Strabo, Geography
H.C. Hamilton, Esq., W. Falconer, M.A., Ed.

Further, the length of the inhabited earth is measured on a line parallel with the equator, as it is in this direction that its greatest length lies: in the same way with respect to each of the continents, we must take their length as it lies between two meridians. The measure of these lengths consists of a certain number of stadia, which we obtain either by going over the places themselves, or roads or ways parallel thereto. Polybius abandons this method, and adopts the new way of taking the segment of the northern semicircle comprised between the summer rising and the equinoctial rising. But no one ought to calculate by variable rules or measures in determining the length of fixed distances: nor yet should he make use of the phenomena of the heavens, which appear different when observed from different points, for distances which have their length determined by themselves and remain unchanged. The length of a country never varies, but depends upon itself; whereas, the equinoctial rising and setting, and the summer and winter rising and setting, depend not on themselves, but on our position [with respect to them]. As we shift from place to place, the equinoctial rising and setting, and the winter and summer rising and setting, shift with us; but the length of a continent always- remains the same. To make the Don and the Nile the bounds of -these continents, is nothing out of the way, but it is something strange to employ for this purpose the equinoctial rising and the summer rising.