Martianus Capella

[595] Finally, hemispherical bowls which are called horologia, or sundials, are adjusted according to variations in latitude. Gnomons are replaced at distances beyond five hundred stadia, as the shadows are depressed or elevated according to the location of the dials. Consequently, the longest day of the year at Meroe lasts twelve and two thirds hours, at Alexandria fourteen, in Italy fifteen, and in Britain seventeen. At the summer solstice, when the sun is borne toward the celestial pole, it bathes the regions that lie beneath in continuous daylight; but when it makes its wintry descent, it causes northern regions to shiver in darkness that lasts six months. Pytheas of Massilia reported that he found such a condition on the isle of Thule. Those discrepancies of the seasons, unless I am mistaken, compel us to admit that the earth is round.

[608] But the two regions, or belts, referred to above - one of which is covered with ice because of its proximity to the chill of Plaustrum [the Wain], the other deserted because of antarctic winds - do not have antipodes of their own. However, because of their diametrically opposite locations, they become antipodal to each other. Inhabitants here observe no risings of the celestial bodies, except the planets, which do not course over their heads but have risings near the middle of one side. Fixed stars are visible for six months, and for six months they are out of sight. The equator is the circle that mark risings for them, and only six signs of the zodiac are visible. Days and nights are of six months' duration, and the poles are visible directly overhead. But the one region, adorned with the luster of Septentrio, and the other, with the star Canopus, have no acquaintance with the remaining portions of the sky. [609] That most learned man Pytheas has disclosed what conditions are like in those regions; but I myself have traversed them: there is no portion of the earth's surface that is not known to me.
The full measurement of the earth's circumference - to convert into Roman miles the figure in stadia which I recorded in the calculations above - is thirty-one thousand five hundred miles. At this point, it would be well to include the opinion which Ptolemy records in his Geography.