¹ This article is an outgrowth of an edition of the Iliad. In this connection I gratefully acknowledge a Minor Grant made by the American Council of Learned Societies for clerical and research assistance.

^1a Or, possibly, some hyperlogician wished the adjective to be in agreement with αμα, and not with the nearer appositive íχώρ.

² For my restricted use of the term interpolation compare Athetized Lines 123 n.;cf. also Cauer, Grundfr. d. horn. Kritik³ 118. It is impossible to classify (and so date) lines 340–2 definitely as long as we are restricted to grounds of intrinsic probability. I consider the ascription of them to some λvτιcóς of the 5th century the most probable explanation of their authorship.

³ Or íχώρ is mystically substituted for αμα by abstinence.

⁴ Plato, Theaetetus 176b.

⁵ Compare Franklin Edgerton, The Bhagavad Gītā 2.28 (Harvard Oriental Series 31, 1944).

⁶ Nillson, A History of Greek Religion 203.

⁷ Class. Rev. 12.247–8 (1898). He ends with an emendation (χαρ) which is accepted by Verral, but brushed off by Lawson.

⁸ I may quote from the NED the phrase: ‘The azure ichor of this elite of the earth’, Ford, Handbk. Spain 1.295.

⁹ Compare Burnet 200 f., 231.

¹⁰ Burnet 27.

¹¹ Strictly speaking, χαρ is not attested. In Aeschylus Suppl. 850 χαρ stands in a meaningless jumble of letters; and the scholiast (M) says τρ πιθνμíαν χαρ ∊π∊ν. That is the source of Theognostus, from whom the equation χαρ = πιθνμíα passed (via Hermann on the passage) to Wackernagel; compare below. Headlam demanded χαρ on morphological grounds (χαρ: íχανν = μχαρ: μηχανν) and we can now see that a short vowel in ablaut with ā(i)—compare below—would be most unusual. Headlam is rightly followed by Liddell and Scott.

¹² They may or may not prove the point. If so, λδωp has (like ταλάωρ) changed its gender under the influence of some synonym. The second quotation is hopelessly corrupt.

¹³ For a review of the earlier stages of the problem compare Bechtel, Lexilogus zu Homer 182–3 (Halle, 1914).

¹⁴ More or less confused in the tradition with íσχανάω.

¹⁵ A misunderstanding of the word is seen in Proclus on Plato's Cratylus lxxxv (39.15 Pasquali): λyϵι, yάρ τιs αύτν (poetarum) άχνας τος πνητας, παρά τν στρησιν το χϵω íδíωs οτως καλσας. Such ‘popular etymology’ led to Attic χηνíα (Aesch. Aristoph.) and to the form of some Alexandrian poet found in Hesychius' gloss ϵχveς' πνητϵς.

¹⁶ I may add: αρ (akin to Skt. asr-k, Lat. asser) is glossed both with αμα and with ψvχή by Hesychius. For Empedocles the blood is the seat of intelligence.

¹⁷ In Benveniste, Origines de la formation des noms en indoeuropéen (Paris, 1935) it does not seem to be treated. No mention of χώρ is made; in E 340 χωρ, olov is apparently read. The length of the vowel is marked in χaρ, though an acute accent is also printed.

¹⁸ There is also the type χωρ -oρoς (ό) ‘dandruff’ of less immediate interest for my problem. I am indebted to Professor Buck for information that will soon be generally accessible in his work on Greek nominal suffixes.

¹⁹ Has something of the sort happened at an earlier period in the case of δωρov: Lat. donum?

²⁰ Compare Bechtel, Lexilogus 274–6.

²¹ Note that when Apollo quotes to Zeus these two lines (363–4), he halves the emotional rhetoric by changing the second to:

ούδ κϵ μηνύσαιμ‘, ούδ‘ ν μήνυτρον άροíμην.

In the first passage it may be questioned how far the asyndeton extends. In the next lines (265–6) the MSS have oτϵ ... ούδ', which evidently needs correction. Allen-Halliday (following Baumeister) change to normal syntax oόδ ... ούδ', while Hermanns ο τι ... ούδ' is more in keeping with the immediate context. If that is judged desirable, ο τι ... oύκ may be considered, or (with least change) oτϵ ... οτ': the two lines paired, but joined asyndetically to what precedes.