The geographical chapter in Al-Battānī’s _Astronomy_, probably compiled from a redaction of Ptolemy’s _Geography_ and translated by Plato of Tivoli in our period, gave a description of the seas of the world.[1317] The Indian Ocean, Al-Battānī said, extends from the land of the negroes to the extreme limits of India, a distance of 8000 miles. Its width was 2200 miles, of which 1900 (Plato of Tivoli mistakenly translated this 3900) reach south of the equator. What lands lay beyond are not specified. From this sea four gulfs run into the land: first the Barbaric Sea, which extends into the “land of the negroes,” or Ethiopia, and may be the Gulf of Aden or possibly even Mozambique Channel; second, the Green Sea (Mare Viride), or our Red Sea, which reaches towards Hyla (Ailah?); third, the Persian Gulf (Mare Persicum); and, fourth, a second Green Sea, running out to the east towards China (“Thinae”) and representing the Bay of Bengal or possibly the China Sea. In the Indian Ocean there are some 1370 islands, among them a very large one called “Tibiariae” (Taprobane), or Sarandib (Ceylon), opposite the eastern coast of India, 3000 miles in circumference, full of great mountains and rivers, quantities of rubies and hyacinths, and surrounded by fifty-nine lesser isles. The traditional account of the many isles of the Indian seas so persistent in Arabic literature arose unquestionably from familiarity with the vast Malay Archipelago or at least with its western portion. On the other hand, whether we may assume, as some have done, that the exaggerated classical and Arabic estimates of the size of Ceylon had their origin in rumors of the existence of Australia[1318] is an obscure problem which we cannot attempt here to solve. SCYTHIA AND CENTRAL ASIA