We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 11 February 2009
Barrett has given a brief account of the affiliations of Hn with the manuscripts which he has collated. He derives his information about the readings of Hn from the reports of nineteenth-century editors, and he does not report this manuscript in his apparatus criticus. He concludes that ‘In three instances (72, 641, 817) Haun. has the truth, or an approximation to it, where the rest of our tradition is at fault (though in two of them there are traces of the same reading in O) … in each case the reading can be accounted for as a lucky accident, and so I judge it in fact to be.’ Of the other four manuscripts, which editors have not collated, he gives no account. ‘From a number of readings cited by Turyn it appears that they are all more or less closely related to Haun.’ K. Matthiessen has voiced a mild regret that Barrett did not settle the question by collation. I have collated these five manuscripts from photographs or microfilms.
1 See Turyn, A., The Byzantine Manuscript Tradition of the Tragedies of Euripides (Urbana, 1957), pp. 357–8Google Scholar.
2 See Turyn, pp. 329–33.
3 See Turyn, pp. 351–2.
4 See Turyn, pp. 348–51.
5 See Turyn, pp. 345–8.
6 Barrett, W. S., Euripides: Hippolytos (Oxford, 1964), pp. 75–6Google Scholar.
7 Which I summarise below, p. 39.
8 I discuss these instances below, pp. 39—40.
9 Matthiessen, K., Studien zur Textüberlieferung der Hekabe des Euripides (Heidelberg, 1974), p. 32Google Scholar.
10 They roughly correspond with those adopted (with fuller explanation) by Barrett, pp. 92–3.
11 I report these other manuscripts from my own collations. A fresh collation of manuscripts collated by Barrett might seem wholly superfluous. It did, indeed, confirm, what one expected, that his reports are of almost superhuman accuracy. But it was necessary, because (reasonably enough) Barrett does not report isolated errors in his mss. (nor does he regularly report P). We need the evidence of these isolated errors (and of P) if we are to form a true picture of the affiliations of PvHnOxNvN. I shall show elsewhere (in a discussion of the mss. of Medea) that the isolated errors of C reveal its affiliation to the gnomologium Escorialense (which was not available to Barrett: see below).
Since Barrett's edition was published the following new manuscript material has become available: the gnomologium Barberinianum, c. 1300 A.D. (edited by Matthiessen, K., Hermes 93 (1965), 148–58)Google Scholar, which I designate gB, the gnomologium Escorialense, early fourteenth century (edited by Matthiessen, , Hermes 94 (1966), 398–410)Google Scholar, which I designate gE, and P. Oxy. 3152, second century A.D. (in addition, some new readings in P. Oxy. 2224 are published in Oxyrhynchus Papyri 44 (1976), 34–5)Google Scholar. The gnomologium Vatopedianum, designated by Barrett as ‘gnomol.’, I designate gV. Corrections in L by Triclinius I designate Tr. Otherwise I designate the mss. by the same symbols as Barrett.
12 I shall not report corrections by the second hand of N: for their source, see below, p. 39 n. 18.
13 cf. 823 π⋯λiν] π⋯λαi E.
14 Not earlier, because at 1077 N has δεiκν⋯εi with V, while the Aldine has πην⋯εi with the rest.
15 ‘…a printed text, most probably the Aldine edition’ (p. 348). Of the other printed texts which come into question, we may rule out that of Lascaris (c. 1494), which has fundamental differences from that of the Aldine and N (cf. Sicherl [below, n. 16] 206 n. 5); also the ed. Hervagiana2 (1544) and the edition of Canter (1571), both of which correctly give πρ⋯ϲ for πρ⋯ν at 1238. (Canter has further divergences.) So far as I can see, we cannot rule out the ed. Hervagiana1 (1537) or the ed. Brubachiana1 (c. 1558; I have not investigated the ed. Brubachiana2, c. 1560), whose texts appear to be faithful copies of the Aldine's (n.b. Hervag.1 repeats the Aldine's misprint at 1396 [see below on this page]; it is corrected in Hervag.2 and Brubach.1).
16 It is not, however, a direct copy of either of these mss.: see Sicherl, M., ‘Die Editio Princeps Aldina des Euripides und ihre Vorlagen’, Rh. Mus. 118 (1975), 205–25, esp. 206–12.Google ScholarThe only divergence which I have noticed between LP and the Aldine is 1442 πατρ⋯ NAldBOAV3: πατρŸϲ HnOxNvVCDELPB3. But the reading of LP is unmetrical, and the Aldine editor was capable of restoring the truth without the help of other mss.
17 The first printed editions to correct the Aldine's misprint are the ed. Hervag.2 (of which N cannot be a copy) and the ed. Brubach.1 (of which, in theory, it can): see above, p. 38 n. 15.
18 N2 appears to have derived most of his corrections from the Aldine. In numerous places his corrections agree with unique errors of P reproduced in the Aldine (e.g. 53 τ⋯πων] β⋯μων N2PAld). But that his source was the Aldine, not P, is suggested by 1010 oἰκ⋯ϲεiν codd. et P:-⋯ϲαi N2Ald.
19 See the instances cited above, pp. 37–8. (§5(b) ad fin.). I observe (in view of the agreement of Nv with M at 503) that M is closely related to O (see Barrett, p. 65). N, also, has unique agreements with M at 132 (above, p. 37) and 144 (above, p. 40).
20 Cited on p. 37.
21 The Aldine restored half of the truth: ἔνεπε δ' ἔννεπε. Similarly at 573 ἔνεπε AldOx: ἔνν-codd.
22 The line numbers are Barrett's. The (very fragmentary) evidence of P. Mil. Vogl. 44 for the hypothesis is given by Barrett, pp. 431–2. I have reported this papyrus only where it is available. Where I do not report it, it may be assumed to be not available.
23 See above, pp. 35–6, for their unique agreements in the poetical text.
24 At 1 αἴӨραϲ κα⋯ πoϲεiδ⋯νoϲ MVPNvN: αἴ- ἵππoλ⋯τηϲ κα⋯ π- <B> OACHnOx (αἴ- υἱ⋯ϲ κα⋯ π- B2): π- κα⋯ α1F34;- D: π- E.
25 See above, p. 36.
26 See above, p. 38.
27 ‘Verba τ⋯λoϲ … ⋯π⋯Өηκε inutilia', said Wecklein. O actually omits 11–12 τoȋϲ … χρ⋯νωi, and P. Mil. Vogl. 44, which is not available at the crucial point, evidently had a text different from (and shorter than) ours at 10–12.
28 See below, p. 42 (§10 (a), (c)).
29 Scholia in Euripidem ii (1891), pp. 1–2Google Scholar.
30 See below, p. 42.
31 See above, pp. 38–9.
32 Everything which I say about the Aldine in this section holds good for the edd. Hervag. (1537, 1544), the ed. Brubach.1 (c. 1558), and the edition of Canter (1571). The only difference which I have noticed between any of these editions and the Aldine is that ed. Brubach.1 writes Τρoiζ⋯νi in the margin as a correction for Θ⋯βαiϲ at 25, and Canter puts this in the text. – According to Turyn (p. 346) N copied the hypothesis of Hecuba from the Aldine; cf. Matthiessen (above, n. 9), 129.
33 See below, p. 43.
34 See above, p. 39 n. 18.
35 See above, pp. 38–9.
36 ‘Summi momenti est codex in Hippolyti scholiis quorum optimam recensionem praebet’, Schwartz, ii p. iiiGoogle Scholar. Cf. Turyn, p. 348, Barrett, pp. 78–81.
37 The lemmata are given twice: once in the continuous text of the scholia, and once in the margin. I mention this in view of the reports on 19 and 123.
I draw attention to the lemmata to the scholia on Andromache in Ottob. graec. 339 (Turyn, p. 355)Google Scholar. Again, they are for the most part ignored by Schwartz; but they are of great interest.
38 Except in the case of Σd, where I have checked the lemmata only at 19 and 447.
