^{page 157 note 1} Travels, tr. von Hammer, i. 1, 167.

^{page 157 note 2} Ibid. i. 2, 49.

^{page 157 note 3} Ibid. i. 1, 24. Evliya states that the Rose Mosque, having become a church, was turned over to the Moslems as the price of Bayezid II.'s retirement from Constantinople. Bayezid made a demand of this sort in 1391, but it was not complied with (Ducas, 49 B). For the real history of the Mosque (S. Theodosia) see van Millingen's, Churches of Constantinople, pp. 164 ff.Google Scholar

^{page 158 note 1} Travels, i. 2, 51.

^{page 158 note 2} Belin, , Histoire de la Latinité de Constantinople, pp. 215 ff.Google Scholar The Church of S. Paul is mentioned about 1400 by Clavijo (Hakluyt Soc. Edn. 49).

^{page 158 note 3} Two, bearing date 1423 (Atti Soc. Lig xiii. 322 (3)) and 1433 (B.S.A. xi. 54), had been recorded earlier.

^{page 158 note 4} These had been hidden under the wooden floor, but were known to exist in the 'sixties (De Launay, , cited in Atti Soc. Lig. xiii. 273).Google Scholar

^{page 158 note 5} In Hammer-Hellert, , Hist. Emp. Ott. xviii. 71.Google ScholarEvliya, (Travels, i. 1, 25Google Scholar; i. 2, 49, 51) says it was built by the Caliph Omar Abd-ul-Aziz during the fifth siege, which he dates A.H. 92.

^{page 159 note 1} D'Ohsson, , Tableau de l'Empire Ottoman, i. 285Google Scholar; Byzantios, Scarlatos, Κωνσταντινούπολις, ii. 46.Google Scholar

^{page 159 note 2} For the Arab accounts see Brooks J.H.S. xviii., xix.

^{page 159 note 3} See the account of the siege and the disposition of the Arab forces in Bury's, Later Roman Empire, ii. 402 ff.Google Scholar

^{page 159 note 4} De Adm. Imp. xxi. iii. (p. 101 B).

^{page 160 note 1} Bury, , Eastern Roman Empire, p. 279.Google Scholar

^{page 160 note 2} See the passages cited by Ducange, , CP. Xliana, ii. (p. 164 P), xv.Google Scholar

^{page 160 note 3} Ibn Batuta, tr. Lee, p. 83, note.

^{page 160 note 4} The idea is much older; cf. Schiltberger's, Travels, ed. Telfer, , p. 66 (c. 1400).Google Scholar

^{page 161 note 1} In Ray's, Collection of Curious Travels and Voyages, i. 311Google Scholar; cf. Shaw's, Travels in Barbary, p. 246.Google Scholar The fear of Christian attack during Friday prayers was not without reason; there was an unsuccessful plot for the surprise and recapture of Rhodes at this hour in 1525 (Torr, , Rhodes in Modern Times, p. 33)Google Scholar, and down to the last century Christians were locked out of the walled city of Rhodes at prayer-time on Fridays (Jowett, , Christian Researches, p. 416Google Scholar; Turner, W., Tour in the Levant, iii. 117Google Scholar; Elliot, C.B., Travels, ii. 175.Google Scholar George Borrow, in the thirties, found the same tradition and practice current at Tangier (Bible in Spain, ad fin.). The same idea occurs also in a Greek folk-story from Trebizond (Polites, Παραδόσεις, No. 22).

^{page 161 note 2} See below, p. 171.

^{page 161 note 3} Gerlach, , Tagebuch, p. 102.Google Scholar This is the prophecy of the ‘Yellow Race’ generally interpreted of the Russians, and evidently a composition of this time when Ivan the Terrible was consolidating his empire. It was revived in the early years of the last century, when the Russian menace was still more apparent, and is cited by Hobhouse in his Travels.

^{page 161 note 4} These appearances are pictured and described by the Venetian cartographer Camotti.

^{page 161 note 5} Hammer-Hellert, , Hist. Emp. Ott. vii. 44.Google Scholar The extreme susceptibility of the Turks to interpret extraordinary events in the most gloomy sense is illustrated by their apprehensions when the Bosporus froze in 1669: they were ‘so frightened that they looked upon it as a dismal Prodigy and concluded that the world would be at an end that year’ (Smith, T. in Ray's, Voyages, ii. 46).Google Scholar

^{page 161 note 6} Hammer-Hellert, , Hist. Emp. Ott. vii. 282 P.Google Scholar

^{page 161 note 7} Jardin des Mosquées in Hammer-Hellert, Hist. Emp. Ott. xviii.

^{page 161 note 8} D'Ohsson, , Tableau, vii. 325.Google Scholar

^{page 162 note 1} This phase of affairs was made good use of by the rising Protestant powers, England and Holland. The first English treaty with the Porte was made in 1581, an embassy being established next year. The Dutch Capitulations date from 1610. Elizabeth certainly made capital out of the distinction between ‘Protestant’ England and ‘idolatrous’ Spain (see Pears, , in Eng. Hist. Rev. 1893, pp. 239 ff.)Google Scholar, and James followed her precedent. He is said to have styled himself to the Porte ‘Verus fidei contra idolatras falso nomen Christi profitentes [!]…propugnator’ (Ambassade de j. de Gontaut-Biron, ed. T. de Gontaut-Biron, p. 36).

^{page 162 note 2} Hammer-Hellert, op. cit. vii. 51.

^{page 162 note 3} Lives of the Norths, ii. 134.

^{page 162 note 4} Knolles, , History of the Turks, p. 899Google Scholar, where the decree of expulsion is given.

^{page 162 note 5} Hammer-Hellert, , Hist. Emp. Ott. vii. 51.Google Scholar

^{page 162 note 6} In 1578 a Constantinople letter (Charrière, , Négotiations de la France dans le Levant, iii. 787Google Scholar) mentions a complaint preferred by ‘dix ou douze Mores de Granate habitans icy…’ The rush began later: cf. Relaz. di M. Zane in Alberi, iii. 390 (1594): ‘di Spagna concorrono ognigiorni Mori in Constantinopoli, che si chiamano mondesari come si uscissero solamente di Granata, ma in effetto tutta la Spagna n'e contaminata e subito giunti levano il talbante’ (i.e. avow themselves Moslems); cf. also the same Relazione, p. 440. Later still (1608–10) the French embassy espoused the cause of the Moors fleeing from Spain through Marseilles, though official efforts on their behalf were not always successful; cf. Ambassade de J. de Gontaut-Biron, Table Analytique, p. 443, and Index, s.v. ‘Grenadins.’

^{page 163 note 1} Travels, tr. von Hammer, i₂. 51; cf. ibid. p. 53, ‘a great number of them are Arabs and Mogrebias.’

^{page 163 note 2} Hammer-Hellert, , Hist. Emp. Ott. vii. 287.Google Scholar

^{page 163 note 3} Ibid.

^{page 163 note 4} Smith, T. in Ray's, Voyages, p. 40Google Scholar: ‘S. Paul and S. Anthony were both taken away some years since from the Christians and turned into Moschs, the former of which is now known by the name of Arab Giamesi, or the Mosch of the Arabians.’ An earlier notice of the seizure of Paul, S. is given by Duloir, (Voyage (1654), p. 14)Google Scholar: Comidas, (Descr, di Costantinopoli, 1794, p. 59)Google Scholar seems certainly wrong in assigning the seizure to the reign of Suleiman (1520–1566), when the Moors to whom he attributes it were not yet fled out of Spain. But the Christians may have been dispossessed earlier. S. Paul's is not mentioned among the Latin churches of Galata by Breuning (1579, Reyss, p. 89).

^{page 163 note 5} Voyage, Letter XII. ‘La mosquée des Arabes fut confisquée sur les Dominicains, il y a environ 100 ans pour servir aux Mahometans Granadins.’

^{page 163 note 6} Hammer-Hellert, , Hist. Emp. Ott. vii. 232Google Scholar; the Jardin des Mosquées gives the date 1591.

^{page 163 note 7} Constantiniade, p. 108.

^{page 164 note 1} But their connection with medicine and the University of Salamanca lasted far into the next century (Smith, T. in Ray's Voyages, ii. 58).Google Scholar

^{page 164 note 2} Knolles, , History of the Turks, p. 917.Google Scholar

^{page 164 note 3} Ibid. and des Hayes, , Voyage, p. 125.Google Scholar

^{page 164 note 4} Meyer's, Konstantinopel, 7 Aufl. 199Google Scholar; cf. Grosvenor, , Constantinople, ii. 698.Google Scholar

^{page 165 note 1} According to the Jardin des Mosquées (p. 73) the mosque measures 66 × 51 paces and has forty-two vaults.

^{page 165 note 2} Brooks, , in J.H.S. xviii. 186Google Scholar: Bury, , Later Roman Empire, ii. 311.Google Scholar Abu Sufian was the title of the Caliph Moawiya.

^{page 165 note 3} This procedure is not uncommon at other Moslem shrines (cf.Evans, , J.H.S. xxi. 204)Google Scholar: the principle involved is that of contact with sacred objects at second hand.

^{page 165 note 4} Hammer-Hellert, , Hist. Einp. Ott. xv. 261.Google Scholar; cf. Jardin des Mosquées (ibid. xviii. 73)

^{page 165 note 5} Travels, i. 1, 25.

^{page 165 note 6} Ibid. i. 1, 167.

^{page 166 note 1} See the Jardin des Mosquées (xviii, c) in Hammer-Hellert, op. cit. xviii, pp. 18 (185, Hassan Hussein Mesjidi), 33 (333, Kahriyeh Djami), 35 (349, Hodja Mustafa Djamisi).

^{page 166 note 2} Carnoy, and Nicolaïdes, , Folklore de Constantinople, p. 116.Google Scholar This tradition seems to be subsequent to the Jardin des Mosquées, in which the saint of this mosque is called a ‘Companion of Eyoub.’

^{page 166 note 3} See especially Evliya, , Travels, i. 2, 15.Google Scholar

^{page 166 note 4} Ibid. i. 26.

^{page 166 note 5} ‘Graves of the Arabs in Asia Minor,’ in B.S.A xix. 182–190.

^{page 166 note 6} For a Moslem saint of this sort discovered in 1845 near Larnaca, see Ross, , Reisen nach Kos, etc., p. 188.Google Scholar

^{page 166 note 7} ProfWhite, (of Marsovan), in Trans. R. Vict. Inst. xxxix. 155.Google Scholar

^{page 167 note 1} B.S.A. xix. 189.

^{page 167 note 2} Biliotti, , Rhodes, p. 501.Google Scholar The name Arab Koulesi is at least as old as Beaufort, (Piloting Directions for Mediterranean, 1831, p. 300)Google Scholar, whose survey took place in 1811. The Moawiya tradition I cannot find before Biliotti.

^{page 167 note 3} In the less reputable field of brigandage the recent exploits of certain redoubtable ‘Arabs’ are still locally remembered (cf. Georgeakis and Pineau, , Folklore de Lesbos, 323Google Scholar; Deschamps, E. in Tour du Monde, 1897, 183 (Cyprus)).Google Scholar

^{page 168 note 1} Tsoukalas, , Πєριγραφὴ Φιλιππουπόλєως, p. 27.Google Scholar

^{page 168 note 2} Polites, , Παραδόσєις ii. 677.Google Scholar

^{page 168 note 3} Λαογραφία, i. 690.

^{page 168 note 4} For the ‘Arab’ in Turkish folk-stories, see Kunos, , Türkische Volksmärchen aus Stambul, preface, p. xviiiGoogle Scholar; for the Greek area, where he is generally called Αράπησ (Μῶροσ in the Ionian Islands, Σαρακηνὸσ in Crete), see Polites, , Νєοελλ. Μυθολογία, pp. 133Google Scholar, 145 ff., and Παραδόσεισ Nos. 419 ff., with the learned note on 419; also Carnoy, and Nicolaïdes, , Folklore de C.P., p. 149.Google Scholar The ‘Arab’ appears early in Greek folklore as the famulus of a sorcerer; see an anecdote of Photius in Bury's, Later Roman Empire, p. 445.Google Scholar

^{page 168 note 5} Polites, Παραδόσєις, Nos. 419–445 inclusive; Pashley, , Crete, ii. 39Google Scholar; Cockerell, , Travels, p. 151Google Scholar; StClair, and Brophy, , Residence in Bulgaria, p. 55Google Scholar; Turner, W., Tour in Levant, iii. 512Google Scholar; Perrot, , L'île de Crète, pp. 103 ff.Google Scholar

^{page 168 note 6} Polites, op. cit. Nos. 455–462; cf. Hobhouse, , Travels, i. 529Google Scholar (haunted houses); Palgrave, Ulysses, p. 59 (haunted bath). In Egypt a talisman which prevented the silting up of a branch of the Nile in the eighteenth century took the form of a negro with a broom (Lucas, , Voyage fait en 1714, p. 339).Google Scholar

^{page 168 note 7} Polites, op. cit. No. 433 (= Leo Allatius, De Graec. opin.), 166 and references given in the note (p. 1108); Lawson, , Modern Greek Folklore, p. 276.Google Scholar

^{page 168 note 8} This is strongly brought out by the Turkish folk-stories (Kunos, loc. cit.).

^{page 168 note 9} The well-known Bridge of Arta story affords a good illustration (Polites, Παραδόσєις, No. 169, note, and 481–483 incl.; also in Νέος Ελληνομνη;´μων i. 39: Sainéan, in Rev. Et. Rel. xlv. 359 ff.). The story occurs all over the Balkan area and as far east as Kourdistan (M. Sykes, Dar-ul-Islam, p. 160).

^{page 168 note 10} For the immolation of a human victim with this object στοιχειώνω see Polites, Παραδόσєις, No. 424 with the note, and 483. The ghost-guardian must be appeased with blood by the finders of the treasure (ibid. No. 404).

^{page 169 note 1} E. G. the guardian of the treasure at the Roman baths called after her Αράπισσα at Sparta (B.S.A. xii. 407) and the ghost Αραπατζέλλα of the Kamares cave in Crete (Folklore, xxiv. 359).

^{page 169 note 2} The porphyry head built into the castle of Roumeli Hissar is said to be that of an Arab woman petrified for mocking the workmen (Grosvenor, , Constantinople, i. 168)Google Scholar, but this is hardly a parallel.

^{page 169 note 3} Above, p. 167.

^{page 169 note 4} With this compare Dev Euren, ‘Ruin of the Ogre’ another figure familiar to folk-tale (Von Diest, , Tilsit nach Angora, p. 38).Google Scholar

^{page 169 note 5} Ainsworth, , Travels, ii. 5, 6.Google Scholar

^{page 169 note 6} e.g. in the fortress commanding the bridge at Chalkis, andat Athens (Dodwell, , Tour, i. 305Google Scholar; cf. Kambouroghlous, , Ἱστ. Ἀθηναίων, iii. 125).Google Scholar

^{page 169 note 7} Ramsay, (Pauline Studies, p. 182Google Scholar) comments on the fact that ancient sites frequently bear names compounded with kara, none with siakh, though both words mean ‘black,’ from which he infers that the word implies awe or mystery. The difference between kara and siakh is primarily one of language, kara being vernacular Turkish, siakh Persian.

^{page 169 note 8} Mariti, , Travels in Cyprus, tr. Cobham, , p. 41.Google Scholar

^{page 169 note 9} F. W. H.

^{page 169 note 10} Records of the Past, vi. 101.

^{page 169 note 11} G. Weber, Dinair, passim.

^{page 169 note 12} Spratt, , Crete, i. 44Google Scholar; Pashley, , Crete, i. 194Google Scholar; cf. B.S.A. xix. 68.

^{page 170 note 1} Pauline Studies, p. 182.

^{page 170 note 2} On this point see Polites, Παραδόσєις, No. 419, note.

^{page 170 note 3} Smyrnakes, , Ἅγιον Ὄρος, p. 471.Google Scholar

^{page 170 note 4} Mariti, , Travels in Cyprus, tr, Cobham, , p. 41.Google Scholar

^{page 170 note 5} Hackett, , Church in Cyprus, p. 421.Google Scholar For a similar alleged conversion of a Moslem saint to Christianity, see Schiltberger (ed. Hakluyt Soc. p. 40).

^{page 170 note 6} Travels, ed. Cobham, p. 41.

^{page 170 note 7} Hackett, , Church in Cyprus, p. 421Google Scholar; Lukach, , Hdbk. oj Cyprus (1913), p. 47.Google Scholar

^{page 170 note 8} Von Hammer, , Osman. Dichtkunst, i. 214Google Scholar: a Kadri convent named Tourabi Tekke exists at Constantinople (Brown, , Dervishes, p. 317).Google Scholar

^{page 171 note 1} For his legend see Delehaye, in Anal. Bolland, xxvi. 247 ff.Google Scholar

^{page 171 note 2} Latrie, Mas, Trésor de Chronologie, p. 911.Google Scholar

^{page 171 note 3} Prognoma sive presagium Mehemetanorum, dated, by the introductory letter, 1545. The prophecy is also published in the Turkish collections of Lonicerus.

^{page 171 note 4} Das Ausland (Munich), 1828, No. 93, p. 372. It will be noted further that ‘seven’ and ‘twelve’ are mystic numbers.

^{page 172 note 1} Ed. Wright, p. 130; cf. Procopius, , de Aedif., 182 BGoogle Scholar; cf. Schiltberger, , Travels, ed. Hakluyt Society, p. 80Google Scholar and note, and for Mandeville's sources, Bovenschen, in Zeitschr. f. Erdkunde, 1888, p. 211.Google Scholar

^{page 172 note 2} Kizil Elma dicunt esse urbem aliquam forlissimam et munitissimam imperiálem (Georgiewicz' commentary), whence doubtless the anonymous writer in Ausland draws the erroneous inference that ‘Red Apple’ was a synonym for any strong city.

^{page 172 note 3} Travels, tr. von Hammer, i. 1, 53. A Russian pilgrim (Khitrovo, , Itin. Russes, p. 91Google Scholar) notices a statue of Leo the Wise which had this property. For other stories of carbuncles that lighted buildings see King, C. W., Natural History of Precious Stones, p. 239.Google Scholar

^{page 173 note 1} There may be a play on this in a Turkish couplet quoted by Gibb, (Ottoman Poetry, iv. 24).Google Scholar

^{page 173 note 2} ‘Red Apple’ for pomegranate has an exact verbal parallel in the Latin name (Malum Punicum) of the same fruit. The Arabic for pomegranate is rumman, which gives a distinct point if the ‘Red Apple’ means Rome.For the curiosity of the subject I note here that there is a mountain called Kizil Elma Dagh (‘Red Apple Mountain’) in the Troad: the name is not derived from the colour of the mountain, possibly from its shape (as apparently its ancient name Κότυλοσ ‘wine-cup’). Other Kizil Elma mountains are shown in R. Kiepert's map above Bartin in Paphlagonia and near Kestelek on the Rhyndacus.

^{page 173 note 3} Gibb, , Ottoman Poetry, iv. 25Google Scholar, note. The globe on the dome is probably meant.

^{page 173 note 4} Loc. cit.

^{page 174 note 1} Avvisi di Costantinopoli, Venice, 1538 (B.M. 1315 d. 18/2).

^{page 174 note 2} Buondelmonti, , Liber Insularum (1420)Google Scholar, ed. de Sinner, p. 72.

^{page 174 note 3} Schepper, , Missions Diplomatiques, 13 G.Google Scholar In this year the marble lion of the Bucoleon was said to have turned its head away from Europe and towards Asia. Such stories are rather the effect than the cause of superstitious fears.