¹ For versions of the general theory see Den Boer, W., Laconian Studies, 1954, pp. 10 f.Google Scholar; Prakken, D. W., Studies in Greek Genealogical Chronology, 1943, ch. 1;Google ScholarMitchel, F., ‘Herodotus' Genealogical Chronology’, Phoenix 10 (1956), 98 f.CrossRefGoogle Scholar

² Schwartz claimed to find a use of the generation of 33¾ years in his reconstruction of Kastor—Göttingen, 1894. The case for generation dating is basically that in some form it must have existed, but whatever the manipulations of later chronographers, more room should probably be allowed for the oral preservation of chronological details than the orthodox picture seems to admit. Otherwise we have to allow that a wider range of values for a generation was used than seems credible.

³ In order to introduce some method into this examination, I have made particular use of Powell, J. E., A Lexicon to Herodotus, and the entries under and is not used for periods of more than one year in Herodotus.Google Scholar

⁴ So How and Wells, i. 221; Stein, , Herodot. i. 164;Google Scholar Rawlinson, i. 221.

⁵ Mitchel, , op. cit., p. 63, seems to have been the first to realize this, and gives by far the best and fullest discussion of this passage.Google Scholar

⁶ Miller, M., ‘Herodotus as Chronographer’, Klio 46 (1965), 113, emphasizes the recognizable pattern obtained in multiples of 33¾ which leads her to believe that at 2. 142.2 Herodotos is using the generation for the first time and gets his figures wrong. But rather it can be used as an argument that Herodotos did not calculate with this generation value.CrossRefGoogle Scholar

⁷ The extra three years may be the three years following Kyros' overthrow of Astyages and the beginning of his reign, since he was the avenger of the Heraklids, and Herodotos mades it look as if the fall of Sardis comes early in his reign.

⁸ I ignore the problem of Dionysos' date. The emendation of 1,600 to 1,000 does not affect the issue. I also realize that my discussion ignores the possibility that ‘900’ is based on a thirty-year generation and ‘800’ on a forty-year generation. What I give is what seems to me the simplest solution. There are other periods too which I have discussed under ‘three generations to a 100 years’ which could be multiples of 30 and 40.

⁹ How and Wells, ad loc., make the same point but according to an arithmetical reckoning.

¹⁰ Twenty generations before Dareios according to the Spartan king lists takes us back to Herakles.

¹¹ There is an unaccountable contradiction involved with both these periods and the main account of the Egyptian kings. See How and Wells, i. 440. [See now Lloyd, , loc. cit.)Google Scholar

¹¹ This is the only fragment of Thrasyllos -FGrHist 253. For Ptolemy of Mendes see FGrHist 611. He is just as unknown a figure. Cf. RE 23.2, No. 74, 1861–2.

¹³ Cf. the Thirty Years' Peace (Th. 1.23.4) and an alliance between colonists to Naupaktos, and Opuntians, (IG ix. 1.334.11 f.)Google Scholar

¹⁴ This is surely preferable to supposing a generation of twenty-three years cf. Miller, M.Klio 41 (1963) p. 32.CrossRefGoogle Scholar |But cf. Forrest, W. G., CQ 19 (1969), 101.|CrossRefGoogle Scholar

¹⁵ Mitchel, F., op. cit.,Google Scholar gives a useful discussion of passages involving levkai and argues similarly that no precise values are implied for a generation [cf. Lloyd, , op. cit., esp. p. 180].Google Scholar

¹⁶ Seventeen for generation d; nineteen for references in terms c generations.