370 STRABO. CASAUB. 602. warm spring has failed, but the cold spring flowing from the Scamander along a subterraneous channel emerges at this place ; or, because the water was near the Scamander, it was called the source of that river, for there are several springs, which are said to be its sources. 44. The Andirus empties itself into the Scamander; a without his knowledge in 1793, an acquaintance with these two springs, which present nearly the same phenomena as described by Homer. These springs have since been seen by many travellers ; they are situated at the foot of a small hill on which is Bounar-bachi, and about 6500 toises in a straight line from the mouth of the Mendere. The stream which flows from them never fails, and after having run for some time parallel to the Mendere", it turns suddenly to throw itself into the Archipelago, near the middle of the interval which separates the ruins of Alexandria- Troas from the cape Koum-kale, but still leaving traces of a bed through which it formerly flowed to join the Mendere. We are now convinced that this little river is the Scamander of Homer, that the present Men- dere is the Simois of that poet, and that the ancient Ilium, which was near the sources of the Scamander, must have been situated on the heights of Bounar-bachi. In the time of Homer these two rivers united together and discharged themselves into the sea by the same mouth : but the course of the Sca- mander has been changed for a long time, since, according to Pliny, (v. c. 33,) a part of its waters spread themselves over a marsh, and the re- mainder flowed unto the ^Egsean Sea, between Alexandria-Troas and Sigeum. This ancient author therefore gave to the little river (which he called Paloescamander, the old Scamander) exactly the same course which the stream Bounar-bachi still follows. This change of direction in the course of the river appears to me to have been anterior to the time of Demetrius of Scepsis, for this alone can explain his error. For, no longer finding a stream which runs on the left of the present Mendere, and which might represent the Scamander, he thought proper to transfer this latter name to the Simois, and to look for the site of the Ilium of Homer, as also of the plain which was the scene of the combats de- scribed by the poet, on the right of this river. Thence he is persuaded that the town of the Ilienses occupied the same site as the ancient Ilium, and that the stream of the Tschiblak was the Simois. I must remark that the Mender^ is a torrent, the waters of which fail during a great part of the year, whilst the stream of the Bounar-bachi always continues to flow. This advantage is probably the reason why it preserved the name of Scamander to the sea, although it ran into the bed of the Simois and was far inferior to this torrent in the length of its course. Hence it may be perceived how the name of Scamander, now changed into that of Mendere, has remained attached to this ancient mouth, how ultimately it was given to the whole course of the Simois, and how De- metrius of Scepsis was led into error by the change in the course of the true Scamander, and by the transfer of its name to the Simois. Voyage Pittoresque de la Grece par M. de Choiseul Gouffier. Le Voyage dans la Troad, par If. Lechevalier. The Topography of Troy, W. Gell. Gossellin.
