¹ The fieldwork for this article was undertaken during the tenure of the Annual Scholarship (1958–59) and Fellowship (1960) of the British Institute of Archaeology at Ankara, and the University of Edinburgh's Tweedie Fellowship; the opportunity to prepare it for publication was generously afforded by Lincoln College, Oxford. My chief acknowledgements for advice and criticism on many points of detail are due to Professor D. Talbot Rice, Mr. M. R. E. Gough, Mr. J. B. Ward Perkins and Dr. M. H. Ballance. The photographs are the work of my wife, who accompanied me throughout. Preliminary reports appeared in AS. X (1960), pp. 26–8Google Scholar; XI (1061), pp. 6–7, and ILN., 20th August, 1960.

² XIV, 3. For Lycian topography see Magie, D., Roman Rule in Asia Minor (Princeton, 1950), I, pp. 516 ffGoogle Scholar, and II, pp. 1370 ff. The steepness of the mountains may be gauged from the fact that in July 1959 snow for iced sherbet was regularly brought down from Alaca Dağ to Demre by camel and donkey in a single day. (This trade ended in 1960, when manufactured ice from Finike came with the new road.)

³ We only encountered one case of malaria, at Dereağzı in 1959.

⁴ Much of de Planhol's, XavierDe la Plaine Pamphylienne aux Lacs Pisidiens, Nomadisme et Vie paysanne (Paris, 1958Google Scholar) is relevant, mutatis mutandis, to Lycia, to which it does contain a few general references.

⁵ XIV, 3.

⁶ Broughton, T. R. S., Roman Asia (An Economic Survey of Ancient Rome, IV, Baltimore, 1938), Pt. 2, Chap. I and pp. 728–9, 779–84Google Scholar; Jones, A. H. M., Cities of the Eastern Roman Provinces (Oxford, 1937), pp. 96–110Google Scholar; W. Ruge, in RE. (s.v. Lykia); D. Magie, loc. cit. For the inscriptions, TAM. (Tituli Asiae Minoris, Vienna, 1901–1944Google Scholar).

⁷ For Lycian cedar see Pliny, , HN. XII, 132Google Scholar; XIII, 52; XVI, 137.

⁸ Attested by, e.g., the Peutinger Table; more important perhaps was the inland road from Limyra to Oenoanda, Cibyra and, ultimately, Laodiceia. Scylax, (Periplous, 100Google Scholar) says that to traverse Lycia by land takes twice as long as by sea. Mellaart's, remarks (AS. IV (1954), p. 176Google Scholar) on this coastline, to explain apparent absence of Bronze Age settlement, must be modified; Lycia's anchorages outweighed the dangers of her reefs and cliffs, at least in classical and post-classical times.

⁹ Vell. Pat., II, 102.

¹⁰ Act. Apost., XXI, 1, and XXVII, 5–6.Google Scholar The apocryphal Acta S. Pauli, written towards the end of the second century, tell of St. Paul's conversion of Thecla in Iconium, and how she followed him first to Pisidian Antioch and then to Myra, where he worked several miracles before returning to Sidon; Schmidt, C., ΠΑΥΛΟΥ, Acta Pauli, nach dem Papyrus der Hamburger Staats- und Universitäts-Bibliothek (1936)Google Scholar. Ramsay argued that these Acts were founded upon historical sources, but the details cannot be pressed. A fifth-century Life of St. Thecla (Basil. Seleuc., I; Migne, , PG. LXXXV, 550–2Google Scholar) confirms that Myra's usual access was by sea, for Thecla's journey from Pisidian Antioch to Myra is said to have been both by land and by sea.

¹¹ As elsewhere, synagogues may have provided the first foothold; for Jews in Lycia, with particular mention of Phaselis, see I Maccabees, XV, 23Google Scholar (c. 140 B.C). For Early Christian Lycia in general, see Schultze, V., Altchristliche Städte und Landschaften II, Kleinasien (1922–1926), pp. 188–209Google Scholar.

¹² Migne, , PG. XVIII, 9–408Google Scholar.

¹³ Inscription from Arycanda, ; TAM. II, 785Google Scholar; Grégoire, H., Recueil des Inscriptions grecques chrétiennes d'Asie Mineure, I (Paris, 1922), 282Google Scholar.

¹⁴ Anrich, G., Hagios Nikolaos, 2 vols. (Teubner, 1913–1917Google Scholar); Halkin, Fr., Bibliotheca Hagiographica Graeca (Brussels, 1957), II, p. 139Google Scholar. For Eudemus of Patara, Mansi, II, 695.

¹⁵ Mansi, III, 322.

¹⁶ Ep. 218; Migne, , PG. XXXII, 809Google Scholar.

¹⁷ Mansi, III, 570.

¹⁸ Malalas, , Chronographia, XIV (ed. Dindorf, , p. 365Google Scholar).

¹⁹ Mansi, IV, 1219–26.

²⁰ Le Quien, , Oriens Christianus, I, 968Google Scholar.

²¹ Mansi, VI, 565–80.

²² Mansi, VII, 576–80; twenty-one bishops signed, sixteen in person and five (each manibus dolente) by proxy.

²³ Le Quien, I, 989 (quoting Cedrenus).

²⁴ Cf. Ostrogorsky, G., History of the Byzantine State (transl. Hussey, J. M., Oxford, 1956), p. 71Google Scholar. The conciliatory gesture was the condemnation of the Three Chapters, the supposedly Nestorian writings of Theodore of Mopsuestia, Theodoret of Cyrrhus and Ibas of Edessa; Mansi, IX, 376.

²⁵ Mansi, IX, 389–96.

²⁶ Literary evidence is the Vita Nicolai Sionitae (see below, note 30), passim; also of the sixth century is Moschus, Pratum Spirituale, 135 (Migne, PG. LXXXVII), a cautionary-tale about a Lycian convent of forty nuns. The abundant archaeological evidence is the main subject of this paper.

²⁷ See below, pp. 148 ff.

²⁸ Procopius, de Aed., makes no mention of Lycia.

²⁹ For the church of SS. Priscus and Nicholas, on the water's-edge of the Golden Horn, see ibid., I, 6.

³⁰ 8. This work (henceforward cited as VNS.) is published in Anrich, G., Hagios Nikolaos (Teubner, 1913–1917Google Scholar) and is discussed by Robert, L., Hellenika, X (1955), pp. 197–208Google Scholar.

³¹ VNS. 8–9, 27–38, 39.

³² ibid., 23–4, 20.

³³ ibid., 15–19; Robert, loc. cit., pp. 197–9. Nicholas also cultivates vines (VSN. 63); extensive vine-terracing was found around Alakilise (cf. also Sir John Mandeville, 1322, for wine from Myra).

³⁴ Robert, loc. cit., pp. 204–5, has a list of thirty-five place-names.

³⁵ VNS. 4.

³⁶ ibid., 52–3. Probably the great plague of 542–3; on this see Bury, J. B., History of the Later Roman Empire (reprinted New York, 1958), II, pp. 62–6Google Scholar. Philip of Myra was one of Lycia's four signatories at Constantinople (553).

³⁷ Ostrogorsky, op. cit., p. 104.

³⁸ Theophanes, , Chronographia (ed. de Boor, ), p. 345Google Scholar. The Arabs had come to Finike for timber. For Lycia in this and the following period, see Tomaschek, W., “Zur historischen Topographie von Kleinasien im Mittelalter,” Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften, Wien, CXXIV, 8 (1891), pp. 43–53Google Scholar.

³⁹ Ostrogorsky, op. cit., pp. 86 ff.

⁴⁰ ibid., p. 119.

⁴¹ ibid., p. 140.

⁴² Theophanes, p. 424.

⁴³ ibid., p. 385.

⁴⁴ ibid., p. 465.

⁴⁵ ibid., p. 483.

⁴⁶ Monachus, Georgius, De Michaele et Theodora, 4 (ed. Bekker, , p. 814Google Scholar).

⁴⁷ Ostrogorsky, op. cit., p. 155.

⁴⁸ Le Quien, I, 969.

⁴⁹ ibid.

⁵⁰ Schlumberger, G., L'Épopée Byzantine (Paris, 1905), III, pp. 188 ff.Google Scholar, quoting Scylitzes; Cedrenus, Georgius, Historiarum Compendium (ed. Niebuhr, ), p. 511Google Scholar.

⁵¹ ibid., p. 512; cf. an inscription at Myra recording restoration by Constantine IX Monomachus and Zoe, dated 1042–3; Rott, H., Kleinasiatische Denkmäler aus Pisidien, Pamphylien, Kappadokien und Lykien (Leipzig, 1908), p. 340Google Scholar; CIG. VIII, 107Google Scholar; Grégoire, op. cit. (note 13), 291.

⁵² Anrich, op. cit. (note 14), pp. 434–49.

⁵³ Brownlow, W. R. B., Saewulf (London, 1892Google Scholar). Which Alexandria? In either case the reference is probably to Arabs, not Turks.

⁵⁴ Planhol, op. cit. (note 4), p. 87; Runciman, S., History of the Crusades, II (Cambridge, 1952), pp. 264 ffGoogle Scholar.

⁵⁵ Planhol, op. cit., p. 86.

⁵⁶ ibid., p. 85.

⁵⁷ Lloyd, Seton and Rice, D. S., Alanya (Alā'iyya), London, 1958, p. 3Google Scholar.

⁵⁸ Gesta Ricardi; cf. Tomaschek, op. cit. (note 38), p. 47.

⁵⁹ Le Quien, I, 970.

⁶⁰ Tomaschek, op. cit., pp. 46 and 49.

⁶¹ e.g. at Finike, Myra, Muskar, Dereağzı (all embodying classical masonry) and Karabel.

⁶² Cf. Wächter, A., Der Verfall des Griechentums in Kleinasien im XIV Jahrhundert (Leipzig, 1903), p. 34Google Scholar, for Myra's prosperity in the fourteenth century and virtual eclipse in the fifteenth; also Beck, H. G., Kirche und theologische Literatur im Byzantinischen Reich (Handbuch, XII, 2, 1, Munich, 1959). pp. 170–1Google Scholar.

⁶³ Tomaschek, op. cit., p. 49.

⁶⁴ ibid., p. 47 (Otto Heinrich Pfalzgraf bei Rhein).

⁶⁵ Professor A. W. Lawrence, who recently visited Kekova, assures me that the castle should be dated, on grounds of style, in the last quarter of the fifteenth or beginning of the sixteenth century.

⁶⁶ Perhaps by the time of da Uzzano, Antonio (1442), whose Della Mercatura (Lisbon/Lucca, 1776Google Scholar) mentions Myra, with the words: “… del cavo Santo Nicolao della Stamita …” (pp. 233–4)Google Scholar.

⁶⁷ Cf. a document dated 1620, Rotṭ, op. cit. (note 51), pp. 295–7. Also, Satha, K. N., Mes. Bibl. III (1872), p. 577Google Scholar.

⁶⁸ Beaufort, Francis, Karamania (London, 1817Google Scholar). Off Lycia he met Cockerell; ibid., p. 106; cf. C. Cockerell, R., Travels in Southern Europe and the Levant, 1810–1817 (ed. Cockerell, S. P., London, 1903), pp. 164 ffGoogle Scholar. See also Demargne, P., Fouilles de Xanthos, I, pp. 15 ffGoogle Scholar.

⁶⁹ Texier, C., Description de l'Asie Mineure, III (1849), pp. 202, 232Google Scholar, and Pl. 205 (Dereağzı); pp. 205, 238, and Pl. 222 (Myra). For the restoration at Myra, see Rott, op. cit., p. 326.

⁷⁰ e.g. Travels and Researches in Asia Minor (London, 1852), p. 366Google Scholar: “…the province of Lycia which has never been corrupted by the Roman or Christian styles….”

⁷¹ Spratt, T. A. B. and Forbes, E., Travels in Lycia, Milyas and the Cibyratis (London, 1847Google Scholar).

⁷² “A tour of Lycia by Mr. D. E. Colnaghi, 1854,” appendix to Newton, C. R., Travels and Discoveries in the Levant (London, 1865Google Scholar).

⁷³ Benndorf, O. and Niemann, G., Reisen in Lykien und Karien (Reisen in südwestlichen Kleinasien, I, Vienna, 1884Google Scholar); Petersen, E. and von Luschan, F., Reisen in Lykien, Milyas und Kibyratis (Reisen in südwestlichen Kleinasien, II, Vienna, 1889Google Scholar). For Dereağzı, , Reisen, I, p. 131 and Pl. 38Google Scholar.

⁷⁴ Reisen, II, pp. 38–40Google Scholar, Figs. 26–8.

⁷⁵ Rott, op. cit. (note 51), pp. 74–80 (K. Michel) and 295–344.

⁷⁶ op. cit. (note 13).

⁷⁷ The tradition may be exemplified by the Lycian tombs and the Hadrianic granary at Andriake, (Reisen, II, p. 41Google Scholar).

⁷⁸ Reisen, II, Fig. 27, and Rott, op. cit., Pl. facing p. 16 and Pls. 119–21 (Alakilise).

⁷⁹ See above, note 26.

⁸⁰ The survey was generously authorized by the Turkish Department of Antiquities, to which I wish to record my gratitude for the necessary permit. Turkish hospitality is proverbial and every assistance was afforded us at every level.

⁸¹ Reisen, II, pp. 223–4Google Scholar; Rott op. cit., pp. 76–7.

⁸² Reisen, II, pp. 38–40Google Scholar; Rott, op. cit., pp. 318–24 (Aladjakisle). A church of this name is ascribed to Lycia by Texier, C. and Pullan, R. Popplewell, Byzantine Architecture (London, 1864), p. 168Google Scholar). The description, however, is a curious hybrid, derived from two other churches, viz. Alahan in Isauria and the citadel-church at Alanya; Lycia (for Cilicia) was yet a further confusion.

⁸³ Cf. St. John of the Studion; van Millingen, A., Byzantine Churches in Constantinople (London, 1912), pp. 35 ffGoogle Scholar.

⁸⁴ Cf. Reisen, II, Fig. 27; Rott, op. cit., Pl. 119.

⁸⁵ No evidence was found for a central dome (conjectured ibid., p. 320).

⁸⁶ See below, p. 146.

⁸⁷ Líntels and pseudo-capitals project slightly from the plane of the masonry; plaster provided a flush surface.

⁸⁸ See below, p. 146. Cf. Rott, op. cit., p. 323 and Pl. facing p. 16; Kautzsch, R., Kapitellstudien (Berlin/Leipzig, 1936), p. 152Google Scholar.

⁸⁹ Noted in Reisen, II, p. 40Google Scholar.

⁹⁰ First reported in AS., X (1960), p. 28Google Scholar, and ILN., loc. cit. (note 1).

⁹¹ Cf. Figs. 11 and 14; this point was made to me by Mr. G. U. S. Corbett.

⁹² Rott, op. cit., p. 316; eine dreischiffige, südöstlich orientierte Basilika mit einem Querschiff, dessen nördlicher Arm in einer Apside endigte. There is now no visible trace of this transept, but the absence of pilaster-responds flanking the apse corroborates Rott's observation.

⁹³ See below, p. 146.

⁹⁴ See below, note 138, for key to photograph.

⁹⁵ South of this fortress (Byzantine masonry with tile-bands, on foundations of classical ashlar) is a terrace with extensive remains of massive masonry, including Doric column-drums; fragments of Byzantine vine-scroll carving suggest an intrusive church or chapel.

⁹⁶ This is an ancient road (see Fig. 2 and Pl. XLVc), rock-cut and revetted (average width 2 m.), traced from Gödeme through Muskar and Karabel to Çağman; cf. ILN., loc. cit. Other similar roads observed were Muskar to Alakilise and Alakilise direct to Koṣkerler (the latter steep, and significantly qualified locally as a boṣ merkep yol; above it, in the cliff-face, are rock-cut cells, perhaps monastic).

⁹⁷ For first reports of this monastery, see note 90.

⁹⁸ A typical block in the south semi-dome was 1·55 m. long, 0·77 m. high; the blocks are not dressed on the inner face.

⁹⁹ This formula occurs at Aphrodisias (Grégoire, op. cit. (note 13), 265), and at Synnada, (MAMA. IV, 99Google Scholar; M. H. Ballance informs me that in 1955 he found it too on a lintel, ibid., 98, unnoticed by the editors). Ζωή alone is recorded at Termessus (Grégoire, 320).

¹⁰⁰ The motif recurs in painted plaster at Alacahisar (see p. 136), but without the relief-line.

¹⁰¹ In the southern and eastern apses of the triconchos the window-pillar has mouldings on only three sides, the outside being left plain. The pillar of the northern window, however, has mouldings on all four sides, showing that this window looked into the baptistery, rather than out over it.

¹⁰² The letters appear to have been cut with a running-drill. μέσατος is equivocal. If a noun, is this Nicholas' father? For Μέσατος as a proper-name see Lesky, A., “Die Datierung der Hiketiden und der Tragiker Mesatos,” Hermes (1954), pp. 1 ff.Google Scholar, and Professor Lloyd-Jones, H., Appendix to Loeb ed. of Aeschylus II (1957), pp. 595 ff.Google Scholar; add Wessely, C., Studien zur Paläographie und Papyruskunde, iii, viii (1904–1908), 402Google Scholar, where μέσατος is more easily taken as a proper-name than as a noun = “arbitrator”. (There is thus no evidence for μέσατος in this latter sense, as claimed by LSJ., s.v.) If, however, an adjective, is it an ethnic (a possibility kindly suggested by L. Robert)? Or is it equivalent to μέσος, qualifying Nicholas' rank? This last use is mainly poetical (cf. LSJ.,s.v.), but is perhaps the most acceptable in the present context: a ship could have several ναύκληροι, who would presumably be distinguished (cf. Bass, G. in Arch. Anzeiger, 1962, 563Google Scholar). For ναύκληροι in the Early Byzantine period, see especially the Rhodian Sea Law (ed. Ashburner, W., Oxford, 1909Google Scholar) and Lopez, R., “The Role of Trade in the Economic Readjustment of Byzantium in the Seventh Century,” Dumbarton Oaks Papers, XIII (1959), pp. 69–85Google Scholar. (I am indebted to Professor Lloyd-Jones, Professor Bean and Miss J. M. Reynolds for help with this inscription.)

¹⁰³ Projection of console 0·50 m., height 0·28 m., width of top surface 0·30 m.

¹⁰⁴ Either from the end (the cross has a tiny hole at its centre, as if for a pin) or from beneath (the acanthus leaf has a back-curling tip).

¹⁰⁵ On the west side of the northern jamb, the following graffito (letters 0·03 m.): διακόν∣[ο]ν καὶ μον∣[α]χοῦ.

¹⁰⁶ This enceinte (apparently erected hastily in Early Byzantine times) is double, a few metres separating the two walls; the huge quoins (some 2 m. long) are roughly squared, other blocks are shapeless and stand c. 5 m. high without mortar. At a higher level, south of the summit, a carefully-built gateway leads into a vestibule, which has doorways to left and right; the construction is curious, each end of the gateway-lintel underlying the end of a doorway lintel.

¹⁰⁷ For first reports, see note 90.

¹⁰⁸ Confirmed by beam-holes for the upper floor.

¹⁰⁹ The lintel is 2·38 m. long, 0·70 m. wide, 0·68 m. high. Dr. M. H. Ballance detected in the photograph the faint traces of an inscription (unnoticed by us on the site), which can confidently be restored:

Ἡ πύλη τοῦ Κυ^. δίον· δίκαιοι] ε [ἰσ][ται ἐν αὐτῇ.

This is a quotation from Psalm CXVIII, 20, and occurs often on church-doorways; cf. Rott, op. cit., p. 322 (and note), and Grégoire, op. cit., p. 92, for an example at Alakilise.

¹¹⁰ Information in a letter, October 1960.

¹¹¹ Mostly arched window-heads with rich carving, apparently considerably later than the sixth-century carving at Alakilise, and Karabel, . Cf. Reisen, II, p. 40Google Scholar.

¹¹² Cf. K. Michel in Rott, op. cit., p. 75 (no measurements or illustrations). I am puzzled by discrepancies, which I only noticed subsequently and have had no opportunity to check, between his account and mine.

¹¹³ See previous note.

¹¹⁴ Also, Spratt and Forbes, op. cit., p. 105; Reisen, I, p. 131Google Scholar and Pl. 38. Particularly relevant parallels are the churches of St. Clement at Ankara and the Koimesis at Nicaea.

¹¹⁵ cf. Spratt and Forbes, op. cit., p. 123.

¹¹⁶ This site overlooks a small rocky island; cf. Saewulf (between Myra and Finike): ‘venimus ad insulam que Xindacopo vocatur, quod latine interpretatur Sexaginta remule.” Tomaschek (op. cit., p. 40,) restored [pen]insulam.

¹¹⁷ Also, Spratt and Forbes, op. cit., pp. 133–4.

¹¹⁸ cf. Rott, op. cit., p. 338. Two figures of Christ in dark blue administering the Communion to two groups of Apostles, six to right (led by St. Peter), six to left (led by St. Paul); the two hindmost appear to be washing their hands; deep blue background, light green floor. On stylistic grounds a date after 1300 seems appropriate. On the east wall of the exonarthex, south of the main doorway, the standing figure of the prophet Habakkuk (Ambakoum), with yellow halo, white robes and folds delineated by sharp red lines, was recognized by an inscription painted in red letters on a white scroll held in front of the body: + Κ(ύρι)ε, ἠσα∣κείκοα τ [ὴν] ∣ ἀκωήν σου ∣ καὶ ἐφωβ[? ή∣θην,] κατενόη∣σα τά ἔργα σου ∣ καὶ . cf. Septuagint, Habakkuk III, 2. An abbreviated quotation characterizes Habakkuk in the church of the Theotokos Pammakaristos, Constantinople; cf. van Millingen, op. cit. (note 83), Pl. 41. Width of scroll 0·16 m., length 0·40 m.; height of figure c. 1·50 m.

¹¹⁹ cf. Rott, op. cit., p. 75.

¹²⁰ ibid., p. 76.

¹²¹ There were four figured panels in each of the two standing bays. East bay—north side, lower: standing group of elderly men (Pl. XLVIIb), with children, to right of damaged central male figure; north side, upper: Virgin crowned by two flanking Angels, Magi on left, two women on right, one standing, the other (named ἡ Σαλόμ[η]) kneeling; south side, lower: destroyed; south side, upper: haloed Virgin (in dark clothing) and Child, bearded man (in white clothing) standing on right. West bay—northern panels: destroyed; south side, lower: female figure on left in front of building, approached by stooping haloed man from right (Annunciation ?); south side, upper: Crucifixion, with two kneeling figures. The colours are faded, but red, blue and white predominate, on a light ground; faces are painted red-on-white.

¹²² Rott, op. cit., p. 342.

¹²³ Beaufort, , Karamania, p. 20Google Scholar, mentions “a small Christian chapel in a cove, called Xera” at the western end of the island, presumably not this church.

¹²⁴ M. H. Ballance has observed that Tekfur Sarayı in Istanbul has similar chequerboard arches: the present church may be c. 12th century, although its proportions look earlier.

¹²⁵ Perhaps spanning the ancient road, of which there are traces nearby.

¹²⁶ Spratt and Forbes, op. cit., p. 135.

¹²⁷ Reisen, II, p. 53Google Scholar.

¹²⁸ Reisen, I, p. 130Google Scholar.

¹²⁹ Texier, op. cit., pp. 199–200, Pls. 192–3; Spratt and Forbes, op. cit., p. 105. Both were demolished before Rott's visit.

¹³⁰ op. cit., p. 91.

¹³¹ cf. TAM. II, 822–3Google Scholar.

¹³² Cut in the rock at many points; several are flanked by the apocalyptic letters. For sundry other remains, see Rott, op. cit., pp. 298, 324, 343 (note).

¹³³ Lying beside the western façade's south pier, to the western side of which it surely belongs.

¹³⁴ Lying at the northern end of the narthex; height 0·37 m., lower width and depth 0·53 and 0·30 m.

¹³⁵ Height of capitals 0·46 m., width of abacus 0·70 m., lower diameter 0·42 m.

¹³⁶ cf. Rott, op. cit., Pls. 120–1.

¹³⁷ ibid., Pl. 116.

¹³⁸ Key to photograph (in which capital lies upside-down): A, vine-scroll; B, splayed acanthus-leaves; C, wind-blown acanthus to left (height 0·14 m.); D, wind-blown acanthus to right (height 0·22 m.); E, acanthus-leaves set obliquely.

¹³⁹ See above, note 109.

¹⁴⁰ Nothing comparable is recorded in Lycia. In Pamphylia a sixth-century church at Antalya has rich, probably local carving, but this is of a different character; cf. Ballance, M. H., “Cumanın Cami'i at Antalya: a Byzantine Church,” PBSR. (1955), p. 114Google Scholar.

¹⁴¹ Salonika: cf. the drilled acanthus of the Salonikan ambo in Istanbul (Volbach, W. F., Frühchristliche Kunst, 1955, Pls. 78–9Google Scholar) and the capitals of St. Demetrius (Kautzsch, R., Kapitellstudien, Pl. 29)CrossRefGoogle Scholar. Constantinople: cf. St. John of the Studion (Deichmann, F. W., Studien zur Architektur Konstantinopels, 1956, Pls. 17–18Google Scholar) and St. Polyeuktos (Mango, C. and Ševčenko, I., “Remains of the Church of St. Polyeuktos at Constantinople,” Dumbarton Oaks Papers 15, 1961, pp. 243–7CrossRefGoogle Scholar). These parallels, however, are only valid in some small points of detail.

¹⁴² Breccia, E., Le Musée Gréco-Romain d'Alexandrie, 1925–31 (Bergamo, 1932), pp. 60–3Google Scholar, Pls. 41–51; ibid., 1931–32 (Bergamo, 1933), pp. 36–45, Pls. 13–50. cf. also Kitzinger, E., “Notes on Early Coptic Sculpture,” Archaeologia, LXXXVII (1938), pp. 181 ff.CrossRefGoogle Scholar, and Harris, J. M., “Coptic Architectural Sculpture from Oxyrhynchos,” Year Book of the American Philosophical Society, 1960, pp. 592–8Google Scholar. Miss Harris is planning a detailed study of the multitudinous fragments, showing them to belong to the fifth and early sixth centuries.

¹⁴³ Notably Greek (“Maltese” and foliate) at Alakilise and Muskar, Latin at Alakilise and Karabel, two at the latter site being monogram-crosses.

¹⁴⁴ For vine-scroll with birds, see p. 147; cf. also Fig. 7 above, and Rott, op. cit., Pls. 121 (Alakilise) and 128 (Myra), with which cf. Ramsay, W. M., The Revolution in Constantinople and Turkey (London, 1909), Pl. 33 (Konya)Google Scholar.

¹⁴⁵ A small block shown to us at Karabel, but said to come from Devekuyusu; carved area 0·17 m. long, 0·08 m. wide.

¹⁴⁶ Note fragment of larger bird below. Height of block 0·29 m.

¹⁴⁷ Lying in church of St. Nicholas; cf. Rott, p. 333. Height 0·35 m.

¹⁴⁸ Fragment of screen lying just west of cross-in-square chapel; length 0·58 m., height 0·27 m., thickness 0·15 m.

¹⁴⁹ Not inscribed as in Syria and (thence) Cilicia.

¹⁵⁰ Muskar; see note 92 above.

¹⁵¹ For the cross-in-square plan, cf. Church of the Prophets, Apostles and Martyrs (464–5); Kraeling, C. H., Gerasa, City of the Decapolis (New Haven, 1938), Pl. 41Google Scholar. The barrel-vaulting, massive masonry and severely rectangular exterior are typically Syrian features. For the triconchos-plan, see below.

¹⁵² cf. e.g. Crema, L., L'Architettura Romana (Turin, 1959), p. 564Google Scholar (Sardis) and p. 635 (Ardea).

¹⁵³ cf. e.g. ibid., Fig. 805 (Piazza Armerina). For other Roman examples of this plan, see de Villard, Monneret, Les Couvents près de Sohâg (Milan, 1925–1926), pp. 49 ffGoogle Scholar.

¹⁵⁴ Abu Mina (cf. Perkins, J. B. Ward, “The Shrine of St. Menas in the Maryût,” PBSR. XVII (1949), pp. 26 ff.Google Scholar); Hermoupolis Magna (Megaw, A. H. S., “Ἡ βασιλικὴ τῆς Ἑρμουπόλεως,” ΠΕΠΡΑΓΜΕΝΑ ΤΟΥ Θ' ΔΙΕΘΝΟΥΣ ΒΥΖΑΝΤΙΝΟΛΟΓΙΚΟΥ ΣΥΝΕΔΡΙΟΥ Θ′ΕΣΣΑΛΟΝΙΚΗΣ (Athens, 1954), pp. 287 ff.Google Scholar); the White and Red Monasteries (Deyr el-Abiad and Deyr el-Ahmar) near Sohâg (Monneret de Villard, op. cit., passim); the church of Denderâ (ibid., pp. 48–9, Fig. 52).

¹⁵⁵ Church of St. John the Baptist at Jerusalem (cf. ibid., p. 55, Fig. 54).

¹⁵⁶ Episcopal palace, Bosra (cf. ibid., p. 53, Pl. 66).

¹⁵⁷ The scale too is important: the difference in engineering problems between roofing a small chamber and a major bay is such as to constitute a difference in kind.

¹⁵⁸ e.g. Hermoupolis Magna and the White Monastery (see above, note 154).

¹⁵⁹ e.g. the Red Monastery (see above, note 154).

¹⁶⁰ The present domes of the White and Red Monasteries are late Coptic, but a dome is conjectured over the square bay of the church of Denderâ (cf. Monneret de Villard, op. cit., p. 48).

¹⁶¹ See above, note 154; the early-fifth-century basilica of Paulinus at Nola appears to have had this plan, but may well be derived from Egypt (cf. Monneret de Villard, op. cit., p. 56).

¹⁶² It is worth noting that a sixth-century triconchos in Asia Minor is recorded in the Life of St. Theodore of Sykeon, 55.

¹⁶³ Comparison of details suggests that this was the prototype for Devekuyusu and Alacahisar, and probably Dikmen.

¹⁶⁴ A relative chronology for these Alaca Dağ churches can be attempted, whereby Alakilise and Muskar stand at the beginning of the series, followed by Karabel, Devekuyusu and Alacahisar; but absolute dating is extremely difficult, both because of the lack of firmly dated comparative material elsewhere and because of the danger of applying general theories of formal development to a remote mountain-area.

¹⁶⁵ Anrich, op. cit., pp. 238–40, set out reasons, from internal evidence in the VNS., for locating the Sion monastery on the west side of Alaca Dağ, equally accessible from Myra and Arneai; this argument is now strengthened by the discovery of an ancient road (see above, note 96) running north from, Myra, through Muskar and Karabel to Çağman, thence probably forking right to Arycanda, left to Arneai. The site was on a hill (VNS. 12), the plan was multi-apsed (ibid., 4, ἐπαρξάμενος τῇ τῶν κογχῶν περιχαραγῇ), the building-blocks so large that divine assistance was claimed for their erection (ibid., 39). Nicholas' tomb was on the south side of his church, where were also the relics of various martyrs (ibid., 80); if it is rightly identified, this would explain the domed-chapel (slightly postdating the main building) with its sarcophagi, and, perhaps, the barrel-vaulted chamber built west of this.

¹⁶⁶ Building was begun while Nicholas was still a boy (ibid., 4 and 7); certainly he was well established before 442 (episode of the plague and famine, ibid., 52–3). It is unfortunate that the dates of Nicholas archbishop of Myra, who approved the site (ibid., 4), are unknown.

¹⁶⁷ cf. Perkins, J. B. Ward, “The Italian Element in Late Roman and Early Medieval Architecture,” Proceedings of the British Academy, XXXIII (1947), pp. 163–83, esp. 177 ffGoogle Scholar. id., “The Shrine of St. Menas in the Maryût,” PBSR. XVII (1949), pp. 26 ff. The relevance of the Lycian churches in this context cannot be stated in precise terms, but it is worth recalling that Haghia Sophia's two architects came from south-west Asia Minor (Procopius, , De aed., I, 1, 24Google Scholar).