of the monsoon was realized at a still earlier date. And the Periplus distinctly indicates the southward trend of the Indian coast-line. Roman penetration of Africa gave Ptolemy some new details; he also conceived the Nile as formed by two headstreams arising in two lakes, possibly on the strength of some hearsay of the facts, and he marked the Mountains of the Moon in remoter Africa, which again suggests hearsay of the heights of Ruwenzori, Kenya, and others. The Romans had penetrated Ethiopia, and possibly the region of Lake Chad, and Ptolemy also used other sources of information about North Africa which are unknown from previous writers, but are completely vague and impossible to follow. Of the shape of the continent he was almost completely ignorant; he just realized that an indentation occurs in the Gulf of Guinea, but gave it nothing like its proper value, and carried the coast thence south-westward till the continent is broader at the southern limit of his knowledge than it is at the north. Ptolemy’s work on physical features was on the whole poor, and he neglected the human side of geography. Discarding the idea of the circumfluent ocean, he supposed the extension of unknown lands northward in Europe, eastward in Asia, and southward in Africa, beyond the limits in which he attempted to portray their outlines; and he even suggested a land connection between south-eastern Asia and southern Africa. Before his time the precision of mathematical method had far surpassed that of the topographical material to which it was applied. Pausanias, a Greek probably of Lydia and about contemporary with Ptolemy, wrote a description (_Periegesis_) of Greece, which, apart from the archæological value which is its chief interest, contains references to various phenomena of physical geography, while as a detailed topographical work it stands alone in the literature of which an outline has thus far been given. CHAPTER III. THE DARK AGE