The breadth of the earth, from south to north, is commonly supposed to be about one-half only of its length, viz. 4490 miles; hence it is evident how much the heat has stolen from it on one side and the cold on the other: for I do not suppose that the land is actually wanting, or that the earth has not the form of a globe; but that, on each side, the uninhabitable parts have not been discovered. This measure then extends from the coast of the Æthiopian ocean, the most distant part which is habitable, to Meroë, 1000 miles[746]; thence to Alexandria 1250; to Rhodes 562; to Cnidos 87-1/2; to Cos 25; to Samos 100; to Chios 94; to Mitylene 65; to Tenedos 44; to the promontory of Sigæum 12-1/2; to the entrance of the Euxine 312-1/2; to the promontory of Carambis 350; to the entrance of the Palus Mæotis 312-1/2; and to the mouth of the Tanais 275 miles, which distance, if we went by sea, might be shortened 89 miles. Beyond the Tanais the most diligent authors have not been able to obtain any accurate measurement. Artemidorus supposes that everything beyond is undiscovered, since he confesses that, about the Tanais, the tribes of the Sarmatæ dwell, who extend towards the north pole. Isidorus adds 1250 miles, as the distance to Thule[747]; but this is mere conjecture. For my part, I believe that the boundaries of Sarmatia really extend to as great a distance as that mentioned above: for if it were not very extensive, how could it contain the innumerable tribes that are always changing their residence? And indeed I consider the uninhabitable portion of the world to be still greater; for it is well known that there are innumerable islands lying off the coast of Germany[748], which have been only lately discovered. The above is all that I consider worth relating about the length and the breadth of the earth[749]. But Eratosthenes[750], a man who was peculiarly well skilled in all the more subtle parts of learning, and in this above everything else, and a person whom I perceive to be approved by every one, has stated the whole of this circuit to be 252,000 stadia, which, according to the Roman estimate, makes 31,500 miles. The attempt is presumptuous, but it is supported by such subtle arguments that we cannot refuse our assent. Hipparchus[751], whom we must admire, both for the ability with which he controverts Eratosthenes, as well as for his diligence in everything else, has added to the above number not much less than 25,000 stadia.