In his edition, Diels has given up all attempt to arrange the fragments according to subject, and this makes his text unsuitable for our purpose. I think, too, that he overestimates the difficulty of an approximate arrangement, and makes too much of the view that the style of Herakleitos was “aphoristic.” That it was so, is an important and valuable remark; but it does not follow that Herakleitos wrote like Nietzsche. For a Greek, however prophetic in his tone, there must always be a distinction between an aphoristic and an incoherent style. See the excellent remarks of Lortzing in _Berl. Phil. Wochenschr._ 1896, pp. 1 sqq. Footnote 331: Both Bywater and Diels accept Bergk’s λόγου for δόγματος and Miller’s εἶναι for εἰδέναι. Cf. Philo, _leg. all._ iii. c, quoted in Bywater’s note. Footnote 332: The λόγος is simply the discourse of Herakleitos himself; though, as he is a prophet, we may call it “the Word.” It can neither mean a discourse addressed to Herakleitos nor yet “reason.” (Cf. Zeller, p. 630, n. 1; Eng. trans. ii. p. 7, n. 2.) A difficulty has been raised about the words ἐόντας αἰεί. How could Herakleitos say that his discourse had always existed? The answer is that in Ionic ἐών means “true” when coupled with words like λόγος. Cf. Herod. i. 30, τῷ ἐόντι χρησάμενος λέγει; and even Aristoph. _Frogs_, 1052, οὐκ ὄντα λόγον. It is only by taking the words in this way that we can understand Aristotle’s hesitation as to the proper punctuation of the fragment (_Rhet._ Γ 5. 1407 b 15; R. P. 30 a). The Stoic interpretation given by Marcus Aurelius, iv. 46 (R. P. 32 b), must be rejected altogether. The word λόγος was never used like that till post-Aristotelian times. Footnote 333: I have departed from the punctuation of Bywater here, and supplied a fresh object to the verb as suggested by Gomperz (_Arch._ i. 100). Footnote 334: Cf. Herod, i. 8. The application is, no doubt, the same as that of the last two fragments. Personal inquiry is better than tradition. Footnote 335: