68 EARLY GEOGRAPHICAL SPECULATIONS. [CHAP. chosen for the delimitation of Asia were rivers -the Phasis on the side towards Europe, and the Nile in the direction of Africa. A stream of water at all times suggests itself at first sight as the ' most natural boundary between tracts of land, though a chain of mountains or an isthmus forms a far more certain and definite limit. Hence the Caucasus, though it is mentioned by Aeschylus as the highest of mountains1, and therefore must have been known to the Greeks before his time, was ignored as the dividing line between Europe and Asia ; and the Isthmus of Suez, which forms the natural line of separation between Asia and Africa, was passed over in like manner. At a later time the importance of isthmuses from this point of view was brought prominently forward ; so much so, that the land which intervenes between the Euxine and the Caspian was treated as nn isthmus, and was regarded as the boundary of Asia on that side*; but this principle of division implies a wider survey of the field of geography than was possible at an early age. By some the Phasis continued to _ - .be treated as the limit even in the time of Hero- Boundary of Europe and dotus8; but in the meanwhile the advance of the Asia* Greek colonies in the direction of the Palus Maeotis had initiated another view, and that piece of water together with the Cimmerian Bosporus at its entrance and the Tanais which flowed into it were regarded as separating the two continents— an opinion which ultimately prevailed. , It appears to have beeh adopted by Hecataeus, for he treats the town of Phanagoria, which lay on the eastern side of the Cimmerian Bosporus, as being in jAsiai^' while in Aeschylus both views are represented, for in the Promethew Vinctus the dramatist describes lo as passing from 1 Prom. VwcL 719: vpb car irpbs aMv Katf/caow 8 Strabo, i. 4. 7: rofa 81 rots l