^{page 61 note 1} Φιλίστωρ, vol. ii. p. 274. Copied thence into the Arch. Anzeiger, 1861, p. 234, and Bullettino, 1862, p. 86. See especially the article by Wolter, , Ath. Mitth. xv. 1890, p. 246Google Scholar, to which I refer below.

^{page 61 note 2} The inscription is published on p. 16 of the last volume (xvii) of this Journal. The early notices mention a headless bust and two heads. The second of these may have been the head which is now fitted to the bust.

^{page 61 note 3} Meisterwerke, p. 624. English edition, p. 381.

^{page 62 note 1} We have usually followed Ehrenberg's map (Leipzig, 1889) in spelling the name Tramythia, but Tramithia is as near to the local pronunciation. The form suggests a derivation from Steph. Byz. mentions a place called (v.l. ) in Cyprus and derives the name The name would be formed like that of in Attica = from The form is further attested by Nicand. Ther. 844

The Melian seems parallel to and in the same island. For place-names in Greece derived from trees see Tozer, , Highlands of Turkey, ii. p. 107Google Scholar. Sibthorp, (Florae Graecae Prodromus, ii. p. 256Google Scholar) and Fiedler, (Reise, i. p. 539Google Scholar) say that Pistachia Terebinthus grows abundantly in the Greek islands. Both give the modern Greek name as Mr. Bickford Smith gives as the Cretan form.

^{page 65 note 1} Ath. Mitth. xx. p. 161. Taf. iv.

^{page 65 note 2} In position and perhaps in form it resembled the ‘house-altars’ of Pompeii. Cf. Overbeck, , Pompeii, p. 268Google Scholar, Presuhn, Die Neuesten Ausgrabungen, i. Taf. iv., vi. Taf. ii.

^{page 69 note 1} De Rossi, Musaici cristiani, xvii., xviii. A coloured paper cast and a coloured drawing by Zeri are exhibited in the South Kensington Museum.

^{page 69 note 2} A variant from the Vulgate of Isaiah v. 1. Bull, crist. 1879. Taf. ix., x. Kraus, , Gesch. der Christl. Kunst, i. p. 298.Google Scholar

^{page 70 note 1} Rev. Arch. iv. (1847), Pl 78, p. 661. Traces of a fish-panel were found in the same church. This juxtaposition of earth and sea, conventional in pre-Christian mosaic, and retained perhaps because to the Christian the fish as well as the vine had a mystic meaning, is seen in other early basilicas of North Africa, e.g. at Tipasa, Mélanges ď Arch, et ď Hist. tom. xiv., and at Sertei, , Mélanges G. B. de Bossi, p. 345.Google Scholar

^{page 70 note 2} Renan, , Mission de la Phénicic, Pl. xlix. p. 607.Google Scholar

^{page 70 note 3} Photograph in Quarterly Statement of Palestine Fund, 1894, p. 261. Cf. Mr. A. S. Murray's note, ib. 1895, p. 126.

^{page 70 note 4} To the fourth century too the British Museum authorities ascribe the Carthage pave ment of the Months. Its decorative design of cypress-like trees growing in vases and converging towards the centre is a very beautiful development of the older tree-patterns. Archaeologia, xxxviii. Pl. ix.-xiii. It was originally planned for a dome rather than a floor. Cf. Garrucci, , Arte cristiana, iv. Tav. 255.Google Scholar

^{page 71 note 1} The ceiling-mosaics in the side-chapels of S. George at Salonica imitate not only the painted coffers but also the cornice-mouldings of late classical architecture. Texier and Pullan, Pl. xxxiv.

^{page 71 note 2} Compare a tile-work design of vines, grapes and birds from ‘the south gate of the Tope Maidan, Teheran, 17th cent.’ reproduced in the Cross Gallery of the South Kensington Museum.

^{page 71 note 3} De la Blanchère, , Musée Alaoui, p. 25Google Scholar.

^{page 72 note 1} The idea of fish swimming in a circle is Used with equal effect, as Mr. Cecil Smith points out, on a series of red-figured plates from South Italy, among the latest examples of Graeco-Italian painted pottery, which were probably intended to be used as fish-plates at table. Brit. Mus. Vase Catal. F. 259 ff.

^{page 72 note 2} Found between Tipasa, and Cherchel, . Bulletin Archéol, de la Comité des Travaux Historiques, 1893Google Scholar, Pl. x.

^{page 72 note 3} Aelian, , N.A. xii. 43Google Scholar. M. Gauckler does not mention the passage. I am in many ways indebted to his full and interesting essays on the mosaics from Sousse, (Rev. Arch. 1897, (2), pp. 8–22Google Scholar; fishing-scene, Pl. xi.) and Oudna, (Monuments Piot, iii. pp. 177–229Google Scholar; fishing-scene, p. 198). In Plato's Sophist the art of fishing is resolved into and

^{page 73 note 1} Lvsons, Rel. Brit. Rom. I., Pt. ii., Pl. vii.

^{page 73 note 2} E.g. at Silchester, , Arehaeologia, lv. p. 241Google Scholar, and at Halicarnassus, Newton, Hist, Discoveries Pl. 41. The latter can hardly be earlier than the 4th century.

^{page 75 note 1} Clarae, Vol. IV. Pl. 695, Figs. 1614, 1615. The former, at Deepdene, = Michaelis, , Ancient Marbles in Great Britain, p. 280Google Scholar. For other instances of the type, see Roscher, , Myth. Lex., p. 1133Google Scholar (Dionysos in Art, by Thrämer). A pardalis sometimes takes the place of the nebris. In our figure the artist has compromised: the hoofs are cloven, but the mask is a panther's.

^{page 75 note 2} Cf. Furtwängler, , Masterpieces, p. 323, 5.Google Scholar

^{page 78 note 1} The worship of Athene, as the very archaic character of the xoanon on coin and column-relief shows, was much older than that of the Tyche of Melos. But the latter patriotic cult may have been established as early as the 4th century B.C., by the remnant of the old population whom Lysander sent back. Cf. the Tyche made by Praxiteles for Megara, and his ‘Bona Fortuna’ which was at Rome when Pliny wrote. The people of Antioch were doubtless following an established fashion when they set up a statue of the Fortune of their city early in the 3rd century. See Wolters' article in Ath. Mitth. xv. For the Melian type of Tyche bearing the infant Plutus, cf. the statue at Thebes, Paus. ix. 16, 1.

^{page 78 note 1} The inscriptions relating to these and other associations among the Greeks have been collected by Ziebarth, Erich, Das griechische Vereinswesen, Leipsic 1896Google Scholar. The facts which follow are drawn in the main from this work and from Foueart's Associations réligieuses.

^{page 78 note 2} Published by Wide, S., Ath. Mitth. 1894, p. 257Google Scholar, and by Maass, E., with fuller commentary, Orphcus, p. 18.Google Scholar

^{page 79 note 1} The twelve priests mentioned as conducting a Dionysiac festival in Patmos, in a passage quoted by Maass, , Orpheus, p. 52Google Scholar, from the Acts of John, were probably the officers of an association of μύσται or βάκχοι. This description by a hostile writer shows us a Bacchic society at its worst, just as the rules of the Iobacchi show one at its best.