46 "TABLEAU de la PLAINE de TROTE 



impettiofity *. Moreover, Mr Liston afTured me, that from 

 M. Chevalier's defcription, the Scamander feems to be a 

 more diminutive water than it really is f . 



" For feveral hours," fays Dr Dallavvay, " we traced, with 

 " the utmoft attention the courfe of the Scamander from the 

 " cold or fecond fource, which is a colledlion of fmall fprings, 

 " through the morafs, where for fome miles it is pofitively hid, 

 " till we reached the new canal, and faw plainly the ancient bed. 

 '* The banks of this river, where expofed, are verdant and beau- 

 " tiful, and watered to the brink. M. Chevalier's topography 

 " and general idea, after a fair invefligation, we acknowledged 

 " to be ingenious and plaufible." (p. 347.). 



In charadlerifing the Scamander, M. Chevalier mentions 

 particularly " the tranfparency of the water, which runs upon 

 ** a bottom of fand and round pebbles, betwixt two verdant 

 " banks " 



Dr Dallaway fays of the two rivers : " SImois has broad 

 " fands, with a fudden and rapid current ; Scamander is tran- 

 " jfparent and regularly full, within a narrow channel, and fo 

 " they continue to be till their jundlion, before they reach the 

 " fea.- (p. 348.). 



M. Chevalier further defcribes his having pafTed the Sca« 

 mander upon an old willow ftretched acrofs, near a mill. Mr 

 Liston alfo mentioned to me this mill, and his having crofled 

 the current in a fimilar manner. 



Tbe 



* If this hypothefis of Mr Liston be well founded, perhaps it may be inferred 

 that the Scamander remains in the fame ftate in which it was in the days of Ho- 

 mer, occafionally flowing into the Simois, but commonly, by what is thought a new 

 canal, into the ^gean Sea. And if this is admitted, it may aflift Mr Heyne in ob- 

 viating a difficulty which occurs to him in his Eflay on the Topography of the Iliad. 

 See Appendix, No, III. 



f Perhaps I may be partly to blame for this, by calling it, in the tranflation, a 

 rivulet, (p. 13. 15.), and once a rill, (p. 2J.)" The original is ruij/eau, which might 

 have been rendered ajlfeam. 



