¹ Προικόννησος, Constantinople, 1895. Practically the only new discovery recorded in the present paper is the Hellenic fortification above the village of Marmara.

² His notes are chiefly geological, as are Spratt's, in Quarterly Journal Geol. Soc. 1850, p. 218Google Scholar; Sibthorp, (in Walpole's Voyages, p. 48Google Scholar) has collected some notes on the natural history. An article by K. Zachariades in the first volume of Ξενοφάνης (pp. 404–412) contains little but educational statistics.

³ Cf. The Field, Sept. 16 and 23, 1905.

⁴ This involves a considerable but not uninteresting détour, as the steamers call at several Thracian ports going and coming. They leave twice a week at sunset, arriving at Marmara the following noon.

⁵ Etym. Mag. s.v. Προικόννησος

⁶ Sch. Ap. Rh. ii. 279.

⁷ Plin. v. 40, but Scylax (94) distinguishes them as two islands.

⁸ Plin. Neuris.

⁹ Theoph. cont. 437 B.

¹⁰ Theoph. cont. ad loc. (Προικόννησον) ἤν τινα Νεβρίαν ὠνομασμένην ἐκ τῆς κατὰ χρησμὸν δεδομένης προχόου ἄποικοι Σαμίων μετωνόμασαν οἷς ἀφικομένοις πρὸς νῆσον . . . ἀριστοποιουμένοις τε ἀπορία ἐτὑγχανεν ὁπόθεν ὕδωρ κομίσαιντο γυνὴ δέ τις ἔφη αὐτοῖς, εἰ ἔχετε πρόχοον δώσω ὑμῖν ὔδωρ λαβόντες, ώς δ χρησμός, καὶ γῆν ἐξῃτήσαντο, τῆς δὲ καὶ ταύτην δεδωκυίας Πρόχοον τὴν νῆσον ὠνόμασαν καὶ τοῖς ἀργυροῖς νομίσμχσιν πρόχοον εἰκόνιζον

¹¹ See B.M. Cat. Mysia, pp. 178–9, where they are assigned to a period c. 330–c. 280 B.C., i.e. after the Cyzicene conquest and deportation. An older coin (about 400 B.C.) is published by Wroth, in Num. Chron. 1904, 306Google Scholar, Pl. XV. 14. Some of them bear magistrates' names. The copper coins are fairly common in the island, very rare outside.

¹² Schol. ad Apoll. Rhod.; Etym. Mag.

¹³ I heard a similar modern legend to the effect that when God created Rumili and Anatoli there Avere certain stones left over which He ordered the Devils to carry from the latter to the former. The Devils dropped the stones half-way, whence the Marmara islands.

¹⁴ App., Bell. Mith. i. 75Google Scholar. Coronelli, (Isolario, 281Google Scholar) saya the islands were called Prichiones ‘mercè che gl' Imperatori Greci le davano per dote alle Sorelle ò Figlie che maritavano.’

¹⁵ Tomaschek, , Silzb. Akad. Wien 124 (1891), p. 3Google Scholar.

¹⁶ 245, Ducange.

¹⁷ Str. 587. Theoph. cont. (437 B) is probably in error as to the Samians. Etym. Mag. (s.v. Προικόννησος) speaks of Milesians in the same legend.

¹⁸ Hdt. iv. 138.

¹⁹ Hdt. vi. 33.

²⁰ Paus. viii. 46; cf. Demosth, . in Polycl. § 5Google Scholar, in Mid. § 173.

²¹ Theoph. cont. 196 (Russians in the reign of Theophilus); ibid. 299; Cedr. ii. 227 (Saracens from Crete in 866); G. Pachy. ii. 529 (Catalans in 1307).

²² Stephanus (son of Romanus Lacapenus), A.D. 945 (Cedr. ii. 325, Zon. iii. 481, Theoph. cont. 437, Leo Gram. 330, Sym. Mag. 753–4), and Basilius Petemos (Cedr. ii. 342), Theophano, A.D. 970 (Zon. iii. 521).

²³ The patriarchs, Nicephorus 815 (Cedr. ii. 56, Zon. iii. 325), Michael Cerularius 1058 (Scyl. 644), Arsenius 1258 (G. Pachy, i. 271—for the monastery of Suda see Gedeon, p. 12—cf. ii. 83. Niceph. Greg. i. 95), and of the saints noticed below: Nicolas, Macarras, John, Hilarion, Theodore, Stephanus, and Philetaerus; all but the last were exiled during the iconoclastic period. Hierocles' ἐξορία and Photius 82, Bekker, (cf. Vita Chrysostomi, lxxv. 22Google Scholar, Migne) show that this was the recognised use of the island. Philetaerus is said by the Synax. C'politanum to have been sent to the quarries.

²⁴ Under Leo Armenius (Migne, , P.G. cv. 912Google Scholar).

²⁵ Under Balbus, Michael; cf. Anal. Boll. xvi. 140 ffGoogle Scholar.

²⁶ c. 713.

²⁷ c. 845.

²⁸ S. Timotheus is said to have come to the islands under Justinian and to have, converted the inhabitants from their barbarous manner of life—they lived by plunder from wrecks and from boats which put in during stormy weather. The Life of Timotheus is discussed by Gedeon, who pronounces it most untrustworthy and even devoid of truth in local colour (pp. 120–123).

²⁹ c. 834. Migne, , P.G. cxvi. 669–72Google Scholar.

³⁰ Migne, , P.G. c. 1178Google Scholar.

³¹ May 19 in Acta SS. (under Maximian).

³² Ignatius (879) is the first archbishop in Gedeon's list.

³³ Gedeon, p. 14.

³⁴ Printed in the preface of Marmora's, AndreaHistoria di Corfu, 1672Google Scholar (Gk. and Lat.), and Dapper, , Archipel, p. 491Google Scholar (French). It is discredited by Hopf, (Veneto-Byzantinische Analekten in Silzber. K. Akad. zu Wien, 1860, xxxii, p. 508Google Scholar). Cf. Gedeon, p. 152. Finlay's copy of the Historia di Corfu has the following MS. note: ‘This is a forgery: the title proves it. It may hare been framed on some document of Manuel of Epirus, Emperor of Thessalonica, 1230–1232. The indiction would really be xii. ιβ′.’

³⁵ Villehardouin, § 245.

³⁶ Lequien iii. 945 (Marmorensis); cf. the thirteenth-century Provincial in Latrie, Mas, Trésor. A seventeenth-century Latin Mission to Marmara is mentioned by Carayon, (ed. Le grand, , p. 57Google Scholar).

^36a Pachy. vi. 25.

³⁷ de Gongora, L., Real Grandeza de la Republica di Genova (Madrid and Genoa, 1665–7)Google Scholar, Tit. viii. No. 22 (May 26, 1315): ‘Asia Minor cum Comitatibus, Baroniis, Civitatibus, Insulis adjacentibus videlicet Fenosio, Marmora, Tenedo,’ etc.

³⁸ Gedeon, p. 219. The Kadi was similarly under the Mollah of Galata (d'Ohsson, , Tableau, ii. 281Google Scholar) so that the islands were administered as a suburb of Constantinople. Hence Coronelli's reference to the Subashi, who was the local representative of the Voivode. The revenues of Marmara were sold for five purses (£350), those of Aphisia and Koútali for 400–600 dollars Pouocke). Coronelli, (Isolario, 281Google Scholar) records that the late Sultan Mahomet IV. gave the islands to his mother for the construction of the Yeni Validieh Mosque, and that the marbles were used for that building.

³⁹ § 62 …ab alio latere oppidum elevatur paucorum inhabilantium ubi scopuli aliqui praevalent Martelli (Add. MS. 15,760, f. 38 vso. c. 1489), gua vero meridiem occasumque spectat oppidum habet eiusdem. quo insula nomine, cetera deserta atque inculta iacent.

⁴⁰ Pérégrinations (1606), p. 373.

⁴¹ Viz. Cadicui (Galimi), Palastra, Gamialo, Antoni, Klazak-Arbanitochori (together inland), Parasleio, Marmora. Gamiato and Arbanitochori are now chiftliks.

⁴² ‘All the inhabitants Xtians, all Albaneses except Marmora, where there are Greekcs’ (Add. MS. 22,914, f. 29 v°); cf. Palerne, op. cit. p. 373: (at Palatia?) ‘Leur coiffeure les-femble à celle d'Esclauonie: elles portent leur robbe ouuerte par deuant où elles mettent un grand' pièce d'ouurage d'or, & de soye de toutes couleurs qu'elles font elles mesmes.’ Coronelli, (Isolano, p. 281Google Scholar) says: ‘(le Donne) essendo naturalmente bellissime, vestono panni soprafini, portano la sottoveste di seta, di sopra una Giubba con bottoni di Argento dorato, ed in testa un grosso involto rotondo di seta gialla, come Felpa guarnita d'oro, giugnendo l'orificio sino alle spalle.’ The costume of a woman from Marmara in (Anon., ) The Costume of Turkey (London, 1802, PL XIII.)Google Scholar has, however, nothing remarkable about it. The Cle. de Marcellus, (Souvenirs d'Orient, i. 167Google Scholar) wrote in 1820: ‘Les femmes de l'île de Marmara portent de longues robes à bandes de diverses couleurs, et une haute coiffure chargée de fleurs et de mousseline blanche.’ Shalvars are now practically universal as far as I saw. A custom which may have a bearing on the race-question is noted by Coronelli (loc. cit): ‘Altro costume ritingono quest' Isolani … che pagando certa somma di denaro, il Subascì permette à Sposi di converrai liberamente, e di poter ancora dormire con le Spose prima di conjungersi sacramentalmente insieme; restando però gravide sarebbero quelli condannati al remo.’ This is also reported of the ‘Pistic’ villagers (in the Brusa district), who appear to be of North-Greek origin.

⁴³ The island is a nahié under Erdek: the smaller islands form a second nahié with a mudir at Pasha Liman.

⁴⁴ In Covel's day Marmara village was known also as Πὑργος (Add. MS. 22,914, f. 29 v.).

⁴⁵ The place is known as Καστράκια: it is immediately above the ‘chapel’ (of S. Elias) shown on the Admiralty Chart 2242 [955].

⁴⁶ Scylax, § 94, νῆσος κοὶ πὁλις Προκόννησος

⁴⁷ i. 288; cf. 286 and Gedeon 128. The harbour to the north (Petali Liman) is mentioned by the Russian abbot Daniel (1105) as a stage between Constantinople and Jerusalem (ed. Noroff, pp. 5, 6).

⁴⁸ Texier, , Descr. de l'Asie Mineure, ii. 167Google Scholar; Gedeon, 155.

⁴⁹ Admiralty Chart 2242 [955]. The height of the hill is 1340 feet.

⁵⁰ The site is marked on Pococke's map. The other chiftlik, Tetrágono, seems also to have been a village down to 1766 (Gedeon, pp. 155, 219). It is probably the ᾿Αρβανιτοχώρι of Covel's map: the site is an ideal one for Albanian shepherds.

⁵¹ Pl. XLIII, in the small edition (L'Univers, xii.), where it is called ‘Restes du, Palais de Justinten’: this he does not explain or justify, saying only that ‘le monument antique de Palatia se compose de deux épaisses murailles construites en moellons de marbre avec plusieurs ranges de briques intercalées. Une seule fenêtre dont le cintre est en brique existe encore’ (ii. 159). Schweigger (1576, in Feyerabend, 's Reyssbuch, ii. 92Google Scholar) has the following curious note: ‘In Proeconniso ist vor Zeiteu ein schön Amphitheatrum, Schauhauss oder Spielhauss von lauter Marmor gewesen eines aus der sieben Wunderwercken der Welt’—presumably a con fusion with Cyzicus.

⁵² They were formerly worked by corvée, the rayah quarrymen receiving certain privileges from the government (including the right to wear white turbans) in return (La Motraye, A., Voyages, i. 472Google Scholar; Dallaway, J., Constantinople, 368Google Scholar; cf. Sandys, 27). This probably dates from the grant of Mahomet IV. White turbans were worn as a distinction by the inhabitants of the Mastic villages in Chios, another imperial appanage.

⁵³ Texier (ii. 158) says that the greater part of the island is composed of it: ‘Le calcaire marbre cristallin, d'un blanc éclatant, constitue la majeure partie de l'île. Le granit gris ne se montre que dans la montagne qui domine le port de Galimi.’

⁵⁴ Spratt, however (loc. cit. p. 218), describes it as ‘in some parts quite as white, pure, and crystalline as the Parian, but generally more resembling the cippoline (sic) of Carysto.’

⁵⁵ The quarries are mentioned in the second of the ‘Letters of Brutus’ to the Cyzicenes, and in the Cod. Theodos. xi. 28Google Scholar.

⁵⁶ Str. 589.

⁵⁷ Vitr. ii. 8, cf. x. 7.

⁵⁸ Phot., Bibl. 229, ed. Bekker, Google Scholar.

⁵⁹ Zos. ii. 30; Theoph. cont. 141, 145, 146, 147; Evagrius, , Hist. Eccles. iii. 28Google Scholar, cf. Migue, , Patr. Gr. xxxv. 281Google Scholar; Paul. Silent., 576, 606, 664 (S. Sophia); Const. Rhodius, 122 (colonnade of Burnt Column), 670 (floor of Holy Apostles); cf. Lethaby, and Swainson, , St. Sophia, p. 237Google Scholar; Strzygowski, , Wasserbehälter, p. 255Google Scholar, who traces the marble by masons’ marks to Ravenna, , Proconnesian marble is also mentioned as the material of a cultus statue in Rome in ‘Passio IV. Coronatorum’ (Berl. Sitzb. 1896, pp. 1299, 1302)Google Scholar, and of an imaginary temple in Zosimus, Περὶ ᾿Αρετῆς (in Berthelot, , Alchimistes Grecs, ii. 111Google Scholar). In Const. Rhod. Proconnesian and (coloured) Cyzicene marble are distinguished, the latter being probably the red-veined variety formerly quarried at S. Simeon near Artaki on the Cyzicene peninsula.

⁶⁰ Sandys, 27.

^60a Coronelli, , Isolario, 281Google Scholar.

⁶¹ La Motraye, i. 472.

⁶² Smith, Thos., Notilia Cp. p. 118Google Scholar; Hobhouse, 819.

⁶³ Codinns, Περὶ τῶν Ταφῶν passim; Const. Porph., de Caerim. i. 643 ffGoogle Scholar. The funerary inscriptions C.I.G. 3268, 3282, 3386 (Smyrna), and Dumont, , Inscrr. de la Thrace, 70Google Scholar, mention Proconnesian sarcophagi.

⁶⁴ Connetted by Gedeon (p. 112) with the monastery mentioned by Pachy. i. 271 [᾿Αρσέ νιον] περιώριζον ἀνὰ τὴν Προικόννησον τψ῀ ἐκεῖσε μοννδρίῳ τῷ ἄνω τῆς ἐγχωρίως λεγομένης Σούδας κειμένψ ἐγκατακλείσαντες

⁶⁵ Also found at ‘S. Elias,’ Salonica (Texier and Pullan, Pl. LII.); Demetrius, S. on Ossa (Mezières in Arch. des Miss. III. (1854), 225Google Scholar; Tozer, , Highlands of Turkey, ii. 75Google Scholar); Titus, S. at Gortyna (Fyfe, Archil. Rev. xxii. (1907), p. 60)Google Scholar; Hagioi Saranda, near Sparta; Nea Moni in Chios; and a small church within the walls at Aegosthena.

⁶⁶ Cf. above, p. 10. I have, however, been told by a Kutalian that it survives also to a limited extent at Prastos.

⁶⁷ Dallaway, 397, records that Klazaki turned Turk to escape paying kharalch, but was rewarded by a doubling of the tax: this would account for the dwindling of the village, but it is fair to say the incident was unknown to the villagers of Aphloni.

⁶⁸ Pp. 120–3.

⁶⁹ The well is in a covered chamber, in the roof of which is a hole where flour was miraculously ground and produced for the saint.

⁷⁰ No particular tree.

⁷¹ It contains a picture of the Virgin, which is regarded as un oracle for the success of ventures. Three crosses are made on it with a coin, after which, if the coin stisks (is accepted), the omen is propitious. This method of divination is common in other parts, e.g. near Artaki on the Cyzicene peninsula (Sestini, , Penisola di Cizico, i. 26)Google Scholar, and has analogies in ancient practice.

⁷² Sathas, , Μϵσαιωνικὴ Βιβλιοθήκη, iii. 604Google Scholar.

⁷³ From Covel, MS. Add. 22,912, f. 160 vso. For the superstition in question see Politis' Παραδόσεις

⁷⁴ Sir Daniel Harvey, Ambassador at the Porte, 1672–81. The other initials I do not recognise.

⁷⁵ Moryson, Fynes, Itinerary, ii. 88Google Scholar.

⁷⁶ v. 40.

⁷⁷ ii. 585. Uzzano, mentions the island as Andanun, p. 226Google Scholar.

⁷⁸ Syuaxarion, Aug. 21. Bassa was martyred under Maximian; it is, however, uncertain whether she died in Halone or at Cyzicus: according to local legend her remains were washed ashore at Halone. Her ἁγίασμα is still to be seen there (Gedeon, p. 37).

⁷⁹ For at least two centuries (Gedeon, p. 194).

⁸⁰ Pérégrinations (1606), p. 372: ‘Alloina a este autres foys deshabitée jusques à ce qu'elle fut à un certain Bashat rené Albanois de Nation qui l'envoya peupler des gens de son pays & sont tous Chrétiens.’ This was written in 1582. Gedeon (55) remarks that the names in Halone are curious and foreign-sounding. The island is sometimes called Βούργαρα from a supposed Bulgarian colony. Covel gives the names Βουλγάρι Arnoutcui, indifferently to the village of Pasha Liman.

⁸¹ He was exiled to Proconnesus, founded a monastery of S. Anna, and lived in a cave called Κισσοῦδα: see Vitas. Stephani junioris (Migne, , Pair. Grec. c. p. 1178)Google Scholar. The effigy of S. Stephen appears ou a Byzantine bishop's seal (Schlumberger, , Sigillographie, 199, p. 732Google Scholar).

⁸² A similar house at Agrapha is described by Woodhouse, (Aetolia, 38Google Scholar) and there is another above Volo. The type is a combination of the ‘Pyrgos’ with the normal Turkish timber-framed house.

⁸³ Zachariades (p. 405) describes it as λίαν θαυματουργός adding that many pilgrims flock to it at the panegyris from the islands and the Kapu Dagh, θεραπευὁμενοι διὰ τῶν αυνηθῶν τρόπων, ξυλοχοπημάτων δηλαδὴ καὶ τῶν λοιπῶν

⁸⁴ Squeezes of this have been sent to the editors of the Corpus.

⁸⁵ Le. Bruyn, 67. It no longer exists.

⁸⁶ 94.

⁸⁷ Ap. Steph. Byz. s.v. Βέσβικος

⁸⁸ Anal. Boll. xvi. 159: διὰ τὸ ἀπεῖναι τὴν κατευθείαν, ὡς οἶμαι, οὔτω καλουμένη(!)

⁸⁹ Theoph. i. 774 (Sons of Constantine VI., 812 A.D.). Acta SS. and Synaxaria, Mar. 28 (Hilarion of Dalmaton, c. 787; for his ἀγίασμα cf. Gedeon, p. 73); Apr. 1 (Macarius of Peiecete, c. 829); Apr. 17 (John of Kathara, c. 713); Dec. 26 (Theodore Graptos; cf. Migne, , Patr. Gr. cxvi. 669–672Google Scholar).

⁹⁰ Two villages were seen by Palerne in 1582 (Pérégrinations, p. 371).

⁹¹ Gedeon, p. 63: Αἱ δὲ γυναῖκες τοῦ χωρίου κυρίως αἱ μητέρες, ἀδελφαί καὶ σύζυγοι τῶν ἀποδημούντων ναυτῶν, πᾶσαι πιστεύουσιν τοὐλάχιστον ἐπίστευον ἄλλιτέ ποτε—ὅτι δῆθεν εἴπερ ὸ ναυτιλλόμενος ἔμενε μακρὰν τῆς πατρίδος ὑπ᾿ ἀνέμων ἀντιξόων κωλυόμενος νὰ ἐπιστρέψῃ ἤρκει νὰ θυμιάσωσιν αὖται περιφερόμεναι περὶ τὰ ἐρείπια ταῦτα τρίς, ὔπως μεταβάλωσι τὴν διεύ θυνσιν τῶν ἀνέμων καὶ εὐκολὑνωσι τὴν ἐπάνοδον

⁹² Κουτάλι=spoon: the sky-line of the island, with the high conical hill of S. Elias at one end, suggests the image.

⁹³ Cf. Niceph. Greg. lxxxviii.; Cantac 1. 251, 313.

⁹⁴ Gedeon, p. 83.

⁹⁵ Ib. p. 79.

⁹⁶ P. 67.

⁹⁷ ii. ch. 22.

⁹⁸ v. 40.

⁹⁹ Cf. Steph. Byz. s.v. Besbicus.

¹⁰⁰ Cf. Steph. Byz. s.v.

¹⁰¹ Of these only S. Andreas is cultivated (Malkotzes, in Ξενοφάνης i. p. 255), and none are inhabited. Sathas, (Μεσ. Βιβλ iii. 565Google Scholar) cites a Sigillion of 1626: περὶ τοῦ ἐν τῇ νήσῳ ἀντικρὺ Μηχανιῶνος μονυδρίου τοῦ ῾Αγίου ᾿Ανδρέου ὑπὸ ᾿Ιωάννου Μούρμουρα κτισθέντος. The island was, however, uninhabited at the end of the century (Coronelli, , Isolarlo, 281Google Scholar).

¹⁰² P. 317.

¹⁰³ Gedeon, 54, etc.

¹⁹⁴ Dapper, 491.

¹⁰⁵ Cf. Gedeon, pp. 23, 52, 54, 65, 129, 130. Cf. Karydones, S., Τὰ ἐν Καλλονῆ Μοναστήρια p. 25Google Scholar.

¹⁰⁶ It is curious to remark that the two pre eminently sea-going villages (Aphtoni and Koútali) preserve the most distinct traces of their Albanian origin—a fresh example of the adapt ability of that race to sea-faring life exemplified in the history of Hydra and Spetsa.