B. xi. c. IT. 17- GEORGIA. 227 of the Euxine Sea, and the extreme boundary of naviga- tion, for in this sense we are to understand the proverbial saying, " To Phasis where ships end their course." Not as if the author of the iambic intended to speak of the river, nor of the city of the same name upon the river, but Colchis designated by a part, because from the city and the river there remains a voyage of not less than 600 stadia in a straight line to the recess of the bay. This same Dioscurias is the commencement of the isthmus lying between the Cas- pian Sea and the Euxine. It is a common mart of the nations situated above it, and in its neighbourhood. There assemble at Dioscurias 70 or, according to some writers who are care- less in their statements, 1 300 nations. All speak different languages, from living dispersed in various places and with- out intercourse, in consequence of their fierce and savage manners. They are chiefly Sarmatians, but all of them Cau- casian tribes. So much then respecting Dioscurias. 17. The greater part of the rest of Colchis lies upon the sea. The Phasis, 2 a large river, flows through it. It has its source in Armenia, and receives the Glaucus, 3 and the Hippus, 4 which issue from the neighbouring mountains. Vessels ascend it as far as the fortress of Sarapana, 5 which is capable of contain- ing the population even of a city. Persons proceed thence by land to the Cyrus in four days along a carriage road. 6 Upon the Phasis is a city of the same name, a mart of the Colchians, bounded on one side by the river, on another by a lake, on the third by the sea. Thence it is a voyage of three or two 7 days "to Amisus and Sinope, on account of the softness of the shores caused by the discharge of rivers. 8 The country is fertile and its produce is good, except the rS)V ovrwv p,sfi, or careless of the truth. Kramer observes that these words are inconveniently placed in the Greek text. 2 The Rion. 3 The Tschorocsu. 4 The Ilori. 5 Choropani. 6 The point of embarkation on the Cyrus (the Kur) is supposed to have been Surham, the ancient Sura. 7 Gossellin, Groskurd, and Kramer, all agree that there is here an error. Kramer is of opinion that the conjecture of Gossellin may be adopted, viz. " eight or nine," instead of " three or two," the letters T and B being a corruption of H and 9. 8 Coray's proposed reading is adopted, KOTO, for KOI. Q 2