illustrate an earlier style of Phoenician dress, afterwards exchanged for a long aba. Instead however of claiming the girdle of the Limniti statuette as the peculiar property of Phoenicia or Cyprus, it may more justly be looked upon as an appropriate ‘active service’ costume. The girdle is generally so recognized in the East. In the O. T. it is continually mentioned, and always in connection with some form of exertion [cf. e.g. 1 Sam. xviii. 4, 1 Kings ii. 5, 2 Kings i. 8, Isa. viii. 9, xxii. 21, Exod. xii. 11, Job xii. 18, and indeed O. T. passim. The use of the metaphor, ‘a girdle about the loins,’ speaks even more plainly. The handling of the bronze in the statuette seems to indicate a leather girdle (for which cf. e.g. Matth. ii. 4, 2 Kings i. 8); but the workmanship is not sufficiently good to make this certain.

^{page 90 note 2} Cf. the Marach bronze illustrated, ‘Perrot and Chipiez’ II. fig. 43 (E. T.).

^{page 90 note 3} That the statuette is male raises no objection. The ear-tire is generally associated with female heads; but for the opposite practice see e.g. Cesnola, , Atlas. pl. xxiii. 55Google Scholar, and cf. also a rude helmeted ἄγαλμα from Limniti now in Berlin [Antiquarium, 2nd Cyprus case], and its counterpart, a terra-cotta head of ordinary size, also in Berlin [ibid.].

The ear-ornament is generally accompanied by a profusion of jewellery—gorget necklaces and stomachers; in at least one case the nose-ring is added [Berlin Antiq. 1st Cyprus case, no. 55; from Dali]. An ear-tire was therefore part of religious full-dress. Such adornment naturally suggests a cult of Aphrodite; and in fact a statuette from Papho shows a female figure carrying, besides these ornaments, a dove [Berlin, ib. no. 64]. It is not necessary to call the statuette an Aphrodite; it may be only a mortal woman wearing the sacred costume of the goddess she worships [cf. Cesnola, , Atlas, pl. x. 12Google Scholar, wrongly described as an Aphrodite]. In the Brit. Mus. is an Astarte-like figure from Kameiros, with similar ear-tire [1st Vase R. Table-case B], and the ornament is not infrequent at Kameiros, which was under the same Phoenician influence as Cyprus. In two, among the several, examples found this season at Limniti, the ear is closely covered by a sort of muslin cap, then drawn together and goffered. Ordinarily the ornament takes the form of a tassel split lengthwise; when Greek style obtains the upper hand it gradually disappears, but is represented for long by a large disc turned full to the front [apparently the sculptor intends a rosette].

The interest of the practice consists in the light it throws on the extent to which Cyprus was, in daily life and manners, oriental.

^{page 90 note 4} In Cyprus especially the rule obtains that the primitive is not per consequence old. A bronze like that in Perrot, Phoenicia, vol. II. fig. 1, is not to be accepted off-hand as an example of early Phoenician work. Perrot remarks: ‘to the feet [of the statuette], which are bare, still hang the sullage-pieces, which may be taken as evidence of the extreme age of the bronze.’ But such ‘evidence’ is extremely doubtful: the sullage-pieces were not left on because the maker did not understand the use of a file, any more than in the Limniti figurine; nor is there anything in Perrot's bronze to connect it with Phoenician art beyond the fact that it comes from Latakiyeh, where a considerable trade is done in ‘Alexandrian’ goods.

It would be an advantage if the term ‘primitive’ were never used in archaeology without the addition of a date. Nothing can be more misleading than simply to describe an object as ‘primitive’: such a description is generally an intentional ambiguity. The ‘primitive,’ ‘Mycenaean, and earliest Cypriote’ pottery' is manufactured still in the island, and may be had any day in the bazaar at Nicosia; it is not intended for the archaeological market, but is simply the ware in common use. ‘Primitive’ vases with geometric devices, especially the concentric rings, are plentiful down to Roman times [and so Colonna Ceccaldi rightly states, Mons. ant. de Chypre, &c. p. 279]; the best examples come from 4th century and Ptolemaic tombs.

^{page 91 note 1} For details cf. preceding description; some confirmation is added below. In a poor district like Limniti the use of solid-cast bronze is significant; there is no question of ideal art. Several little bronzes superficially resembling this from Limniti are in the Lang collection, and are catalogued as ‘kings.’ They are however obvious imitations of the Egyptian ‘Pharoah’ type. A similar figure in stone, Cesn., Atlas, pl. xlviii. 286Google Scholar.

^{page 91 note 2} See illustration.

^{page 91 note 3} For the process en cire perdue v. Bischoff Das Kupfer und seine Legirungen, p. 204. It is not much used now. At Limniti this method must have been still somewhat rude, the core being but roughly-shaped and the wax scarcely moulded at all. The result is that the figures have a general, easy roundness which leaves much to be desired in point of accuracy and truth; but the artist can at least claim that this mode of casting was ‘all his own invention.’

^{page 92 note 1} All of rough workmanship and conventional; the legs are mere stumps. The Berlin Antiquarium also has one such animal and part of a second. The subjects are a sheep, an ox (?), a dog (?).

^{page 92 note 2} The former is at present in the British Museum; the latter at Nicosia.

^{page 92 note 3} Examples may be seen in Cesnola, Atlas, pl. xxv.

^{page 93 note 1} This trait is obviously a reproduction of the type of the ἀγάλματα.

^{page 93 note 2} On the right temple the clay has been broken away, but there was evidently a horn there also originally. On another head there would seem to be some remnant of this feature: the appearance presented by the terra-cotta is just as though the horn had been broken and were hanging down from under the stephane by the skin only. This makes me somewhat doubtful whether the horn may not be simply a part of the head-gear, rendered by the artist in an abbreviated fashion; otherwise there must clearly be here an instance of the ‘horned Astarte,’ appearing however not in her own person but in that of her worshipper and imitatress.

This class of figures may help to explain why at Limniti centaurs appear with horns, long hair falling on the neck, and outstretched arms so curved as to suggest a crescent.

Some of the female heads of this type have either a κίτταρις or helmet.

The three classes of heads and the types of ἀγάλματα will be best understood by the accompanying illustrations.

^{page 94 note 1} A fashion hieratic rather than archaistic in purpose, if a distinction can be drawn. It is useful to compare the ‘Artemisia’ and the head known as ‘Aphrodite’ from the Mausoleum, where there may be a similarity of thought. For shape of wreath v. Cesn., Atlas, pl. lxxv.

The meaning of these Cypriote statuettes is by no means as yet completely explained. Two or three points are certain: there is always a more or less strong suggestion of portraiture: consequently there is not at any temple a single fixed type, such as might, though in a humble sense, be treated as ideal and divine. On the other hand, there is a certain element of continuity in the different groups, supplied by a fixed scheme of dress and ornament. With these data to work upon the theory that kings and priests were represented in these statues was early put forward—a theory accepted apparently by the late Dr.Birch, [Proc. Soc. Bibl. Arch. 1883, May 1st, p. 121]Google Scholar. Ermann thinks the theory has been too hastily adopted, and believes that the statues being ex-votos are therefore necessarily deities. ‘Surely a closer inspection would render it almost a certainty that the two long series of bearded heads, one helmeted, the other crowned with a wreath, ought to represent the two forms of Apollo mentioned above [Ἀμυκλαȋος and ‘ϒλάτη‘On the Origin of the Cypriote Syllabary,’ Proc. Soc. Bibl. Arch. 1883, pp. 113 and foll.]. Ermann's view rests on a quite unsound hypothesis; but neither is the alternative theory altogether satisfactory, The Limniti heads afford some help to a third and, I think, preferable explanation. If they represented kings, these could only be the kings of Soli, and there are far too many statues for the number of reigning kings during the period (v. infra) during which this sanctuary was open. A similar objection, though with less force, applies to the theory which makes them priests; moreover the presence of helmeted heads is here a little awkward. The statues are certainly ex-votos; but instead of being images of the god they are those of his worshippers, whether king, priest, noble, or merchant. By an idea very prevalent in at any rate the Hebrew section of the Semitic stock, the devotee honours the deity by conforming to his likeness [cf. also Plato's well-known doctrine, Rep. x.]. The Cypriote statues represent the worshipper under the aspect of one who in assuming some of the attributes of his god pays him the highest honour he can. In one instance in Berlin [Antiq. 2nd Cyprus case] a helmeted head from a statue is an almost exact reproduction of an ἄγαλμα which has been a little more fully rendered than usual. Contrary to the opinion just expressed, this comparison might seem to require the attribution of both statue and ἄγαλμα to the same deity. But the statue cannot be separated from its fellows; nor can its individualism of rendering be overlooked.

^{page 95 note 1} The torso, where preserved, is on the contrary quite free from this influence.

^{page 95 note 2} Sometimes the flower is absent, but the type is not, in strictness, complete without it.

^{page 95 note 3} Terra-cottas of this class are frequent also in other localities. [Cesn., Atlas, pl. xxvi. nos. 66, 67, 69, where the flower is clearly a lotus, showing both the origin and the meaning of the figures. Nos. 68, 70 on the same plate should be contrasted; they are in the round, and have a different purpose.] There are also male figures belonging to the same type [Atlas, pl. li.]. ‘A stele of calcareous stone, with figure in relief [Atlas, pl. xiv.], is also to be compared; it so far varies from the canonic type that the left foot is slightly raised, a freer motif perhaps due to the necessity of placing the stele upright.

At Limniti there were also examples of the Greek method of handling the theme. The figure becomes quite free, almost négligé, and the butt of the sarcophagus-lid appears as a true pedestal. This class was by no means infrequent in the tombs at Poli; it is represented also by a number of terra-cottas from Kameiros [Brit. Mus. T.-C. R^m]. Those found in Cyprus are very possibly imported, at least specimens from Limniti and Poli have come from a single mould. The Kameiros figures show the sense in which the type was understood in Greece. They are reproductions, with omission merely of one or two attributes of a scheme which, in Greece, can only be assigned to Persephone. Similarly the Cypriote variety is almost a copy of the figure of Aphrodite, as used for the body of an alabastos [several vases of this kind from Kameiros are in the Brit. Mus.]. The dead clothe themselves with the character of the goddess with whom they are associated [cf. sup. p. 41, n. 3].

^{page 96 note 1} Were it critically possible it would be natural to follow the hint supplied by several terra-cottas from Cyprus which are caricatures, and suppose that this type, in a temple, has a satiric meaning. For such figures cf. Cesn., , Atlas, pl. xxiv. 57–59Google Scholar, two of which are given also by Perrot, Phoenicia, II. figs. 137 and 138. Cf. also examples in the Lang Collection.

^{page 96 note 2} So apparently the lines of paint, appearing on the better-shaped specimens, are to be interpreted [De dea Syria, 35; for the beard ibid.].

^{page 97 note 1} One of them however cannot with certainty be ascribed to Limniti: style, clay, and type of head rather suggest the neighbouring Vouni. The objects in Berlin have been obtained from Messrs. Constantinides and Richter of Nicosia, who in turn had purchased from the villagers. It is the statements of the latter which are the only ground for designating the find spot.

^{page 97 note 2} De dea Syria § 35.

^{page 97 note 3} Some of the Ionic towns, Ephesus e.g., show the extent to which Greek religion could assimilate itself to the Oriental. In Cyprus, Orientalism tries to assume a Greek dress.

^{page 97 note 4} Strabo says: (i.e. inland from Soli)

(a) The Greek makes it clear that Arsinoe and the Διὸς ἄλσος are to be closely joined: otherwise the article is otiose, there being no famous ‘grove of Zeus’ along the Northern coast. Strabo goes by stages, and Limniti, were it mentioned, would be included in the Soli district.

(b) A grove of Zeus is not antecedently probable in a purely Cypriote and perhaps somewhat rude district. [At the present time it is the most barbarous in the island.] As against this however stands the fact that Soli was a centre of Greek influence. So however, and probably to a greater degree, was Marion (Arsinoe), Ἑλληνίς as Scylax calls it [Perip. 103].

(c) There exists a yet unexplored temple-site at Poli.

(d) The antiquities found at Limniti have no discoverable relation to Zeus, but are closely connected with Apollo in his Cypriote form. The helmeted ἀγάλματα can only represent Apollo: helmeted heads, like those here dedicated by worshíppers, have been found in other τεμένη of Apollo. Other evidence is afforded by details already described.

As to Limniti being Limenia, the writer who suggests this view can never have read Strabo; and when he makes it the ‘harbour of Soli’ he shows an equal ignorance of the topography of the Soli district, and the practical conditions of commerce.

^{page 98 note 1} For Apollo Amyklaios and his warlike character cf. Paus. III. 10. 8: 19. 2: Plut. de Pyth. Or. 402 A: the helmeted Apollo on coins of Kalymna; Ross Insc. Ined. 282: Bull. Gorr. Hell. 1879. 322, compared with Hesych s.v. ἀγρέτην, &c.: — See generally Preller, ‘Gr. Myth.’ cd. Robert, p. 274. Apollo ‘Amyklaios’ is simply Apollo ‘Mikal’; the adoption of the title being helped by a popular etymology. But Resef-Mikal had another side: he became identified with ‘Adonis,’ the ‘greatest of the gods‘ and patron of Byblos [Strabo XVI. ii. 18, p. 755] the father of agriculture, and the same person as Sanchoniathon's Ἀγρούηρος or Ἀγρότης [Sanch. p. 20 ed. Orelli: for Adonis in this character cf. Aug. de Civ. Dei vii. 21]. Apollo-Amyklaios was also a nature-god: and hence his association with Hyakinthos. A Cypriote festival corresponded to the Hyakinthia and had the same meaning [cf. the Apollo Mageirios of Pyla: Ceccaldi, Mons. de Chypre, &c. p. 199Google Scholar, Pyla Inscr. No. 1]. In Greece Apollo, as the god of agriculture and cattle-rearing, is given the name of Nomios [his functions as Karneios and Amyklaios are similar in many points]: in Cyprus this side of his character is probably represented by the Apollo Opaon Melanthios of Amargetti and Kition.

^{page 98 note 2} The connection between Apollo and Aphrodite is expressed also in the person of Kinyras, the favorite of Apollo and his rival in skill on the harp, but also high-priest and darling of Aphrodite. Kinyras is kinnor, ‘the harp.’ Apollo and Aphrodite are joined together, e.g. at Golgos.

^{page 98 note 3} Beside the ἀγάλματα already mentioned I found a cone-shaped object of stone, conceivably a copy of the sacred emblem [similar shaped cones are in Donaldson Arch. Num. XXX. from a bronze coin of the British Museum; ibid. vi., pediment of the Artemision at Ephesos. In the Papho temple the cone has a different form, ibid. xxxi.]: a fragment of a statuette holding a hare [for the hare as sacred to Aphrodite cf. Philo Her. Erot. and for the reason Hdt. III. 108. 3]: a horned centaur [several others are in Berlin] and the hand of a statue holding a dove.

^{page 99 note 1} Possibly we may restore—

The date given is that assigned it by Mr. Hicks, to whom I submitted a squeeze. As the fragment was obtained from the surface by a villager, and as no further portions were found, nor any place where it might have been set up, its connection with the shrine is not quite certain.