thenes having been determined by itineraries, and the latitude of Thule being known from Pytheas, as the same number of degrees from the pole that the tropics were from the equator, this distance was determined. The most important of these parallels was that of Rhodes, since upon this he measured the entire length of the world. He reckoned his measurements of longitude from Cape Sacrum in Iberia (now Cape St. Vincent), this being considered on a parallel with the Straits of Gades (Gibraltar), the Straits of Sicily, Rhodes, the Gulf of Issus (eastern extremity of the Mediterranean), the Caspian Gates (mountain passes south of the Caspian Sea),and Thinæ on the Eastern Ocean. Not having here any astronomical data, his longitudes are less correct than his latitudes. The principal points established are Carthage, the Straits of Sicily, and Rome, which he erroneously places on the same meridian, Alexandria and Rhodes, which he also places on the same meridian, Issus, the Caspian Gates, the source of the Indus, and the mouth of the Ganges. Though using a plane chart, he yet recognized the fact that a degree of longitude on the parallel of Rhodes had not the same value as at the equator, but was as 555 to 700, which is very nearly correct. Dividing now the distances from Cape Sacrum, which he gives us only in stadia, by 555, we find for Carthage a longitude of 21° 15' 40", which is only two degrees in excess of its true longitude; for Alexandria we have 45° 35' 8", which is an excess of between six and seven degrees; for Issus we have 54° 35' 40", making the Mediterranean too long by between nine and ten degrees; for the Indus the longitude of 100° 10' 48" is between twenty-three and twenty-four degrees too great; while for the Ganges 126° 7' 34" is an excess of between forty-five and forty-six degrees. These excesses, it will be seen, increase uniformly toward the east, as they would by using too short a measure for the degree; and since Eratosthenes expressly states that the degree at Rhodes is only four fifths of that at the equator, it has been conjectured that he has used stadia of different values. An argument in favor of this is that, if we use a stadium of such value that there would be 700 to a degree (as at the equator), the length of the Mediterranean would be very nearly correct, nearer indeed than upon any subsequent map down to the eighteenth century a. d., while the mouth of the Indus would be within three degrees of its true longitude. Knowing Eratosthenes's correctness upon other points, one can hardly resist the conviction that he did use stadia of different lengths, and that Strabo and Pliny have failed to quote his statement and explanation of the fact. We are further confirmed in this opinion when we consider that his age believed the inhabitable world to be very nearly twice as long as it was broad, and that this estimate makes its length to its breadth as 106 to 54.
Nevertheless, we have given here a representation of his map, in which 555 stadia, of the same length with the stadia of latitude, are allowed to each degree of the parallel of Rhodes.
