the most part, he depends on the oral statements of well-informed persons, both for the older history of Greece and for the Medika. In barbarian countries he was largely dependent on mere dragoman-knowledge, and the careless talk of the Greek quarter of the town.
His frequent expressions, "the Libyans say,", "the Cyreneans say," seem to refer either to the results of his own inquiries in the country referred to, or to the direct statement of some native. Four times we have a personal authority given (iii.55; iv.76; viii.65; ix.16). "Archias whom I met at Pitane" gives the story of his grandfather; Tymnes, the steward of Ariapeithes, verifies some genealogies; Thersander of Orchomenus, who had dined with Mardonius in Thebes, and Dikaios of Athens, who had lived in exile among the Medes together with Demaratus the Spartan king, vouch respectively for two stories which tell at least of the troubled nerves among the following of Mardonius. A more important source of knowledge lay in the archives of various families and corporations: sometimes, perhaps, Herodotus was allowed to read the actual documents; more often, probably, he had to question the men who possessed them. That would be the case, for instance, with the Delphic oracle, to whose records he plainly owes an immense amount, especially in the earlier books. He draws from the traditions of the Alcmaeonidae (Pericles), the Philaidae (Miltiades), and probably from those of the Persian general Harpagos.
The weakness of these sources may be easily imagined. In his Spartan history Herodotus knows all about Lycurgus, who was of course a fixed saga-figure; then
