^{page 134 note 1} p. 102.

^{page 134 note 2} I am especially indebted to ProfessorFränkel's, HermannParmenidesstudien (Berlin, 1930)Google Scholarand to ProfessorGomperz's, HeinrichPsychologische Beobacktungen an griechischen Philosophen (Leipzig, 1924)Google Scholar.

^{page 134 note 3} On Kranz's arrangement, now accepted by Diels (4th ed. I. p. xxviii). Most probably right, except that frs. 2 and 3 should belong to the 'Aλ⋯θια.

^{page 134 note 4} Assuming Cornford to be right in protesting against the view that fr. 4 implies only two ways. The fragment in itself is clearly incomplete, and is continued by fr. 6.

^{page 134 note 5} It is generally conceded that the δ⋯ξα has strong Pythagorean features; but only for those who hold to the view that the δ⋯κρανοι are Heracleiteans and these alone is this an obstacle to relating the δ⋯ξα with fr. 6.

^{page 134 note 6} Cf. 1, 27; 16, 2; 19, 3.

^{page 134 note 7} Cf. 1, 30; 6, 4; 8, 39, 51, 61.

^{page 134 note 8} 1, 27.

^{page 134 note 9} These are they who traverse the Way of Night mentioned in 1, 11. It is the way back into the δώματα Nυκτ⋯ς; and it has no goal (see below).

^{page 134 note l0} This is certainly the implication of the phrase; cf. 1, 3, where P., the seeker after truth, is already εἰδώς because he knows the goal, the goal on which his will is already set (1, 1—σον τ' ⋯π⋯ θυμ⋯ς ἱκ⋯νει).

^{page 135 note 1} κρῖναι λ⋯γψ (1, 36).

^{page 135 note 2} 1, 35. ἠχ⋯εσσαν goes with γλ⋯σσαν as well as with ⋯κου⋯ν.

^{page 135 note 3} Fr. 8 (Diels).

^{page 135 note 4} Cf. Stenzel, , Metaphysik des Altertums, p. 55Google Scholar.

^{page 135 note 5} The three together: 4, 7 (γνο⋯ης, φρρ⋯αις, ⋯⋯ν). 6, 1 (λ⋯γειν, νοεῖν, ⋯⋯ν). 8, 8 (φατ⋯ν, νοητ⋯ν, ἒστιν). 8, 17 (⋯ν⋯ητον, ⋯νώνυμον, ο⋯κ ἔστιν). 8, 35 (⋯⋯ν, πεφατισμένον, νοεῖν).

^{page 135 note 6} Fr. 1, 32: ώς τ⋯ δοκο⋯ντα | χρ⋯ν δοκιμ⋯ς εἶναι. There is no need whatever to emend. Why it is important Parmenides tells us in 8, 61.

^{page 135 note 7} The number-theories of the Pythagoreans may have been an exception; but Parmenides would still have regarded them as mere hypothesis, i.e. as δ⋯ξα, and therefore not metaphysics.

^{page 136 note 1} Parmenides' thought becomes much clearer if his manner of stating arguments be noted: he very frequently begins with the conclusion, and then states in inverse order the arguments which lead up to it. Each γ⋯ρ- or έπε⋯-clause refers to the clause immediately preceding it, and not generally to the whole argument. The fragment in question affords an excellent example: its last clause is the starting-point of the argument, its first line the conclusion. In subsequent paraphrases the argument will be stated, for clarity's sake, in the more normal, un-Parmenidean order.

It is noteworthy that this method of inverse statement runs through the whole poem. E.g. (1) the prologue begins by mentioning P.'s arrival at the goal, and then goes back to describe the way to it. (2) The Aλ⋯⋯θεια begins by stating the σ⋯ματα to be established, and then gives arguments for them one by one.

^{page 136 note 2} ν⋯ημα=τ⋯ νοεἰν, as in 8, 34.

^{page 136 note 3} πολυπλ⋯κτων merely emphasizes the paradox that out of the world of becoming νο⋯ς can arise.

^{page 137 note 1} This is proved in II. 26–33. There are two main σ⋯ματα (οὖλον, ⋯κ⋯νητον, I. 38), amplified respectively by μουνογενές (what is whole must be single) and ⋯τ⋯λεστσν (what does not move cannot end). That ⋯τ⋯λεστον does not mean τετελεσμ⋯νον (Patin, followed by H. Gomperz) is proved by 1. 27, where ἂναρχσν, ἂπαυστον correspond exactly to ⋯τ⋯λεστσν here.

^{page 137 note 2} These four divisions correspond to the four entities proved in II. 34–41 to be mere ὂνομα:(a) γ⋯γνεσθα⋯ τε κα⋯ ὂλλυσθαι.(b) εϊνα⋯ τε κα⋯ οὐχ⋯. (c) κ⋯νησις. (d) ⋯λλο⋯ωσις. (See below.)

^{page 137 note 3} That is, metaphysically speaking, Becoming amplified and Perishing do not exist. So best, though other translations are possible, Diels' correction of μ⋯ to πῃ (to correspond with Simplicius' paraphrase, , Phys. 78)Google Scholar is inadmissible on account of the abruptness which it would involve, and of P.'s use of ⋯⋯ν throughout the 'Aλ⋯θεια in the single sense of absolute Being.

Note the future ⋯φ⋯σει together with the aor. ⋯ν⋯κε: P.'s complete conviction of the rationality of Being. What logic will not admit cannot have been ordained by Destiny.

^{page 138 note 1} But the real ground for P.'s statement of the homogeneity of Being here is probably given in frs. 2–3. which should then come between II. 21–2.

^{page 138 note 2} The same as the Dike of I. 14.

^{page 138 note 3} Since it is καθ' αὑτ⋯ (1.29).

^{page 138 note 4} Deleting μ⋯ (1. 33) with Bergk, Diels, Burnet, etc. Simpl. certainly had μ⋯, and H. Gomperz wishes to keep it, scanning ⋯πιδεες. But there can be no point here in a ference to Notbeing; and the corruption is slight. The argument is that what has no limit must be utterly indeterminate.

^{page 138 note 5} ⋯τελεὐτητον is precisely equal to οὖλ (Cf. 1. 38). This proof that Being is one and whole is the first stage of the proof that it is completely determinate. The second, and last, stage is to show that it is also qualitatively determinded. The distinction between the two is marked by the negative expression of the first conclusion and the positive expression of the second. Before the latter, however, P. inserts a proof of the non-existence (metaphysically speaking) of the infinite and indeterminate. The passage thus falls into three sections: (a) Being is not infinite (26–33). (b)The infinite does not exist (34–41). (c) Being is finite (42–9).

^{page 138 note 6} Literally ‘The same thing is for thinking, and of which the thought is.’ (οὓνεκεν=ὅτι.) There is thinking on the one hand, and the object thought of on the other, and there is nothing besides; i.e. there is no thought without a real object (cf. Plato, , Parm. 132B)Google Scholar. Hence, when people ‘postulate’ (κατατ⋯θενται—I. 39) entities, they are not thinking, but merely playing with words. But even to such a use of language some kind of mental process must correspond; and this is what P. Calls δ⋯θα. Between this and νο⋯ς there is no point of contact at all, since their objects are fundamentally unrelated.

^{page 138 note 7} The text is difficult, but can be restored with probability. We have it in three forms:

Theaet. 180E—οἶον, ⋯κ⋯νητο τελ⋯θει τῷ π⋯ντ ⋯νομ' εἶναι.

This Simpl. quotes verbatim twice (Phys. 29. 143). The manuscripts of both Simpl. and Plato have οἶον and παντ⋯ but it appears from the comments fo both authors on the line that the above is the right reading (Simpl. 29—μ⋯ν; Theaet. 180E—ὠς ἓν τι π⋯ντα ⋯στ⋯).

II. Simpl. phys. 146—οὐσ' εἰ χρ⋯νος. ἔστιν ἠ ἔσται

ἄλλο π⋯ρεξ το⋯ ⋯⋯ντο. ἗πε⋯ τ⋯ γε Mοῖρ' ⋯π⋯δησεν

οὖλον ⋯κ⋯νη⋯ν τ' ὠν⋯μασ ται …

III. Simpl. Phys. 86—οὐδ⋯ν γ⋯ρ ἔστιν ἥ ἔσται π⋯ρεξ (sic)

ἄλλο π⋯ρεξ το⋯ ⋯ντος, ⋯πε⋯ τ⋯ γε Mοῖ' ⋯π⋯δησεν

οὖλον ⋯κ⋯νητ⋯ν τ' ἔμεναι. τῷ π⋯ν' ὃνο' ἔσται.

The last is both unmetrical and stops short in the middle of a sentence; it is aparently either from memory or at second-hand. In either case the watery οὐδ⋯ν γ⋯ρ ἔστν has no weight at all against the οὐδ' ɛἰ χρ⋯νος of II, which is apparently taken direct from a manuscript of Parmenides. As a palpable stopgap it is pointless ot try to emend it. (Preller—οὐδ⋯ν γ⋯ρ <ἢ> … So Diels. Bergk—οὐδ' ἦν γ⋯ρ ἢ …) II is obviously our best authority, and Diels is right is saying that Plato's quotation is free. The ὠν⋯μασται of II is, however, unmetrical and corrupt; I and II taken together show pretty clearly that the true reading here is ⋯ʿμ' ἔσται. I therefore read:

οὐδ⋯ χρ⋯νος ἔσται ἢλλο π⋯ρεξ το⋯ ⋯⋯ντος ⋯πε⋯ τ⋯ γε Mοῖρ' ⋯π⋯δησενοὖλον ⋯κ⋯νητ⋯ν τ' ἔμεναι τῷ π⋯ντ' ⋯νομ' ἔσται ἔσα …

So Ritter and Preller. ἢλλο is neuter adverbial.

^{page 139 note 1} since it is perfectly uniform. Vide Cornford, p.106, n.

^{page 139 note 2} Reading οὐκ ⋯⋯ν.

^{page 139 note 3} There is no reference to shape, rather to the erroneous essumption of the ⋯ρχ⋯-theorists that BEing can become more or less rea;. So I. 5.

^{page 139 note 4} Reading τοιγ⋯ρ. This must be right, for (a) the line is manifestly the conclusion, not the premiss of the argument (as it must be with the manuscript reading). (b) The manuscript reading gives a very lame conclusion for the whole' Aλ⋯θεια.

^{page 139 note 5} A world consisting of both Being and Notbeing is necessarily infinite, since wherever one stops the other begins. For the turn which Parmenides gives to this sort of view see p. 141, n. 5.

^{page 139 note 6} A single species of ⋯λλο⋯ωσις, representing the genus.

^{page 140 note 1} κατ⋯θεντο, πεποιθ⋯τες εἶναι ⋯ληθ⋯ (8, 39).

^{page 140 note 2} Cf. fr. 8, 12–14, and p. 4, n. 3.

^{page 140 note 3} τετελεσμ⋯νον.

^{page 140 note 4} ταὐτ⋯ν δ' ⋯στ⋯ νοεῖν τε κα⋯ οὔνεκ⋯ν ⋯στι ν⋯ημα

^{page 141 note 1} Cf. Soph. 260A–264B, which proves that the object of δ⋯ξα. as of all statement, is something real. Parmenides and Socrates had only been able ot distinguish between δ⋯ξα and ⋯πιστ⋯μη by assigning to them different objects, viz. Becoming and Being. This left Becoming with no status at all, since it neither was not was not; hence the obscurity in Parmenides' view of it: it is mere name and yet it is the world we live in.

^{page 141 note 2} μορφ⋯ς γ⋯ρ κατ⋯θεντο δὐο γνώνας ⋯νομ⋯ζειν (8, 53).

Cf. ⋯σα βροτο⋯ κατ⋯θεντο πεποιθ⋯τες εἶναι ⋯ληθ⋯ (8, 39).

The use of κατατ⋯θεμαι is apparently not the same in both fragments; but the implication of it in both is certainly that the doctrines in question are mere hypothesis.

^{page 141 note 3} οἶς τ⋯ π⋯λειν τε κα⋯ οὐκ εἶναι ταὐτ⋯ν νεν⋯μισται κοὐ ταὐτ⋯ν.

^{page 141 note 4} In the sense in which P. uses ‘be’ throughout the δ⋯ξα. Cf. Fr. 19—οὔτω τοι κατ⋯ δ⋯ξαν ἔφυ τ⋯δε κα⋯ νυν ἔασι.

^{page 141 note 5} Cf. especially εἶνα⋯ τε κα⋯ οὐχ⋯, the assumption of which is made one of the characteristics of a δ⋯ξα-metaphysic in 8, 40. There it referred to alternation of body and void. Parmenides modifies the doctrine, since he holds that even what is ‘between’ cannot be nothing. It only is not in the sense that it is not the other. That is why he insists that the two Forms are equal in status. (ἴσων ⋯μφοτ⋯ρων, 9, 4).

^{page 142 note 1} Diels 18 A 46 (De Sensu 1 sq.).

^{page 142 note 2} There is no need to take Sextus' explication of the allegory very seriously.

^{page 142 note 3} Note that P. speaks from the chariot (φ⋯ρουσι [I, 1] is present tense), but he goes no further; the Gateway is the goal. The was to it is ἤματος κ⋯λευθος νυκτ⋯ς κ⋯λευθος is the way back, and is traversed by those whom Dike does not admit.

^{page 142 note 4} Fr. 2.

^{page 142 note 5} Cf. Fränkel, p. 158, and fr. I, 10: ὠσ⋯μεναι κρ⋯των ἢπο χερσ⋯ καλὺπτρας.

^{page 143 note 1} Aphrodite? Cf. H. Gomperz, p. 20.

^{page 143 note 2} Frs. 17–20.

^{page 143 note 3} Diels 18 A 46. Cf. Plato, , Epist. 7, 342cGoogle Scholar.

^{page 143 note 4} Parm. 129.

^{page 144 note 1} On this vide Fränkel, , p. 158 sqGoogle Scholar.

^{page 144 note 2} Just as Plato's Timaeus is a myth.