OF THE ROMAN EMPIRE. 233 dred yards broad, and a strong chain could be occa- CHAP. • • XVII sionally drawn across it, to guard the port and city '_ from the attack of an hostile navy". Between the Bosphorus and the Hellespont, the The Pro- shores of Europe and Asia, receding on either side, en- P°" '^' close the sea of Marmara, which was known to the an- cients by the denomination of Propontis. The naviga- tion from the issue of the Bosphorus to the entrance of the Hellespont is about one hundred and twenty miles. Those who steer their westward course through the middle of the Propontis, may at once descry the high lands of Thrace and Bithynia, and never lose sight of the lofty summit of mount Olympus, covered with eter- nal snows °. They leave on the left a deep gulf, at the bottom of which Nicomedia was seated, the imperial residence of Diocletian; and they pass the small islands of Cyzicus and Proconnesus before they cast anchor at Gallipoli; where the sea, which separates Asia from Europe, is again contracted into a narrow channel. The geographers who, with the most skilful ac- The Hel- curacy, have surveyed the form and extent of the '^*'^°" " Hellespont, assign about sixty miles for the winding course, and about three miles for the ordinary breadth of those celebrated straits p. But the narrowest part of the channel is found to the northward of the old Turkish castles between the cities of Sestus and Aby- " See Ducange, C. P. 1. i. part i. c. 16, and his Observations sur Villehar- douin, p. 289. The chain was drawn from the Acropolis, near the modern Kiosk, to the tower of Gaiata ; and was supported at convenient distances by large wooden piles. ° Theveuot (Voyages au Levant, part i. 1. i. c. 14.) contracts tiie mea- sure to one hundred and twenty-five small Greek miles. Belon (Observa- tions, 1. ii. c. 1.) gives a good description of the Propontis, but contents himself with the vague expression of one day and one night's sail. When Sandys (Travels, p. 21.) talks of one hundred and fifty furlongs in length as well as breadth, we can only suppose some mistake of the press in the text of that judicious traveller. P See an admirable dissertation of M. d'Anville upon the Hellespont or Dardanelles, in the Memoires de I'Academie des Inscriptions, tom. xxviii. p. 318 — 346. Yet even that ingenious geographer is too fond of supposing new, and perhaps imaginary measures, for the purpose of rendering ancient writers as accurate as himself. The stadia employed by Herodotus in the description of the Euxine, the Bosphorus, etc. (1. iv. c. 85.) must undoubt- edly be all of the same species : but it seems impossible to reconcile them cither with truth or with each other.
