¹ J.H.S. 1909, pp. 41–52.

² In particular, the number of tolerably certain references by Herodotus is singularly small. There are but two passages in which Herodotus' remarks on questions of geography can be assumed with any confidence to apply to Hecataeus rather than to anybody else (ii. 21, and iv. 36, where the theory of Ocean which Herodotus comments on may fairly be identified with that of Hec. fr. 278). There is no proof whatever that the ‘Ionians’ and other predecessors against whom Herodotus polemises in his Egyptian λόγοι stand for Hecataeus.

³ Readers whom the bold language of the opening passage in the Γϵνϵηλογίαι (fr. 1) may have unduly impressed may have their prejudice corrected by a glance at fr. 341.

⁴ The numbering of the fragments in this article is that of the 1841 edition by C. Müller.

⁵ Diels, , Hermes, 1887, pp. 418–9Google Scholar.

⁶ The following clear instances of this ambiguity of authorship occur among the first hundred items: fra. 19, 25, 32, 43, 49, 58, 61, 90, 92.

⁷ Susemihl, , Geschichte der griechischen Literatur in der Alexandrinerzeit, ii. chs. 35Google Scholar, 37 38.

⁸ Kleine Schriften, i. p. 124.

⁹ The lacuna on pp. 93–4 of the codex (pp. 79–80 of C. Müller's edition) is clearly not large enough to have entirely swallowed Scylax' allusions to the eastern seas, had he made any.

¹⁰ Niebuhr, op. cit. p. 110, shows that the name Scylax was not uncommon in Caria.

¹¹ Susemihl, op. cit. ii. 677–8.

¹² Herodian, (Πϵρὶ Καθολικῆς Προσψδίας ii. 925. 6Google Scholar) shows that one single treatise of Scymnus, the Πϵρίπλους Ἀσίας, ran into 10 volumes.

¹³ 842 b 27.

¹⁴ τὰς ἀλεκτορίδας τῷ μεγέθει πάντων εῖναι μικροτέρας τῶν ὀρνίθων The passage from the De Anima vi. 1. 558 b, also quoted by Mr. Wells, stands in no connexion with fr. 58 of the Γῆς Περίοδος as its contents are markedly different.

¹⁵ Hec. fr. 172. ῾ Εκαταῖος ὁ Μιλήσιος ἐν ᾿Ασίας τεριηγήσει εἰ γνήσιον τοῦ συγγραφέως τὸ βιβλίον Καλλίμαχος γὰρ Νησιώτου ἀνα γράφει

¹⁶ So Diels, loc. cit. pp. 412–6, whose agnosticism Mr. Wells can hardly claim to have confuted.

¹⁷ E.g. in fr. 46.

¹⁸ Witness the list of ῾ πόλιες in i. 149: Λήρισαι, Νέον Τεῖχος Τῆμνος, Κίλλα, Νότιον and that in vii. 123: Λίπαξος Κώμβρεια, Λίσαι, Γίγωνος Κἁμψα, Σμίλα, Αίνεία Similarly, many of the πόλεις mentioned in the Attic tribute lists must have been quite tiny settlements.

¹⁹ Hellanicus frs. 13, 18–9, 21–5, 27, 119, 121, 128 (ed. C. Müller). Pherecydes frs. 116–7. Philistus frs. 9, 14, 18–9, 21, 27, 29, 31–2, 38–9, 40, 43. Attention may also be drawn to a luminous essay by Scluchardt, C. on the earlier Greek conception of a πόλις (Neue Jahrbücher für dan klassische Altertum, 1909, pp. 305–321)Google Scholar, where it is shown that πόλις originally meant ‘fortress,’ but soon came to denote any settlement without reference to size or political status.

²⁰ Tropea, G., Rivista di Storia Antica, 1897, p. 89Google Scholar.

²¹ I.G. i. 237, 239, 240, 242–3, 257, 259. The variant reading Θρᾴκιος φόρος is also found.

²² vii. 72.

²³ Appian, , Bell. Civ. iv. 72Google Scholar. Similarly fr. 108 need not be called into question because Cynus in Opuntian Loeris is therein called a πόλις, and not an ἐπίνϵιον, as with Pausanias (x. 1. 2). The depopulation of Greece in the interval between Hecataeus' and Pausanias' era sufficiently accounts for the discrepancy.

²⁴ Smyth, , Ionic, p. 86Google Scholar.

²⁵ The indiscriminate use of the expressions Περιήγησις Λιβύης and Περιήγησις ᾿ Ασίας in frs. 268–329 shows that Libya is hero reckoned as part of Asia, and that the author of the Γῆς Περίοδος only recognises two approximately equal continents—Asia and Europe. The Oceanic theory mentioned above is derived from Schol. Apol. Rhod. iv. 259: ῾ Εκαταῖος δὲ ὁ Μιλήσιος λέγει ἐκ τοῦ Φάσιδος διελθεῖν εἰς τὸν Ωκεανόν, εἶτα ἐκεῖθεν εἰς τὸν Νεῖλον

²⁶ iv. 36 and ii. 21.

²⁷ The passages in Herodotus do not convey the inclusion of Libya in Asia, nor the possibility of through travel between the Phasis and the Nile.

²⁸ The Πϵρίπλους of Scylax is particularly interesting in this connexion. Whereas the coast of southern Italy receives a full treatment, the description of the remaining coast-land to the Straits of Gibraltar is at once very deficient and incorrect: even Massilia and its colonies seem hardly known to the author.

²⁹ Cf. Macan, , Herodotus iv.–vi., vol. ii. pp. 277–8Google Scholar.

³⁰ It is significant that the two chief events in the extrusion of the Greeks from the farther Mediterranean, the battle of Alalia (535 c.) and the failure of Dorieus' schemes of colonisation (515 c.) fell in the early days of Hecataeus' life.

³¹ Scylax, writing apparently about 350 B.C., possesses some knowledge about the whole length of the eastern Italian coast (ehs. 14–19).

³² Cf. Prašek, , Klio, 1904, pp. 206–7Google Scholar. This scholar clearly goes beyond his evidence in attributing to the author of the Γῆς Πϵρίοδος an extensive knowledge of Iran. But he shows convincingly that Herodotus' information was less complete, and that Herodotus must have drawn upon the Γῆς Πϵρίοδος rather than vice versa.

³³ Fr. 179 Κασπάπυρος πόλις Γανδαρικἠ Σκυθῶν ἀκτή) mentions a town which Marquart, (Philologus, Suppl. x. p. 242)Google Scholar identifies with the starting point of navigation on the Indus and thus makes into a natural place for Scylax to have seen and remembered. Marquart further mentions that this reach of the Indus has a Turanian population, thus justifying its description as ‘Scythian.’ The meaning of ἀκτή is hard to determine: perhaps it signifies ‘fraction,’ ‘enclave’ (=ἀπορρώξ)

³⁴ The substratum of truth in this tale is presumably to be found in some ceremonial mummery and mimic fighting practised by the Pygmies. The distortion of the real facts in fr. 266 is of the same character as in various passages of Herodotus' Egyptian λόγοι, and probably is due to a similar inspiration on the part of the author's native Egyptian informants.

³⁵ Cf. Diodorus i. 37; Strabo, pp. 769–770, 789; C.I.G. 5127 (the Adule inscription); Mahaffy, , The Empire of the Ptolemies, pp. 151–2, 215–6Google Scholar.

³⁶ Tropea, loc. cit. p. 83.

³⁷ Loc. cit. p. 419.

³⁸ The addition ῾ ἔστι καὶ πόλις may be taken from its form and position in the text to stand outside the quotation and to have been appended by Stephanus.

³⁹ Cf. the ῾ ´ Ελληνες Σκύθαι in Hdt. iv. 17, whom Macan (note ad loc.) regards as a half-breed population.

⁴⁰ Boryza may be regarded as a counterpart of Doriscus, which Herodotus (vii. 59) describes as a τεῖχος βασιλήϊον in the days of Xerxes

⁴¹ Hicks, and Hill, , Greek Historical Inscriptions, Nos. 25 (1. 48) and 44Google Scholar.

⁴² Head, , Historia Numorum, p. 89Google Scholar. In this instance there can be no doubt that the form with δ is the more archaic of the two.