very general way the extents of the figures, you ought to adopt some rule. But Eratosthenes takes a line from the Caspian Gates along the mountains, running as it were in the same parallel as far as the Pillars, and then a second line, starting directly from the mountains to touch Thapsacus; and again a third line from Thapsacus to the frontiers of Egypt, occupying so great a breadth. If then in proceeding you give the length of the two last lines [taken together] as the measure of the length of the district, you will appear to measure the length of one of your parallelograms by its diagonal. And if, farther, this diagonal should consist of a broken line, as that would be which stretches from the Caspian Gates to the embouchure of the Nile, passing by Thapsacus, your error will appear much greater. This is the sum of what may be alleged against Eratosthenes. 38. In another respect also we have to complain of Hipparchus, because, as he had given a category of the statements of Eratosthenes, he ought to have corrected his mistakes, in the same way that we have done; but whenever he has any thing particular to remark, he tells us to follow the ancient charts, which, to say the least, need correction infinitely more than the map of Eratosthenes. The argument which follows is equally objectionable, being founded on the consequences of a proposition which, as we have shown, is inadmissible, namely, that Babylon was not more than 1000 stadia east of Thapsacus; when it was quite clear, from Eratosthenes’ own words, that Babylon was above 2400 stadia east of that place; since from Thapsacus to the passage of the Euphrates where it was crossed by Alexander, the shortest route is 2400 stadia, and the Tigris and Euphrates, having encompassed Mesopotamia, flow towards the east, and afterwards take a southerly direction and approach nearer to each other and to Babylon at the same time: nothing appears absurd in this statement of Eratosthenes.