[263] Plato makes Socrates say that he took up the works of Anaxagoras, hoping to learn whether the earth was round or flat (Phaedo, 46, Stallb. i. 176). In Plutarch’s dialogue “On the face appearing in the orb of the moon,” one of the characters is lavish in his ridicule of the sphericity of the earth and of the theory of antipodes. See also Lucretius, De rerum nat., i. 1052, etc., v. 650; Virgil, Georgics, i. 247; Tacitus, Germania, 45.
[264] That extraordinary picture could, however, hardly have been intended for an exposition of the actual physical geography of the globe.
[265] Aristotle, De caelo, ii. 15.
[266] Archimedes, Arenarius, i. 1, ed. Helbig. Leipsic, 1881, vol. ii. p. 243.
[267] The logical basis of Eratosthenes’s work was sound, but the result was vitiated by errors of fact in his assumptions, which, however, to some extent counterbalanced one another. The majority of ancient writers who treat of the matter give 252,000 stadia as the result, but Cleomedes (Circ. doctr. de subl., i. 10) gives 250,000. It is surmised that the former number originated in a desire to assign in round numbers 700 stadia to a degree. Forbiger, Handbuch der alten Geographie, i. 180, n. 27.
[268] The stadium comprised six hundred feet, but the length of the Greek foot is uncertain; indeed, there were at least two varieties, the Olympic and the Attic, as in Egypt there was a royal and a common ell, and a much larger number of supposititious feet (and, consequently, stadia) have been discovered or invented by metrologists. Early French scholars, like Ramé de l’Isle, D’Anville, Gosselin, supposed the true length of the earth’s circumference to be known to the Greeks, and held that all the estimates which have come down to us were expressions of the same value in different stadia. It is now generally agreed that these estimates really denote different conceptions of the size of the earth, but opinions still differ widely as to the length of the stadium used by the geographers. The value selected by Peschel (Geschichte der Erdkunde, 2d ed., p. 46) is that likewise adopted by Hultsch (Griechische und Römische Metrologie, 2d ed., 1882) and Muellenhof (Deutsche Alterthumskunde, 2d ed., vol. i.). According to these writers, Eratosthenes is supposed to have devised as a standard geographical measure a stadium composed of feet equal to one half the royal Egyptian ell. According to Pliny (Hist. Nat., xii. 14, § 5), Eratosthenes allowed forty stadia to the Egyptian schonus; if we reckon the schonus at 12,000 royal ells, we have stadium = 12,000/40 × .525m = 157.5m. This would give a degree equal to 110,250m, the true value being, according to Peschel, 110,808m. To this conclusion Lepsius (Das Stadium und die Gradmessung des Eratosthenes auf Grundlage der Aegyptischen Masse, in Zeitschrift für Aegypt. Sprache u. Alterthumskunde, xv. [1877]. See also Die Längenmasse der Alten. Berlin, 1884) objects that the royal ell was never used in composition, and that the schonus was valued in different parts of Egypt at 12,000, 16,000, 24,000, small ells. He believes that the schonus referred to by Pliny contained 16,000 small ells, so that Eratosthenes’s stadium = 16,000/40 × .450m = 180m.
It is possible, however, that Eratosthenes did not devise a new stadium, but adopted that in current use among the Greeks, the Athenian stadium. (I have seen no evidence that the long Olympic stadium was in common use.) This stadium is based on the Athenian foot, which, according to the investigations of Stuart, has been reckoned at .3081m, being to the Roman foot as 25 to 24. This would give a stadium of 184.8m, and a degree of 129,500m. Now Strabo, in the passage where he says that people commonly estimated eight stadia to the mile, adds that Polybius allowed 8⅓ stadia to the mile (Geogr., vii. 7, § 4), and in the fragment known as the Table of Julian of Ascalon (Hultsch, Metrolog. script. reliq., Lips., 1864, i. 201) it is distinctly stated that Eratosthenes and Strabo reckoned 8⅓ stadia to the mile. In the opinion of Hultsch, this table probably belonged to an official compilation made under the emperor Julian. Very recently W. Dörpfeld has revised the work of Stuart, and by a series of measurements of the smaller architectural features in Athenian remains has made it appear that the Athenian foot equalled .2957m (instead of .3081m), which is almost precisely the Roman foot, and gives a stadium of 177.4m, which runs 8⅓ to the Roman mile. If this revision is trustworthy,—and it has been accepted by Lepsius and by Nissel (who contributes the article on metrology to Mueller’s Handbuch der klassischen Alterthumswissenschaft, Nordlingen, 1886, etc.),—it seems to me probable that we have here the stadium used by Eratosthenes, and that his degree has a value of 124,180m (Dörpfeld, Beiträge zur antiken Metrologie, in Mittheilungen des deutschen Archaeolog. Instituts zu Athen, vii. (1882), 277).
[269] Strabo, Geogr., ii. 5, § 7; the estimate of Posidonius is only quoted hypothetically by Strabo (ii. 2, § 2).
[270] Pliny, Hist. Nat. ii. 112, 113. There is apparently some misunderstanding, either on the part of Pliny or his copyists, in the subsequent proposition to increase this estimate by 12,000 stadia. Schaefer’s (Philologus, xxviii. 187) readjustment of the text is rather audacious. Pliny’s statement that Hipparchus estimated the circumference at 275,000 stadia does not agree with Strabo (i. 4, § 1).
[271] The discrepancy is variously explained. Riccioli, in his Geographia et hydrographia reformata, 1661, first suggested the more commonly received solution. Posidonius, he thought, having calculated the arc between Rhodes and Alexandria at 1-48 of the circumference, at first assumed 5,000 stadia as the distance between these places: 5,000 × 48 = 240,000. Later he adopted a revised estimate of the distance (Strabo, ii, ch. v. § 24), 3,750 stadia: 3,750 × 48 = 180,000. Letronne (Mém. de l’Acad. des Inscr. et Belles-Lettres, vi., 1822) prefers to regard both numbers as merely hypothetical illustrations of the processes. Hultsch (Griechische u. Römische Metrologie, 1882, p. 63) follows Fréret and Gosselin in regarding both numbers as expressing the same value in stadia of different length (Forbiger, Handbuch der alten Geographie, i. 360, n. 29). The last explanation is barred by the positive statement of Strabo, who can hardly be thought not to have known what he was talking about: κἄν τῶν νεωτέρων δὲ ἀναμετρήσεων εἰσάγηται ἡ ἐλαχίστην ποιόυσα τὴν γῆν, οἵαν ὁ Ποσειδώνιος ἐγκρίνει περὶ ὀκτωκαίδεκα μυριάδας οὖσαν, (Geogr., ii. 2, § 2.)