[68:1] The excerpt alluding to Polybius (printed in his text as vi. 2, 2) merely asserts that in the book of Aristodemus of Elis it was stated that no victors were recorded till the twenty-eighth Olympiad, when Corœbus the Elean won and was recorded as the first victor; from which time the Olympiads were then reckoned. Aristotle is reported to have called Lycurgus the founder (fr. 490). Aristodemus was later than Hippias (cf. above, p. 58); and still _it is to his book, and not to old registers_, that the Greek writers refer. The recurrence of the 28th as an improper Olympiad shows that this number had some important place in the whole discussion. I think it likely that Corœbus really belonged to the twenty-eighth after 776, and not to that year. The oldest actual record of a victor which Pausanias could find was from Ol. 33, and this he describes as of extraordinary antiquity. Other details are given in the Appendix. [69:1] _Fasti Hell._, vol. ii. p. vii. [69:2] Cf. above, § 30, note. [71:1] I incline, with Mr. Bent, to place the remains of Santorin before those of Hissarlik, even though they may be in some respects superior in development. As is obvious, the culture of one place need not keep pace with that of another. But the total disappearance from the legends of any mention of the eruption which must have disturbed the whole Ægean Sea, compared with the living memories of Troy, is to me a proof that the latter and its destruction must be far more recent than the former. Mr. E. Abbott, who refers to Bent's _Cyclades_, is disposed to the other view (_History of Greece_, i. 43); and so are Duruy (vol. i. chap. ii. § 1) and Holm. [71:2] Many writers put the Dorian immigration and the resulting changes of population, and emigration to Asia Minor, in the gap. [72:1] i. 131. Busolt, as he informs me, now agrees with this view. [72:2] In two remarkable articles (_Hellenic Journal_ for 1890 and 1891). [73:1] The date of Archilochus, the earliest of the historical figures among Greek poets, used to be fixed about 709 B.C. The researches of Gelzer, _Das Zeitalter des Gyges_, make it certain that this date is wrong, and must be reduced to at least 670 B.C.; for Archilochus names Gyges in an extant fragment, and Gyges appears on a cuneiform inscription as the vassal of an Assyrian king whose time is determinable. Moreover, an eclipse which Archilochus mentions seems to be that in April, 647 B.C., which was total at Thasos, where the poet spent his later years. Even the conservative Duncker (vol. ii. p. 175, English ed.) adopts these arguments. Nevertheless, some recent histories still acquiesce in the exploded date!