Strabo considered that mathematical and astronomical knowledge was indispensable to the science of geography; he says in book i.,[9] that without some such assistance it would be impossible to be accurately acquainted with the configuration of the earth; and that every one who undertakes to give an accurate description of a place, should describe its astronomical and geometrical relations, and explain its extent, distance, latitude, and climate.[10] As the size of the earth, he says, has been demonstrated by other writers, we shall take for granted what they have advanced. We shall also assume that the earth is spheroidal, and that bodies have a tendency towards its centre. He likewise says, the convexity of the sea is a further proof that the earth is spheroidal to those who have sailed; for they cannot perceive lights at a distance when placed at the same level as their eyes, but if raised on high, they at once become perceptible.[11] He also observes, “our gnomons are, among other things, evidence of the revolution of the heavenly bodies, and common sense at once shows us that if the depth of the earth were infinite, such a revolution could not take place.”[12] But Strabo did not consider the exact division of the earth into climates or zones, in the sense in which Hipparchus used the term, and the statement of the latitudes and longitudes of places, which in many instances were pretty well determined in his time, as essential to his geographical description. With regard to the lost continent of Atlantis, Strabo is very cautious in criticising[13] Poseidonius; he observes, “he did well, too, in citing the opinion of Plato, that the tradition concerning the island of Atlantis might be received as something more than a mere fiction, it having been related by Solon, on the authority of the Egyptian priests, that this island, almost as large as a continent, was formerly in existence, although now it had disappeared,” and remarks that Poseidonius thought it better to quote this than to say, He who brought it into existence can also cause it to disappear, as the poet did the wall of the Achivi. The measure adopted by Strabo was the stadium. In book vii. chap. vii. § 4, he says, “From Apollonia to Macedonia is the Egnatian Way; its direction is towards the east, and the distance is measured by pillars at every mile, as far as Cypsela and the river Hebrus. The whole distance is 535 miles. But reckoning, as the generality of persons do, a mile at eight stadia, there may be 4280 stadia. And, according to Polybius, who adds two plethra, which are a third of a stadium, to every eight stadia we must add 178 stadia more,—a third part of the number of miles.” In book xi. chap. xi. § 5, he compares the parasang with the stadium, and states that some writers reckoned it at 60, others at 40, and others at 30 stadia.