¹ For this phenomenon, see Mango, C., Byzantium, The Empire of New Rome (New York 1981) 60–87Google Scholar.

² A generous grant from the American Council of Learned Societies made this research possible. My sincere thanks to them and to my driver, Nail Ildem of Izmir, whose skill and patience made a difficult trip far simpler than it might have been.

³ See the two versions of the pilgrimage of Saint Willibald in Tobler, T. and Molinier, A., Itinera hierosolymitana (Geneva 1879) I. 257Google Scholar (Hodeoporicon, quoted here) and I. 288 (Itinerarium: “Urbis Strabolis monte excelso”). This and many of the sources which follow have been gathered by Tomaschek, W., “Zur historischen Topographie von Kleinasien im Mittelalter”, Sitzungsberichte Wien 124 (1891) 81 f.Google Scholar See also Amantos, K., “Strobilos” in Hellenika 1 (1939) 292Google Scholar; Kambouroglou, D., “Semeioma peri Strobilou”, in Eis Mnemen Spyridonos Lamprou (Athens 1935) 428Google Scholar, contributes little to the subject.

⁴ Synaxarium ecclesiae constantinopolitanae ( = Propylaeum ad Acta Sanctorum Novembris, Brussels 1902) 520, 522Google Scholar.

⁵ Ryden, L., “The Andreas Salos Apocalypse”, DOP 28 (1974) 197–261Google Scholar at 204, sec. 857B.

⁶ Karioupolis has recently been investigated: it apparently originated in the ninth century, but its remains are largely of the thirteenth century and later: see Etzeoglou, H., “Karyoupolis, une ville byzantine désertée”, Byzantion 52 (1982) 83–123Google Scholar.

⁷ The original text was published by Ginzberg, L. in Hebrew in Genizeh Studies in Memory of Doctor Solomon Schechter I (New York 1928) 313–23Google Scholar, and discussed in detail by Krauss, S., “Un nouveau texte pour l'histoire judeo-byzantine”, REJ 87 (1929) 1–27Google Scholar. Although Krauss' attempts to identify the obscure persons and events of this text seem of dubious merit, he did improve it by emending “Strogilion” to “Strobililon” which, however, he identified with the Stribali islands, ancient Strophades, in the Ionian Sea. The last three names are enigmatic: mention of the obscure Osteodoes islands (north of Sicily) rests on an emendation of the original “Asynyad”; Amorgos is represented only by the letters Arm- which could as well stand for the Armenokastron of Andreas Salos; and the last name is transcribed “Istamboli” for which the obvious equation with Constantinople might seem appropriate. For the apocalyptic literature to which this text belongs, see Macler, F., “Les apocalypses apocryphes de Daniel”, RHR 33 (1986) 37–53, 163–76, 288–319Google Scholar, and the various references of Ryden.

⁸ The archon is known from a seal published by Davidson, G. R. in Corinth XII, The Minor Objects (Princeton 1952), no. 2727Google Scholar and assigned to Zacythinos, Euboea. D. in “Meletai peri tēs dioikētikēs diairēseōs kai tēs eparchikēs dioikēseōs en tō Byzantinō kratei”, EEBS 25 (1955) 127–57Google Scholar at 148 f. recognized it as belonging to the Carian town and assigned it to the tenth or early eleventh century. In fact, the elaborate cross which appears on the seal most closely resembles that of Class H of the anonymous bronze coinage, assigned to Michael VII (1071–1078): see the illustrations in Dumbarton Oaks Catalogue III.2 plate lxviii. For the office of archon, see Ahrweiler, H., Byzance et la Mer (Paris 1966) 54–61Google Scholar and Oikonomides, N., Les listes de préséance byzantines (Paris 1972) 342 fGoogle Scholar.

⁹ Constantine Porphyrogenitus, de Thematibus, ed. Pertusi, A. (Vatican 1958) 78Google Scholar.

¹⁰ Theophanes Continuatus 388 ( = Leo Grammaticus 294 = Scylitzes 202).

¹¹ Scylitzes 398, a text which seems to imply that the Thracesian theme stretched as far as Strobilos, a situation quite unlike that described by Constantine Porphyrogenitus a century earlier. During this period, there was apparently a Jewish community at Strobilos: see the following section.

¹² MM VI. 19–21Google Scholar, a document of August 1079, edited with introduction and notes by Vranouses, E., “Chrysobyllon Nikephorou tou Botaneiatou”, EEBS 33 (1964) 52–69Google Scholar.

¹³ MM. VI. 62, 87Google Scholar.

¹⁴ MM. VI. 84 f.Google Scholar, 89.

¹⁵ Saewulf, , ed. Rev. Brownlow, Canon, Palestine Pilgrims Text Society (London 1892) 51Google Scholar. This edition prints the name as “Stromlo”, an easy misreading for “Strouilo”.

¹⁶ Tafel, G. L. and Thomas, G. M., Urkunden zur älteren Handels- und Staatsgeschichte der Republik Venedig (Vienna 1856) 52Google Scholar.

¹⁷ Ibid., 118, 265. The latter document, like the sources cited above, notes 3, 7 and 16, describes Strobilos as an island. This has led to some confusion: Amantos (above, n. 3) supposed that Theophanes Continuatus by specifying Strobilos κατὰ Κιβύρραν meant a place in Lycia rather than Caria (apparently not understanding that the phrase meant merely “in the Cibyrrhaeot theme”). Şahin, S., Bithynische Studien (Bonn 1978) 34 n. 11Google Scholar referring to Synaxarium 520, conjured up an otherwise unknown island in the Sea of Marmara. Actually, the “island” refers to the same Carian Strobilos, and the explanation is quite simple. In Greek, the term nēsos is used rather loosely for peninsulas almost surrounded by water, and for places which project into the sea. The Peloponnesus is the obvious example of the first case; for the Byzantine period, Theophanes 375, referring to the “island” of Amastris, might be added. For the second case, there is a parallel dealing with late antique Caria: in 467, according to Priscus frag. 43 in FHG IV. 110Google Scholar, the “islands” of Cnidus and Cos were devastated by an earthquake. Many other examples could no doubt be found. Such a usage probably originated with sailors: many headlands and ports of the rocky Aegean coast appear to be islands when seen at a distance from the sea. Many instances of this phenomenon in this very area are cited in the seventeenth-century Greek sailing directions published in Delatte, A., Les portulans grecs (Liège 1947) 248 f.Google Scholar: the two harbours of Bodrum, the northern cape of Cos, and Cape Crio (Cnidus). Consequently, these late texts show some apparent confusion between headlands and islands. As will be seen, Strobilos stood on a high hill projecting into the sea which might legitimately be called an “island”.

¹⁸ Géographic d'Idrisi, tr. Jaubert, P. (Paris 1836) II. 134 f.Google Scholar, 129.

¹⁹ Tafel and Thomas (above, n. 16) 477; for Strobilos as an island, see above, n. 17.

²⁰ MM. VI. 183Google Scholar.

²¹ Pachymeres I. 220, 311.

²² Byzantinische Kleinchroniken, ed. Schreiner, P. (Vienna 1975) 666Google Scholar, Chron. 106/3.

²³ Malipiero, Domenico, Veneti, Annali, ed. Longo, F. in Archivio storico italiano 7.1 (1843) 75Google Scholar.

²⁴ Piri Reis, ed. P. Kahle (Berlin 1926); Piri Reis Kitabi Bahriye, ed. Kurdoğlu, F., Alpagot, H. (Istanbul 1935)Google Scholar. Kahle uses MSS of 1570 (Bologna), 1554 (Dresden) and later (Vienna and Mordtmann); all of these derive from a tradition which goes back to 1521, while the Istanbul edition is based on the Aya Sofya MS, derived from another of 1532: see Kahle xx–xxxiv, Istanbul xxxix–xli. The name Sirevolos and similar forms should occasion no surprise: Turkish does not tolerate groups of consonants, but breaks them up. In this case, a vowel has been inserted for ease of pronunciation and the -t- dropped. An exactly parallel formation may be observed in Belisırama, the modern name of the Peristremma river in Cappadocia.

²⁵ Piri Reis, ed. Kahle I. 48 (trans. II.69), cap. 22; Istanbul ed. 213 f. The latter treats the west and south coast separately under the rubric, “Coast of Kördil, Gümüşlük and Karabağ”.

²⁶ Piri Reis, ed. Kahle I.57, II.80. For Pitasa, see, e.g., the British Admiralty Charts (available in various editions and reprintings) and the Mediterranean Pilot (2nd ed., Washington 1925) IV. 346Google Scholar.

²⁷ Pedasa was identified in Paton and Myres (1896) 192–5, 202; cf. Maiuri, A., “Viaggio di esplorazione in Caria”, Annuario 4/5 (1921/1922) 425–32Google Scholar and Bean and Cook (1955) 123 f., 149–51.

²⁸ Bean and Cook (1955) 130; one of the forts was on the headland Uzun Burun (now called Ada Burnu), the other in the hills between Bitez and Müsgebi: see the map of Paton and Myres (1896) pl. XI after p. 356.

²⁹ Since the list follows an alphabetical order, no geographical conclusions may be drawn from it; see Kalfa, Haji, Jihannuma (Istanbul 1145 AH = 1732) 638 f.Google Scholar; the text is analysed in Wittek, P., Das Fürstentum Mentesche (Istanbul 1934) 163–9Google Scholar. This section of the work was written by a continuator: ibid., 163 n. 2. For Sıravala, see ibid., 166 n. 12.

³⁰ Çelebi, Evliya, Seyahatname IX (Istanbul 1935) 218Google Scholar. For Karabağlar, see Wittek (previous note) 169. Wittek 166 n. 12 and 171 mentions Sıravala as occurring in the itinerary of Evliya (IX. 211) between Peçin and Varvil, the ancient Bargylia, an unexpected location in the interior to the northeast of the Halicarnassus peninsula. The printed edition of Evliya, however, gives the name of that place as Muvalasun; it may therefore be removed from consideration in this context.

³¹ For these charts and texts, see Kretschmer (1909).

³² Kretschmer (1909) 106 ff. Dr. Anthony Luttrell and I confirmed the readings of this chart in the Bibliothèque Nationale.

³³ Ibid., 522.

³⁴ Mediterranean Pilot (above, n. 26) IV. 348.

³⁵ By an alternate interpretation, it might be possible to take “ed e entro e garbin mia li” to refer to the size and location of the island which stretches southwest from the cape and is, according to the same Portolan, sec. 250, fifty miles in circumference.

³⁶ Mediterranean Pilot IV. 347 fGoogle Scholar.

³⁷ For Evliya Çelebi, the distance along the coast from Bitez (which he calls Vitiz) was eight miles: see the text quoted below.

³⁸ It is marked Aspat on the Turkish 1:200000 map and is so called by the locals; the name Chifut Kalesi is currently unknown. Chifut or çıfit is a highly derogatory term best expressed by a kind of colloquialism inappropriate here.

³⁹ Sakellion, J., Patmiakē Bibliothēkē (Athens 1890) 44 n. 6Google Scholar. The work was finished in 1862 but met long delays in publication described in the introduction. Sakellion had first gone to Patmos in 1841 and thus could have visited Strobilos at any time during the next twenty years.

⁴⁰ Robert, L., Études épigraphiques et philologiques (Paris 1938) 164–6Google Scholar. Robert's identification was accepted by Bean and Cook (1955) 129 n. 184. P. Kahle, in his edition of Piri Reis, II. 71 n. 5 identifies Aspat with Strobilos, but does not cite his evidence.

⁴¹ Pl. XX(a). It is here labelled “Curro”. (Photo kindly provided by A. Luttrell.)

⁴² Piri Reis, ed. Kahle II.71 ( = I.49, map I.50, n. 9).

⁴³ Piri Reis, Istanbul edition 224, 222, map 223. The printed edition renders the name as Esyot, a variation, like Usput, of no significance.

⁴⁴ For the maps of Piri Reis, see Kahle I.x–xiii, and for the MSS, above, n. 24. Although the Istanbul edition is based on an early MS tradition, the date of the published text is not clear. Aspat is also designated as ruined in Kahle II.71.

⁴⁵ Çelebi, Evliya, Seyahatname IX. 214Google Scholar; my translation.

⁴⁶ Beaufort, F., Karamania (London 1817) 102Google Scholar.

⁴⁷ Hamilton, W., Researches in Asia Minor, Pontus and Armenia (London 1842) II. 38Google Scholar.

⁴⁸ Newton 588–90. See also Bean and Cook (1955) 129, who noted a medieval circuit wall, the monastery of the Virgin, spoils, rock-cut tombs and ruins of the deserted village.

⁴⁹ Cousin, G. and Diehl, Ch., “Inscriptions d'Halicarnasse” BCH 14 (1890) 90–121CrossRefGoogle Scholar at 120. The inscriptions are reproduced in Grégoire, H., Recueil des inscriptions grecqueschrétiennes d'Asie Mineure (Paris 1922) nos. 232–3Google Scholar.

⁵⁰ Grégoire assigned the inscriptions to the seventh century for no evident reason. My thanks to Professor Ihor Ševčenko for his helpful comments.

⁵¹ See above, n. 12.

⁵² For this pottery in the context of an Ottoman site, see Crane, H., “Archaeological notes on Turkish Sardis”, Muqarnas 4 (1987) 43–58 at 53CrossRefGoogle Scholar.

⁵³ The inscription finds an exact parallel in CIG 8922 from the region of Antioch (and thus probably of the 5–6th c.), incised on the wall of a church: σταυρὸς πεσόντων ἀνάστασις.

⁵⁴ I did not visit the church, but report the observations of Mr. Robert Carter.

⁵⁵ For examples in a similar style, see Grabar, A., Sculptures byzantines du moyen age II (Paris 1976) 127–50Google Scholar with plates civ–cxxxix passim.

⁵⁶ The wall uses rough smallish blocks no more than a metre long. They seem to represent a poorer version of the distinctive masonry used in the early Mausolan walls of Halicarnassus (Bean and Cook [1955] Pl. 15b) and Myndus (Bean, G., Turkey Beyond the Maeander [London 1971] pl. 31Google Scholar; Bean and Cook [1955] 110f.). There seems no reason to suppose that they represent the remains of an earlier Lelegian settlement, for those use masonry of a quite different type; see the illustrations in Radt, W., Siedlungen und Bauten auf der Halbinsel von Halikarnassos (Tübingen 1970)Google Scholar; cf. Paton, W. and Myres, J., “Karian Sites and Inscriptions”, JHS 16 (1896) 188–236CrossRefGoogle Scholar at 203, who remark on the unsuitability of the site for a Lelegian settlement. The larger wall at Strobilos, which stands some three metres high, was so overgrown that it could not be studied in detail. Both walls could represent terracing for houses, but this would not change the chronology, for similar terracing near Mylasa has been dated to the fourth century: see Radt, W., “Kuyruyklu Kalesi”, IstMitt 19/20 (1969/1970) 165–76, pl. 30.3Google Scholar and 171 n. 25 for reference to similar masonry in Caria.

⁵⁷ Bean and Cook (1955) 168; cf. Bean (above n. 56) 115–27.

⁵⁸ Newton 590 f. identified it with Termera (and so it appears in Kiepert's Atlas von Hellas), but that has been convincingly located at Asarlık, a well-fortified high hill some five kilometres to the northwest: Paton, W., “Excavations in Caria” JHS 8 (1887) 64–82CrossRefGoogle Scholar; cf. Paton and Myres (1896) 203 f., and Bean and Cook (1955) 116 ff., 147 ff.

⁵⁹ For such walls, see Foss, C. and Winfield, D., Byzantine Fortifications, An Introduction (Pretoria 1986) 131–42Google Scholar.

⁶⁰ These types of masonry, which have not been studied in the context of fortifications, are common in early Ottoman buildings: see the numerous examples illustrated in the works of Ayverdi, E. H., Osmanli Mi‘marisinin Ilk Devri, Osmanli Mi‘marisinde Çelebi ve II Sultan Murat Devri and Osmanli Mi‘marisinde Fatih Devri (Istanbul 1966–1973)Google Scholar.

⁶¹ Arel, A., “Menteşe Beyliǧi Devrinde Peçin Sehri”, Anadolu Sanatı Araştırmaları 1 (1968) 69–101Google Scholar at 71 f. The walls are here curiously called Byzantine, even though there is no evidence that the Byzantines occupied the site.

⁶² Keçikalesi: Müller-Wiener, W., “Mittelalterliche Befestigungen im südlichen Jonien”, IstMitt 11 (1961) 5–122 at 112–16Google Scholar and, for the date, Foss and Winfield (above, n. 59) 188 n. 84; for Torbalı (Metropolis of Ionia), see ibid. 157.

⁶³ See the useful sketch of regional history in Luttrell (1986) 143–7.

⁶⁴ Dolger, F., Regesten der Kaiserurkunden des oströmischen Reiches (Munich 1925) 1390Google Scholar, with list of editions and bibliography; cf. Starr, J., The Jews in The Byzantine Empire (Athens 1939) no. 181Google Scholar, where the text is mistranslated.

⁶⁵ Antoniadis-Bibicou, H., Études d'histoire maritime de Byzance (Paris 1966) 129–37Google Scholar.

⁶⁶ Janin, R., Constantinople byzantine (Paris 1964) 432 fGoogle Scholar.

⁶⁷ Synaxarium (above, n. 4) 702, 904, 908; Theophanes Continuatus 323.

⁶⁸ For Basiliscus, see Janin, R., La géographie ecclésiastique de l'empire byzantin (Paris 1953) I.iii.65Google Scholar, and for the Prodromus, ibid., 455 and idem., “Les églises byzantines du précurseur à Constantinople” Échos d'Orient 37 (1938) 312–51 at 337 f., and “Le Pétrion de Constantinople, Échos d'Orient 36 (1937) 31–51CrossRefGoogle Scholar.

⁶⁹ Church of St. Juliana: Janin, , Géographie ecclésiastique 268Google Scholar; Pétrion, : Constantinople byzantine 407 f.Google Scholar, and in more detail, “Pétrion”; cf. Schneider, A., Mauern und Tore am Goldenen Horn zu Konstantinopel, Nachrichten Göttingen 1950: 6 no. 5, 65–107 at 72 f.Google Scholar, with a plan and information about the walls of this district.

⁷⁰ Byzantios, S., He Konstantinoupolis (Athens 1851) 562 f.Google Scholar A biographical sketch of the writer (1798–1878), who is remembered both for his monumental work on Constantinople and for his tireless efforts as director of national education, may be found in the Megalē Hēllēnike Enkyklopaideia.

⁷¹ It is perhaps instructive to note the remark of Janin, , Géographie ecclésiastique 268Google Scholar, that the churches of Juliana were identical because “Strobilos se trouvait précisément dans le Pétrion” and idem, “Pétrion” where the location of Strobilos, mentioned several times, is assumed as well established.

⁷² Jacoby, D., “Les quartiers juifs de Constantinople”, Byzantion 37 (1967) 167–227 at 167–89Google Scholar.

⁷³ Mann, J., The Jews in Egypt and Palestine under the Fatimids (Oxford 1922) 344 f.Google Scholar

⁷⁴ Starr (above, n. 64) 186 with revised text 245.

⁷⁵ For Strobilos in Bithynia, which I will discuss in a different context, see Şahin (above, n. 17) 29–36.

⁷⁶ Text from Galanté, A., Histoire des juifs d'Anatolie (Istanbul 1939) 188Google Scholar, where information about the findspot will be found. This inscription appears to be identical with that mentioned by Cousin and Diehl (above, n. 49) 120. Although they noted that the stone bore a seven-branched candlestick, the decoration was later described as possibly representing a palm (Frey, J.-B., ed., Corpus Inscriptionum Judaicarum [Vatican 1952] II. no. 756aGoogle Scholar); it may, in fact, have consisted of the design reproduced here from Galanté. The stone seems to have disappeared; no photograph has been published. My sincere thanks to my colleague Prof. Lester Segal for the translation and comments, and to Drs. Mark Pinson and Israel Bartal for further discussion.

⁷⁷ Piri Reis, ed. Kahle I.57 (map), II.89 (text), II. xxiv–xxix (MSS).

⁷⁸ The Jews of Mastaura are known from the documents of the Cairo Genizeh only: see Starr (above, n. 64) nos. 130 and 139; those of Seleucia appear only in a letter of 1137: Goitein, S., “A Letter from Seleucia (Cilicia)”, Speculum 39 (1964) 298–303CrossRefGoogle Scholar.

⁷⁹ The Itinerary of Benjamin of Tudela ed. Adler, M. (London 1907) 14Google Scholar.

⁸⁰ MM VI. 36Google Scholar.

⁸¹ Psellos, , Scripta minora II.299Google Scholar.

⁸² See the inscription first published by Condoleon, A.-E. in REG 12 (1899) 386 no. 11Google Scholar; republished by Salač, A. in BCH 51 (1927) 394 f.CrossRefGoogle Scholar with illustration; its provenance was discussed by Robert, L. in RevPhil 1929.135Google Scholar (= Opera minora selecta II.1101Google Scholar).

⁸³ Şahin (above, n. 17) 35 f.

⁸⁴ Lazari, Vita (Acta Sanctorum Nov. III [Brussels 1910] 502–608)Google Scholar 510.

⁸⁵ Portolans: Kretschmer (1909) 397 (sec. 119), 412 f. (sec. 168, 169, 170); Strophili: Amantos (above, n. 3).

⁸⁶ P. Soustal, Nikopolis und Kephallenia (= TIB 3; Vienna 1981) 265.

⁸⁷ Lemerle, P. et al. , Actes de Lavra (Paris 1970) no. 28, p. 183Google Scholar; see nos. 12, 54 and 57 for the abbots.

⁸⁸ For all these places, see Amantos (above, n. 3), who adduced Strombolo to clarify the description of Strobilos as an island (see above, n. 17).

⁸⁹ The surviving part of the text reads: ἡγεμο]νευόντος Οὐαλ[ηρίου / τοῦ διασημ / οτάτου / Ṁ I. It was thus set up by a governor who bore the title perfectissimus and whose name was Valerius. Although he cannot be identified with any known praeses of Caria, his title indicates a date between Diocletian, who established the province of Caria, and the late fourth century, when its governor gained the rank of clarissimus: see the fasti tabulated in PLRE I.1100.

⁹⁰ For roads in the peninsula, see Paton and Myres (1896) 201 with map, Pl. XI. The remains of this road are discussed by Bean and Cook (1955) 131 and attributed to the time of Mausolus.

⁹¹ In their brief note of the site, Bean and Cook (1955) 131 f. mention a Christian church on the summit of the hill; they are presumably referring to this building.

⁹² For Güvercinlik, see Newton 602 f., and for Salih Adası, Bean and Cook (1955) 132.

⁹³ For Caryanda, see The Athenian Tribute Lists, ed. Meritt, B. et al. (Cambridge, Mass. 1949) I.493Google Scholar and index, s.v., and Bean and Cook (1955) 151–9; and for its classical remains—a fort, various buildings, and rock-cut tombs—ibid., 121 f., and Guidi, G., “Viaggio di esplorazione in Caria”, Annuario 4/5 (1921/1922) 345–96Google Scholar at 363–5 with additional illustrations; cf. Robert, L., “Villes cariennes dans les listes des tributs attiques”, Rev Phil 62 (1936) 276–84Google Scholar (= Opera Minora Selecta III. 1466–77Google Scholar). See also Newton 596–600 who describes classical spoils built into a Byzantine wall. It was Newton who first suggested the identification with Caryanda.

⁹⁴ On these capitals, see Yegul, F., “Early Christian Capitals from Sardis: A Study of the Ionic Impost Type”, DOP 28 (1974) 265–74Google Scholar. The present example, with its Ionic part reduced and its volutes flattened, belongs to the later, simple stage of development represented by St. Irene in Constantinople and St. John in Ephesus, both built by Justinian around 540.

⁹⁵ See French, D., “Milestones of Pontus, Galatia, Phrygia and Lycia”, ZPE 43 (1981) 147–74Google Scholar at 173 f. The text of the present stone is identical with the example from Bodrum published by French, except for occasionally variant abbreviation, preservation of the first line τοῖς κυρίοις ἡμών and the mileage.

⁹⁶ Villagers reported a place called kale, “castle”, above the village, but stated that no remains were to be found there.

⁹⁷ Piri Reis, Istanbul ed. 213.

⁹⁸ Both names appear in Hierocles, but not in the Notitiae. They were presumably ephemeral honorifics given to existing sites, which may have resumed their old names subsequently.

⁹⁹ Newton 579; cf. Bean and Cook (1955) 128.

¹⁰⁰ de Suchem, Ludolfus, de Itinere Terrae Sanctae, ed. Deycks, F. (Stuttgart 1851) 27Google Scholar.

¹⁰¹ For the quarries, see Bean and Cook (1955) 130 and for the stone and its use, Luttrell (above, n. 41) 203 f.

¹⁰² Bean and Cook (1955) 130 f.; Alakilise: Maiuri (above, n. 93) 446, with photo (site called Alakişla).