OF THE EOMAN EMPIRE 251 by the speed and secrecy of his enterprise ; and the dependent cities, as far as Laodicea and the confines of Aleppo,- obeyed the example of the metropolis. From Laodicea to the Thracian Bos- phorus, or arm of St. George, the conquests and reign of Soliman extended thirty days' journey in length, and in breadth about ten or fifteen, between the rocks of Lycia and the Black Sea.**^ The Turkish ignorance of navigation protected, for a v.hile, the in- glorious safety of the emperor ; but no sooner had a fleet of two hundred ships been constructed by the hands of the captive Greeks, than Alexius trembled behind the walls of his capital. His plaintive epistles were dispersed over Europe, to excite the compassion of the Latins, and to paint the danger, the weakness, and the riches, of the city of Constantine.*^* But the most interestmg conquest of the Seljukian Turks w^as state and 1 f r 1 -i-i 1 .1 .i'. r" !• pilgrimase of that oi Jerusalem,^ which soon became the tlieatre or nations. Jemsai ' In their capitulation with Omar, the inhabitants had stipulated the assurance of their religion and property ; but the articles were interpreted by a master against whom it was dangerous to dispute ; and in the four hundred years of the reign of the 82 See Antioch, and the death of Soliman, in Anna Comnena (Alexias, 1. vi. p. i68, 169 [c. 9]), with the notes of Ducange. •HVilliain of Tyre (1. i. c. 9, 10, p. 635) g^ves the most authentic and deplorable account of these Turkish conquests. ^In his epistle to the count of Flanders, Alexius seems to fall too low beneath his character and dignity ; yet it is approved by Ducange (Not. ad Alexiad. p. 335, cSic.) and paraphrased by the abbot Guibert, a contemporary historian. The Greek text no longer exists ; and each translator and scribe might say with Guibert (p. 475), verbis vestita meis, a privilege of most indefinite latitude. [Guibert incorporates the substance of this letter, Recueil, H. Occ. iv. p. 131 sf<^. The best edition of the text (preserved only in Latin) is that of the Count de Riant (1877 and again 1879). . controversy has raged over the genuineness of the document. Riant rejects it as spurious (like Wilken, Raumer, and others). But it was accepted as genuine by Sybel, and has been defended more recently by Vasilievski (Zhurn. Min. Nar. Prosv. 164, p. 325 j^^. 1872) and Hagenmeyer (Byz. Ztsch. vi. i s^t/. 1897). It is doubtless genuine. The objections brought against it are not weighty ; and the critics who condemn it have offered no theory of its origin that is in the least prob- able. It is perfectly incredible that it was composed as a dehberate forgery in the year 1098-9 in the camp of the Crusaders, as Riant tries to establish. Its contents are absolutely inconsistent with this theory. It was probably written long before the First Crusade ; and Hagenmeyer is probably right in assigning it to 1088, when the Empire was in danger from the Patzinaks, and some months after the personal interview of Alexius with Robert of Flanders at Berroea. The letter, of course, has suffered seriously in the process of its translation into Latin.] •* Our best fund for the history of Jerusalem from Heraclius to the crusades is contained in two large and original passages of William, archbishop of Tyre (1. i. c. i-io, 1. xviii. c. 5, 6), the principal author of the Gesta Dei per Francos. M. de Guigncs has composed a very learned M4moire sur le Commerce des Francois dans le Levant avant les Croisades, <ltc. (Mem. de 1' Academic des Ijiscriptions, torn, xxxvii. p. 467-500), em. A.D. 638-1099
