51 Sir John Maundeville, who visited Constantinople in 1322, remarks on the diminution of the empire: ‘For he was emperor of Romania and of Greece, of all Asia the Less, and of the land of Syria, of the land of Persia and Arabia, but he hath lost all but Greece’ (Early Travels in Palestine, p. 130).

52 Cant. ii. 9, 14, 15; Greg. ix. 10, xiii. 3; Ducas, vi.

53 Ducas, i. 6.

54 Cant. iv. 3.

55 Vol. vii. p. 30, edition of Dr. J. B. Bury. The Tartars were still in the Balkan peninsula, and Orchan in 1347, probably just after the marriage of John, sent six thousand Turks to aid Matthew, son of Cantacuzenus, in fighting against the kral of Serbia.

56 Greg. xxvii. 49.

57 Cant. iv. 5 and 6.

58 ἕνεκα ἀσφαλείας πράττειν, iv. 3.

59 Even Gibbon (vii. 30) says, ‘It was in the last quarrel with his pupil that Cantacuzenus inflicted the deep and deadly wound which could never be healed by his successors and which is poorly expiated by his theological dialogue against the prophet Mahomet.’ But the Moslems, both from the north and south, had been fighting in Europe fifty years earlier, sometimes on the side of the Greeks, oftener, as with the Catalans, against them.

60 Heyd’s History of Commerce in the Levant.

61 The Black Death (πανούκλα) was the terrible disease which spread throughout Europe and depopulated most of its large cities between 1346 and 1370. Cantacuzenus, whose son Andronicus fell a victim, gives a vivid and terrible picture of its symptoms, and of its effect upon the population (iv. 8). Dr. Mordtmann, who is not merely distinguished as an archæologist well acquainted with the Byzantine writers, but as a physician of great experience, believes it to have been a black form of smallpox, and not what is usually known as plague, and a well-known specialist in plague, to whose attention I have submitted the account of Cantacuzenus, is disposed to accept the same view.

62 The walls of Galata, both before and after this enlargement, which doubled the area of the city, may still be traced.

63 The demand for slaves, and especially for girls for the harems, was always great. Slaves, indeed, usually formed the most valuable part of the booty in a raiding expedition.

64 Cant. iv. 39.

65 Ibid. iv. 37.

66 Ibid. iv. 37.

67 The statement that he visited Italy and Germany is made by Ducas (i. 11), but it is remarkable that Cantacuzenus makes no mention of it. Muralt (p. 640) suggests that he left Tenedos in the spring of 1352. But Cantacuzenus, writing of the events of 1254, represents John as having passed a whole year in Tenedos. Possibly this would be a year terminating in January 1355.

68 Gregoras, xxix. 25.

69 Rayn. iv. lxiii.

70 iv. 9.

71 The History of Nicephorus Gregoras, as written by an enemy, is a useful corrective. Krumbacher in his account of Byzantine literature speaks of Gregoras as ‘die Hauptperson des 14. Jahrhunderts’ (p. 19). His narrative is described by Cantacuzenus as stamped with ignorance, partiality, and falsehood. Its chief accusation against him is not merely false but improbable (iv. 24). In his own History Cantacuzenus declares that he has never departed from the truth either on account of hatred or the desire to say pleasant things (iv. concluding chapter). What he finds most fault with in Gregoras is the statement that, even during the lifetime of Andronicus, Cantacuzenus had become possessed of a burning desire to become emperor, and that he had consulted certain monks at Mount Athos who were supposed to have the power of divination, in order to learn whether he would accomplish his desire. The story, he declares, is absolutely false. It is brought up because he as emperor protected Palamas in his religious controversies where Gregoras took the opposite side.

72 iv. 9.

73 iv. 24.

74 iv. 28.

75 iv. 17.

76 Greg. xi. 10.

77 Ibid.

78 The Bogomils still exist in Eastern Rumelia. One may be sceptical as to the doctrines in which, according to their enemies, they believed. Apparently they were quietists, searchers after the Inner Light, who would have nothing to do with the worship of Eikons, were possibly Unitarians, and had a tendency in many directions towards what may be called reformation principles. Their teaching was imbued with the Slavic mysticism which is characteristic to-day of Russian literature.

The Bogomils became first noticeable in Bulgaria in the days of King Peter (927–968). Even a few years earlier they are alluded to as certain ‘Pagan Slavs and Manichaeans.’ Later on the Bogomils are spoken of as Paulicians. In Bosnia they became so powerful that the whole country was described as Bogomil. The pope in 1407 promised help to Sigismund against the ‘Manichaeans and Arians’ in Bosnia, and they were beaten and the kingdom dismembered in 1410–11. The Council of Bâle received a deputation from the Bogomils in 1435 and dealt at the same time with them and with the Hussites. In 1443 they lent valuable aid to Hunyadi against the Turks. Persecuted by both the Catholic and Orthodox Churches, many of the magnates who had been forced to become Catholics in order to retain their lands turned Mahometans, and their example was largely followed by the smaller landholders. Among the Mahometans of Bosnia there still exist many customs of Christian origin. Mr. Evans, in Through Bosnia and Herzegovina, states that there are still many thousands of Bogomils in these countries. Herr Asboth, who has been over the country, declares the statement to be too general, and says that he was never able to find any, although he admits that they recently existed. Subject in Bulgaria to persecution from the Orthodox Church, many of them sought escape about a century ago by joining the Church of Rome. Bogomilism spread from Bosnia into Europe, where it gave rise to the Cathari or Albigenses, who acknowledged the Church of Dragovitza in Macedonia as their mother Church. The best account I know of the Bogomils in Bosnia is in J. de Asboth’s Official Tour through Bosnia and Herzegovina, London, 1900.

79 Vol. vii. p. 87.

80 Raynaldus, N. xxxii., professes to give the text of his submission. If his text is genuine it shows that John was under the same delusion as Michael had been: namely, that he could force the Orthodox Church to accept what he wanted.

81 Ducas, xii. Chalcondylas makes a similar statement (i. 45); Canale says that a Genoese doctor restored sight to Andronicus.

82 Sauli, Colonia dei Genovesi in Galata, ii. 260.

83 Urban the Sixth died in 1389.

84 Ducas, xiii.

85 This was in 1097, when, on the invitation of Godfrey de Bouillon, Alexis had reached the city on its water side by taking his boats, in part at least, overland from the Gulf of Moudania to the lake. The object of Godfrey was to prevent the Crusaders being exposed to the demoralisation of plundering a hostile city.

86 Greg. ix. 2 says the Turks had carried off three hundred thousand Christian captives. The Turks fought well, but were exterminated.

87 Cant. iv. 16.

88 Cant. iv. 39.

89 I reserve my description of the Janissaries for a later chapter.

90 Ch. xxiii.

91 Chalc. i. 51, and Phrantzes, i. 11.

92 Du Cange, Familiae Dalmaticae, 230, Venetian edition. The story of this battle is fully described in Die Serben und Türken im XIV. und XV. Jahrhundert of S. Novakovich (Semlin, 1897) and also in Ireček’s History of the Bulgarians (p. 430). Ireček states that as late as the seventeenth century the stone monument of the despot Uglisha’s tomb still existed. Uglisha was one of the three brothers.

93 Chalc. i. 44 says that the sultan immediately beheaded his son; Ducas, that Countouz was blinded (xii.).

94 Cossovo-pol, the Plain of Blackbirds, is between Pristina and Prisrend, to the north-east of Uskub. The town of Cossovo is due south of Prisrend, and about thirty miles distant.

95 Novacovich, p. 335. ‘Gleichwie durch den Krieg an der Maritza das Schicksal Ost-Bulgariens und der serbischen Staaten in Macedonien, ebenso ist durch die Schlacht aus Kossovopolje, den 15. Juni 1389, das Schicksal der nördlichen serbischen Länder und des westlichen Bulgarien entschieden worden, namentlich der Länder des Fürsten Lazar und Buk Brancovic’s.’

96 Sad-ud-din. See also Halil Ganem’s Les Sultans Ottomans, Paris, 1901. Upon the assassination of Murad the custom grew up, which continued till about 1820, of not allowing any Christian belonging to a foreign state to enter the presence of the sultan except with Janissaries holding each arm.

97 Now called Anatolia-hissar. The word hissar means castle.

98 The version of Ducas differs from those of Chalcondylas and Phrantzes, the first of whom knows nothing of the arrangement suggested, but states that Manuel left the city for Italy, while Phrantzes declares that John, having lost the favour of Bajazed, fled to his uncle, who entrusted the city to him during his absence (Phr. pp. 61–3.)

99 Ducas, xx.; Chalc. iv. p. 183. Phrantzes, p. 89, praises Mahomet very highly.

100 Ducas, xxiii.

101 Mersaite declared he failed because of the presence of a noble lady, evidently the Holy Virgin, walking upon and guarding the walls.

102 According to another version he withdrew on account of the famine and plague which prevailed in his army. It is, however, certain that the Turkish revolt in favour of Mustafa took place, and in the following year, 1423, Murad captured the leader, Elias Pasha, and bowstrung both him and the young Mustafa at Nicaea. Before the end of the year he returned to Thrace and took possession of Adrianople.

103 See ante; and also Pachymer, iii. 10 to iv. 25.

104 ‘The Greek Church has had a fossilised aversion to change; boasting that it follows the doctrines and practices of the Apostolic Church, it believes that it has no need of reform.’ Eighteen Centuries of the Orthodox Greek Church, by Rev. A. H. Hore, p. 553 (Jas. Parker & Co.: London, 1899).

The expression ‘fossilised aversion’ is perhaps too strong, though I should be prepared to admit that the Eastern non possumus was at least as obstinate as the Western. The Orthodox Church in countries where it is free, as in Greece and Russia, shows signs of growth, and therefore hardly deserves the adjective ‘fossilised.’ Since 1453 in Turkey it has been comatose.

105 Milman, History of Latin Christianity, 3rd edition, vol. viii. p. 348.

106 While the rival representatives were in Constantinople Murad suggested to John that his friendship under the circumstances would be of greater value than that of the pope. Chalc., Syropulus, and Phrantzes.

107 Phrantzes, pp. 181–6.

108 Vol. vii. p. 108.

109 Second Council of Nicaea, in 787.

110 The copies sent to London and Karlsruhe, as well as the diptych of Rome (the official record) consulted by Niches, signed by the emperor of Constantinople and by thirty-six Latin prelates, contain on this point only the following: ἔτι ὁρίζομεν τὴν ἁγίαν ἀποστολικὴν καθέδραν καὶ τὸν ῥομαϊκὸν διάδοχον εἶναι τοῦ μακαρίου Πέτρου. The pope and forty-two Latin prelates, on the other hand, signed the following: Item definimus S. Ap. sedem et romanum pontificem in universum orbem tenere primatum et ipsum pontificem romanum successorem esse S. Petri.

111 Many of the towers near the Golden Gate bear inscriptions showing that they were repaired during John’s reign. For the inscriptions see Paspates’ Βυζαντιναὶ Μελέται.

112 Caramania was the Turkish state which remained longest outside Ottoman dominion. At one period it extended from the river Sangarius to Adana. Ordinarily its boundaries did not extend further north than Konia. See Stanley Lane-Poole’s Mohammedan Dynasties, p. 134.

113 The island of Chios had for several years been held by a Commercial Company, mostly if not exclusively of Genoese, each of whose members was, apparently, known by the name of Justiniani.

114 Gibbon suggests, on the authority of the Hist. Anonyme de St-Denys, that the French had murdered their Turkish prisoners on the eve of the engagement, and that the sultan was merely retaliating (Gibbon, vii. 37).

115 Chalc. ii. 807.

116 Chap. xv.

117 The word timour is the same as the ordinary Turkish word for iron, demir.

118 Leunclavius, 250.

119 Leunclavius, pp. 250–1, Ven. edition, makes the conquest of Damascus in 1399; Chalcondylas and others, in 1402; the Turkish authors quoted by Von Hammer, in 1401. The statement of the hindrance due to locusts I take from Muralt, 772, who quotes as his authority ‘Bizar,’ a name unknown to me.

120 The Crescent, which Gibbon and other writers assert to have only been employed by the Turks after the capture of Constantinople, had probably been used by them for many centuries previously. It is true that it had been made use of in Constantinople at an early period, and figures on several coins of Constantine, but I doubt whether it was used as the symbol of Constantinople in the later centuries of its history. The Crusades are not incorrectly described as wars between the Cross and the Crescent. The symbol is an ancient one and figures with the star on several coins belonging to about 200 B.C. The Abassid dynasty so used it. Professor Hilprecht considers it a remnant of moon-worship and connects it with the subsequent cult of Ashtaroth, Astarte, or Aphrodite.

121 Though the Turks were a branch of the Tartar race, the Greek authors by this time had acquired the habit of calling the nation which Othman had formed Turks, and all others from Central Asia Tartars, and it is convenient to follow this nomenclature.

122 Von Hammer has shown conclusively that the story of an iron cage is a mistake. It arises from the misinterpretation of the Turkish word Kafés, which has the two significations given above. Two contemporary authors made the blunder, Phrantzes and Arab Schah. A Bavarian, who was made prisoner at the battle of Nicopolis, named Schildberger, and who was present at the battle of Angora, has given a detailed account of the massacre of the Christians, but he does not mention the cage. (His travels between 1394 and 1427 have been translated and published by the Hakluyt Society, 1879.) Neither do Ducas, Chalcondylas, or Boucicaut, though they state that Bajazed died in irons, which he had to wear every night after his attempt at escape. Six Persian authors who wrote the history of Timour are silent about the cage. The oldest Turkish historian recounts, upon the evidence of an eye-witness, that Bajazed was carried about in a palanquin ‘like a Kafés,’ or in the usual kind of grilled palanquin in which ladies of the harem travelled. Sad-ud-din, one of the most exact of Turkish historians, states that the story of the iron cage given by many Turkish writers is a pure invention.

123 I have relied for the account of the battle of Angora and the subsequent progress of Timour, mainly upon Von Hammer (vol. ii.), who is at his best in describing this period of Turkish history. The authorities are carefully given by him. Zinkeisen, in his History of the Turks, calls attention to the deterioration of the Ottoman armies during the reign of Bajazed, and attributes it to the profligacy of the sultan.

124 Chalc. iv. p. 170. Ducas says he disappeared in Caramania; Phr. p. 86, that he was bowstrung. There was, according to Chalcondylas, another son of Bajazed, the youngest, also named Isa, who was baptised and died in Constantinople in 1417. This was probably the son given over as hostage to Manuel.

125 Ducas, xix.

126 Chalc. iv.; Phr. i. 29; Ducas, 19.

127 Official Tour in Bosnia and Herzegovina, by J. de Asboth.

128 i. 37.

129 Ducas, xxiv.

130 Phr. i. 38.

131 In reference to this passage across the Dardanelles, Ducas (ch. xxvii.) gives an interesting piece of information as to the size of the Genoese vessels. There were seven large ships. Murad was in the largest, which contained 1,300 Turkish and Frank soldiers. These ships ‘covered the sea like floating cities or islands.’

132 Ducas mentions expressly that in the same year three Mustafas died, first, the pretender, who claimed to be the son of Bajazed; second, his brother, and, third, the grandson of Atin (ch. xxviii.).

133 De la Brocquière, whose narrative was finished in 1438, states that, when in Galata, the ambassador of the duke of Milan, the protector of the Genoese, told him that ‘to do mischief to the Venetians he had contributed to make them lose Salonica taken from them by the Turks;’ and he adds, ‘Certainly in this he acted so much the worse, for I have seen the inhabitants of that town deny Jesus Christ and embrace the Mahometan religion.’ Early Travels, pp. 335–6.

134 Halil was the one Turkish leader in 1453 friendly to the Greeks. Even at this early date he showed a similar spirit. Chalc. 136, Venetian edition.

135 Phr. ii. 13, p. 180.

136 Possibly Hungary was not mentioned, with the object of leading the Turks to believe that the place of attack would not be nearer than Constantinople.

137 Callimachus, who describes the battle, took part and was wounded in it.

138 I have followed here the version of Ducas (xxxii.). It is doubtful, however, whether this expedition into Caramania ought not to be placed a year earlier. See the authorities quoted by Muralt, p. 856.

139 Chal. vi.; Ducas, xxxii. The latter states that Hunyadi refused either to sign or to swear.

140 The treaty was made in June. According to Muralt, it was broken in the same month. If so, the account of Ducas is incorrect. Murad was informed by George of Serbia of the renewal of war and again took the government into his own hands ‘at the beginning of summer, when the dog-days were commencing.’ Ducas, xxxii.

141 Early Travels, pp. 346–347.

142 Lonicerus, p. 18, speaking of the cardinal, does not go so far. He says, ‘qui Pontifici licere juramenta praesertim hostibus Christiani nominis praestata rescindere contendebat.’ Thurocz (quoted by Von Hammer, p. 307, vol. ii.) and Cambini, p. 13, make similar statements.

143 Liber Jurium, xxii. 57, xxvi. 24, 26. Chalc. vi. Aeneas Silvius states that Eugenius, when he was informed of the treaty, wrote to Cardinal Julian that it was null as having been signed without the papal sanction; that he ordered Ladislaus to disregard it, and that he gave him absolution for so doing. At the same time, he directed the cardinal to do his best to renew the war, in order that the great preparations he had taken in hand might not be fruitless. The statement may be true, but it is difficult to believe that the report of the signature could have reached Rome and that his answer could have arrived to the cardinal before war was declared.

144 The Turkish accounts agree that the crossing was at the Bosporus. Barletius, Book II. p. 38, with whom Leunclavius agrees, says: ‘Si vera est fama,’ merchant vessels transported the army over the Bosporus, receiving a gold coin per man. Bonfinius likewise gives this story of payment and says it was made to the Genoese. Lonicerus, p. 18, says the fleet crossed the Dardanelles. Ducas, whose account I have adopted, states that the fleet only crossed with great difficulty and against the will of the emperor. Chalcondylas makes the transit take place at Hieron, near the Dardanelles (Chalc. 135); one writer, at Asomaton. There is a church of the Asomatoi (the Bodiless, i.e. of Angels) at Arnaoutkeui still existing. See The Constantiade, where the Patriarch gives an account of it. Phrantzes identifies the position on the Bosporus (namely, opposite Anatolia-Hissar) by saying that it was near the narrow part of the Bosporus above the village of Asomaton or Arnaoutkeui: κατὰ τὸ στενὸν ἐγγὺς τοῦ ἀνωτέρου μέρους τῆς τῶν Ἀσωμάτων κώμης (Ph. ch. II. p. 223), which is conclusive as to the locality he wishes to indicate. Ducas also in several places gives the name of Hieron to the straits between Anatolia and Roumelia-Hissar. It is therefore clear that two places on the Bosporus were known as Hieron. The safest passage would be at the Hieron below the Giant’s Mountain.

145 Callimachus.

146 ‘Morbo detentus,’ Lonicerus, 18. Chalc. and others also mention his illness. He was suffering from an abscess in the thigh.

147 On the opposite shore of the lagoon now runs the railway from Varna to Rustchuk.

148 Early Travels, 361.

149 Early Travels, 366.

150 Chalc. p. 138. The account by Phrantzes, p. 198, of the interview between Hunyadi and the king is very well given.

151 Bonfinius states that it was at this moment also that he unfurled the treaty of Szegedin.

152 Leunclavius, 256.

153 Eton’s Travels, p. 332.

154 Gibbon adopts the statement of Chalcondylas (145) that Murad joined the dervishes after Varna, though on other matters regarding his life he relies upon Cantemir, who by implication discredits the story. Chalcondylas states that in the crisis of the battle of Varna, the sultan had vowed that if he were successful he would abdicate and join one of these religious orders. Von Hammer knows nothing of the story, and the whole course of Murad’s life is against the belief that ‘the lord of nations submitted to fast and pray and turn round in endless rotations with the fanatics who mistook the giddiness of the head for the illumination of the Spirit’ (Gibbon, VII. p. 140). Neither Phrantzes nor Ducas mentions his having become a dervish, as they probably would have done if the fact had been known to them. Indeed, the one point in favour of the story was unknown to Gibbon: namely, that some of the dervish sects are liberal or philosophical. They are all religious or pietistic, but many claim that their tenets are independent of Islam. Their explanation of the turning or dancing is that they first look towards Mecca and reflect, God is there; then they make a turn and reflect, He is there also; and so in the complete circle. It should be noted also that there are many dervishes who neither turn nor dance in their devotions. On the subject of the dervishes in Turkey, two useful books are The Dervishes, by J. P. Brown (London, 1868), and, better still, Les Confréries Musulmanes par le R. P. Louis Petit, supérieur des Augustins de l’Assomption à Kadikeuy (Constantinople, 1899).

155 Kroya or Croia, now called Ak-Hissar or the White Castle, is a few miles to the north of Durazzo and a short distance from the Adriatic.

156 Aeneas Sylvius gives the number at 200,000; Chalcondylas at 15,000, which Von Hammer reasonably suggests is an error for 150,000.

157 Bonfinius makes Murad state in a letter to Corinth that eight thousand Hungarians were left dead on the plain: a much more likely number.

158 Von Hammer gives the numbers I have adopted.

159 For the siege of Belgrade see a paper in the English Historical Review, 1892, by Mr. R. N. Bain.

160 ‘Novit majestas imperatoria, Turcorum, Assyriorum, Aegyptiorum gentem: imbelles, inermes, effaeminati sunt, neque animo neque consilio martiales; sumenda erunt spolia sine sudore et sanguine.’ Oratio Romae habita anno 1452 de passagio Cruce signatorum contra Mahometanos suscipiendo. Edita apud Reynaldum [by Dr. Dethier].

161 La Brocquière, 366.

162 Θρῆνος, line 720.

163 According to Scholarius and Manuel the Rhetorician, John shortly before his death declared against the Union. In such a matter, however, both these witnesses are suspect.