Tragicorum Graecorum Fragmenta Selecta

This volume deserves a loud welcome and louder applause. It contains in under 200 pages almost all the major Tragic fragments, including satyr-drama, in bare scholarly essential: Greek texts with terse introductions, ancient secondary or orientatory matter (such as hypotheseis) and apparatus criticus. The technical precision with which the texts are presented, the clarity of the apparatus and the numerous and often convincing conjectural interventions are exactly as one would expect from the acclaimed editor of Euripides. ἄνδρα… στεφανωσάμενον αἰνέσω (James Diggle is a crowned heavyweight in this particular ring …). When has there been such a full selection before, let alone one of this quality?

In a way the volume is a descendant of A.S. Hunt’s almost forgotten Oxford Classical Text selection of 1912, Tragicorum Graecorum Fragmenta Papyracea Nuper Reperta; and it accompanies the similarly conceived OCT volumes of D.L. Page ( Lyrica Graeca Selecta, 1968) and M.L. West ( Delectus ex Iambis et Elegis Graecis, 1980). It appears amid the continuing surge of publications devoted to Tragic fragments, many pushed forward in response to the great enterprise of Tragicorum Graecorum Fragmenta itself (ed. Snell-Kannicht-Radt, Vols. 1-4, 1971-85). There are so many new texts, and so many new studies, that the two-part Euripides-volume which will complete TrGF (ed. Kannicht) is understandably delayed; exactly half of D[iggle]’s pages are given to Euripidean texts. Apart from TrGF itself, even the most noteworthy recent editions of fragmentary tragedians or plays make an impressive number:

for Aeschylus, H. Lloyd-Jones in the ‘Loeb’ (Vol.1, 1957, pp.526-603); H.J. Mette, Die Fragmente der Tragödien des Aischylos (1959); M. Werre-de Haas, Aeschylus’ ‘Dictyulci’. An attempt at reconstruction of a satyric drama (1961);

for Sophocles, R. Carden, The Papyrus Fragments of Sophocles (1974); D.A. Sutton, Inachus (1979); E.V. Maltese, Ichneutae (1982); H. Lloyd-Jones in the ‘Loeb’ (Vol.3, Fragments, 1996) – and Prof. A. Sommerstein of the University of Nottingham has a project to publish annotated editions of Sophoclean fragmentary plays;

for Euripides, after E.W. Handley and J. Rea’s Telephus (1957), the fragmentary plays as generally more extensive have received many individual commentated editions, some of them more than one ( Andromeda, Cresphontes, Erectheus, Hypsipyle, Phaethon, Telephus). There have been, and are in progress, collected or selective editions: H. Van Looy, Zes verloren Tragedies van Euripides (1964); C.F.L. Austin, Nova Fragmenta Euripidea in Papyris Reperta (1968); C. Collard, M.J. Cropp, K.H. Lee, Euripides: Selected Fragmentary Plays, I (1995); F. Jouan and H. Van Looy, Euripide. Fragments, 1ère. Partie (‘Budé’ ed., Vol.8, 1998 — a volume which on pp. vii-x, lv-lxxxiii contains a definitive survey and bibliography of Euripidean “fragmentary studies” since the early 19th Century); most recent, N. Pechstein, Euripides Satyrographos (1998). It is right to recall too the great impulse given to study of the fragmentary Euripidean plays by T.B.L. Webster, Euripides (1967) and R. Aélion, Euripide héritier d’ Eschyle (1983) and Quelques grandes mythes héroïques dans l’oeuvre d’ Euripide (1986);

for the Minores there have been the chapters in B. Snell, Szenen aus griechischen Dramen (1971), the edition of some fragments in G. Xanthakis-Karamanos, Studies in Fourth Century Tragedy (1980) and the generous selection of texts, with German translation and brief notes, and with an Introduction by R. Kannicht, in Musa Tragica, ed. B. Gauly u.a. (1991).

D. has therefore had to take into account a huge range of work published since earlier and usually author-based selections from Tragedy.

‘Studiosae iuventutis in usum potiora tragicorum fragmenta ad arbitrium selegi’ (D., Praefatio v); ‘elegi … longiora ac notabiliora’ (Hunt, TrGF PNR); cf. ‘carmina vel carminum fragmenta praestantiora selegi’ (Page, LGS) and ‘… utile miscuisse dulci’ (West, Delectus). D.’s ‘potiora’, Hunt’s ‘longiora ac notabiliora’, Page’s ‘praestantiora’, and West’s dictum: what are the appropriate criteria for selection in such a volume and in such an OCT “series” as this? Authenticity and probable rather than merely possible ascription? Length alone? Completeness of content or self-containedness, as in the run of passages found in ancient anthologies, which yield the majority of book-fragments? Individual quality? General interest? Particular importance to reconstruction of a play-plot? Adequate representation of a play-scene, such as a dialogue or a lyric part? Mere existence, even, as in the accidental survival of a papyrus fragment, of whatever length or content, and condition?

Length — or brevity, rather — must be a chief criterion, even for the plays with the fullest outline; and in a selection which does not have reconstruction as an objective, it is seldom useful to include very mutilated papyri or purely sententious book-fragments. Selection ‘ad arbitrium’ (D.) must therefore rule. I spent perhaps an hour, certainly no more, scanning Vols. 1-4 of TrGF (Minores, Adespota, Aeschylus, Sophocles), noting fragments suitable for inclusion. For most, my criteria were either visual (sufficiently numerous and complete lines with a reasonable continuum) or swiftly impressionistic (whether a passage was self-contained or nearly so, and whether its content was of significance, interest or merit). The nature of Euripidean fragments dictates a different principle of selection from that appropriate to the other poets and texts, and I discuss this below; and the lack of Vol.5 of TrGF makes selection from Euripides extremely time-consuming — and demanding of energy, for I calculate that to assemble all the very scattered Euripidean fragmentary texts would at present still need a largish box. D. without question took much more than one hour over his choice; but the similarities and discrepancies between my instant selection and his more deliberate one may be of interest.

In Aeschylus, we coincided in seventeen selections, including the obvious plums from (satyric) Dictyulci, F 46a and 47a; from (satyric) Theori, the fascinatingly metatheatrical F **78 (one or two asterisks against the numbers of Snell-Kannicht-Radt take a fragment further away from confident ascription); from Myrmidones, F **132c and 139; from Niobe, F 154a. D.’s total is twenty selections, mine perhaps twenty-five. He omits F 74 (Heraclidae), which I selected, and he was wiser, since this is a very minor piece of description. I omitted but he includes two heavily damaged papyri, F **168 (Xantriae), about thirty lines which contain matter relevant to “ritual” studies and are of some technical interest because of the overlap with testimonia, and F 281a (play unidentified), about fourty lines of a conversation between Dike and (?) Chorus in which the goddess describes her functions. He also includes F 350 (play unidentified), a book-fragment of nine lines in which Thetis complains that Apollo prophesied falsely her son Achilles’ happy destiny; certainly this is of interest to readers of the Iliad. I included rather more fragments, even short ones, attributed to Prometheus Lyomenus (D. gives F 190-3, *195, 196, 199).

In Sophocles, we coincided in eleven selections, mostly from the big papyrus-fragments, including the obvious messenger-report from the possibly satyric and uncertainly authentic Eurypylus, F **210; parts of the probably satyric Inachus, F **269a and **269c; all that survives of the satyric Ichneutae, F 314; from Rhizotomi, F 534 and 535 (book-fragments), Medea gathering poisonous ingredients; the book-fragment F 941, the evocation of sexuality’s tyranny (cf. Euripides F 897 and 898, from unknown plays(s), selected by D. on his p.167); and five book-fragments from Tereus, F **581, 583, 591-3. D.’s total is twenty-four, mine was eighteen, because I was too quick in passing over things which have either inherent interest or some value at least to the picture of Sophocles’ survival: F **10c (Aiax Locrius), ten broken papyrus-lines in which Athena upbraids Ajax for raping Cassandra (D. compares Eur. Tro.69-70); F 255 (Thyestes), a book-fragment of eight lines describing the miraculous growth and ripening of grapes; F 524 (Polyxena), a book-fragment of seven lines in which (?) Agamemnon doubts her sacrifice (cf. in general Eur. Hec.120-2); F 555 and 557 (Scyrii), two short book-fragments, the first overlapped by a papyrus, evoking the miseries of sea-trading and the finality of death; F 659 (Tyro), a book-fragment of ten lines perhaps represented in a papyrus, a striking simile for a girl whose hair is forcibly shorn; and two more fragments incertae sedis, F 871 and **1130.

From the Minores, D. includes only pieces from Critias, Pirithous ( TrGF Vol. 1, 43 F 1, 4a, 5, 7) and Sisyphus (F 19), and from Neophron’s Medea (15 F 1-3). The dramatic and compositional interest of these is incontestable, as is the still disputed ascription to Critias.¹ I made a rather wider selection, wanting also Agathon 39 F 4, together with Theodectes 72 F 6, and the possible antecedent to both these descriptions (in unidentified plays) of the Greek letters making the name of Theseus, Euripides F 382 (Theseus); Carcinus 70 F 5 (play unknown), a book-fragment of ten lines, in which Carcinus, an immigrant to Syracuse, aetiologizes the cult of Sicilian Demeter; Chaeremon 71 F 14 (Oeneus), the lush description of exhausted moon-lit dancers; Diogenes Sinopensis 88 F 7, a book-fragment of 12 lines describing asceticism; Moschus 97 F 6, the well-known book-fragment of 33 lines evoking man’s rise to civilisation, useful for students to set beside e.g. the Prometheus 442-506, Soph. Ant.332-75, Eur. Supp.201-18 — and Critias, Sisyphus F 19, selected by D. himself on pp.177-9; Sositheus 99 F 2, a book-fragment of twenty-one lines describing Lityerses the murderous host, of comparative mythological interest. This last piece is textually very corrupt (although this criterion cannot bar papyri from inclusion), but the others could claim a place on grounds of content or style, to illustrate both the continuing and the changing condition of Tragedy. Like D., I did not choose the sadly mutilated but extraordinarily interesting fragments of (?) Astydamas the Second’s apparently famous Hector (60 F **1.h and i, **2a), or two things to be reasonably disqualified on grounds of genre, Python’s Agen (91 F 1) and the singular Exagoge of Ezechiel (128), despite the latter’s continuum of 269 lines; to have included Exagoge would perhaps have invited desire for at least a sample from the Christus Patiens. The Hector has recently been edited and studied by Xanthakis-Karamanos, Studies, pp. 162-9 (above), Agen is available in Snell, Szenen, pp. 104-37 (above), and Exagoge has received a full annotated edition from H. Jacobson (1983).

From the Adespota D. selects only the best-preserved part of the much-discussed (?) Gyges papyrus ( TrGF Vol.2, F 664.18-33). I chose also three fragments of comparative interest to other plays: F *8.l, a book-fragment of eleven lines perhaps from a Rhesus, in which Hera urges upon Athena the need to punish Paris for his insult in the Judgement (cf. the surviving Rhesus 595ff.); F 129, a book-fragment of nine lines on the power of gold (cf. Soph. F 88, from Aleadae, included by D. on p.35, and Euripides, Danae F 324, not included); and F 665, from a Seven against Thebes, twenty or so papyrus lines in which Polynices, Jocasta and Eteocles fail to achieve reconciliation (cf. the agon of Eur. Pho. 355-637).

For Euripides, the considerations are quite different. First, there is the very large number of fragments from plays whose outline can be reconstructed with more confidence, usually on the basis of papyrus-pieces; here, judicious selection or, much more often, omission among the book-fragments has to be made, and D. chooses only a very few which are of strong individual interest (see e.g. Bellerophon and Stheneboea). Furthermore, such book-fragments may conveniently be read, with useful discussion, in editions of the individual plays (see the fifth paragraph above). In Antiope it is noteworthy (and in my view correct) that D. accepts Luppe’s attribution of P.Oxy.3317 to this play rather than to Antigone (where Kannicht, in my present information, proposes to keep it still, as F *175; so too Jouan and Van Looy in the Budé Euripides [above]).² In this category of play I miss something at least from Andromeda, even the scraps of the remarkable anapaestic opening monody, F 114-7. Oedipus too is unrepresented, despite the overlapping of two book-fragments by the very damaged P.Oxy. 2459, one of them the striking description of changing hues reflected from the Sphinx’s back (formerly Adespota F 541 Nauck ²); and I would have included under Oedipus the purely trochaic F 909 Nauck ², which seems to cohere with the tetrameters attested for the play as F 545 and which is to be admitted by Kannicht as F 545a (cf. Austin, NFEPR on his Fr. 88 = F 545).

Second, there is the problem of selecting from the many Euripidean book-fragments which are of little or no significance to reconstruction and whose style or sententious content are so frequently matched in passages in the surviving plays, where the known dramatic context gives them vitality. An overriding independent interest must be the criterion. Everyone will have expected to find F 282 selected, the condemnation of athletics (Autolycus ‘A’) and it is (p.96), as also (p.166) is the celebration of Earth and Ether, much pondered in antiquity, F 839 (Chrysippus: how one would like to know what bearing and location it had in the play!). Among the rootless fragments I marked out for selection just F 897, 898, 910 and 912, exactly the four which D. chooses.

Now to the meat of it: what do we get to aid our reading? D.’s Praefatio is very brief, describing the background to his work but, most importantly, listing his separate papers upon the play-texts selected, all of them published subsequently to his Euripidea (1994); these bare references, like the statement ‘Papyros … inspexi’, hardly suggest the very thorough and often important fresh scrutiny to which he has subjected not just the primary evidence, but also the legions of editorial conjectures upon these fragments, many of them in fairly inaccessible places. D.’s editorial changes in his main text, and many suggestions confined to his apparatus, can only be fully understood upon studying these papers — and often it is essential to have the volumes of Snell-Kannicht-Radt open as well. I single out the papers on Ichneutae in ZPE 112 (1996) and Antiope in PCPS 42 (1996) for his productive re-examination of papyri, and AntClass 65 (1996) for his work upon Phaethon since his already and unprecedentedly full edition of 1970. For many plays where it is possible to get some hold upon the plot as a whole, or upon the action of at least an individual scene, D. gives us a terse headline summary of content; he cites ancient hypotheseis, of whatever pedigree, where they are useful (Sophocles, Niobe; Euripides, Alexander, Melanippe Sophe, Stheneboea, Hypsipyle, Phaethon, Phrixus ‘A’ and ‘B’ or for Euripides cites Hyginus (Antiope, Phrixus ‘A’). Thereafter, we are largely on our own, as in almost any OCT. D. leaves questions of authenticity largely to inference from the single or double asterisks attached to fragment-numbers (see above). Comparative material, explanations and brief comments upon his own or others’ conjectures are infrequent, much rarer than in the Euripides edition: Aesch. F **132.3 (p.18), 281a.29 (30), 350.1 (32); Soph. F **210.52 (38), 269a.34 (42), 314.183 (53), 535.4 (69), 555.3 (70); Eur. F 360.42 (103), 453.9 (114), Sthen. hypothesis 3 (128); and Critias F 19.18-19 (177) exemplify the range; and these few give appetite for more. The extraordinary number and high quality of his own fresh conjectures upon the text, including very many supplements exempli gratia, and their general significance to the plays, will become clear from the following detailed but far from exhaustive notes (a major reason for my slowness with this review is that so many fragments required very close reading and challenged me to try my own hand at conjecture or supplement — perhaps predictably with the very few results which I care to acknowledge here!). It is understandable that D. offers more conjectures and supplements in the newer papyri than in the older ones and the long-known book-fragments. As in his Euripides he nevertheless occasionally leaves obeli in the text and cites hardly a conjecture, or advances none of his own; in these places his restraint is invariably well reasoned and I too found myself almost always helpless to make a suggestion.

In the following notes the line-numbers within fragments are those of D.’s volume and not necessarily the same as those of Snell-Kannicht-Radt, Nauck or editiones principes of papyri. I refer in abbreviated form to D.’s papers listed in his Praefatio (see above).

Aeschylus: the very first fragment in the volume makes a methodological test-case for selection: in Glaucus Pontius F 25e only the conjectural supplements of 5-6 make it seem reasonable to include the extremely mutilated 1-4; for D.’s supplement in 12 see MusCrit (1995-6). At Danaides F 44.6-7 D. rightly prefers Hermann to Diels (favoured by Luppe, GGA 239 [1987], 27). Dictyulci : in F 46a.10 D.’s γ’ for τ’ emphasises the diune deity Poseidon-Zeus in conformity with the testimonia; F 47a.25 and 48 πάπ[π]ας, 37,46 and 54 Doric forms restored. 38 φίντων printed, a word only recently admitted to LSJ, but of unknown meaning; D. might have added ‘non intellegitur’, I think, or cited Lobel’s Φίντων. 68 χάρις?ιν λαμπραῖς τῆς Ἀ[φ]ροδίτης brilliantly D., MusCrit (above). Edoni (p.10): the heading ‘prima λυκουργείας fabula’ leaves rather a lot for readers to infer about the content. In contrast for Theori or Isthmiastae on p.11 D. offers an extremely helpful summary of presumable content; F **78.1 no fewer than five conjectural supplements are recorded; 11-12 καλλίγραπτον and 36 χρήματα are the sort of places where glossing might help; 25 D.’s προσεν[πεδοῖ is very attractive ( MusCrit). Cares or Europa : F **99.16 αἰχμῆς εἰ ‘ξ by D. improves Francken. Myrmidones : F **132c: supplements ex. gr. in 3, 7, 10; an attractive new reading in 16, ] ἀράσσων for ] ταράσσων; 18 ἀπανθίζων suggested in ἐπανθίμων?. Niobe F 154a: it says much for the energy and acuity of many earlier scholars that D. has felt able to suggest nothing new in this papyrus. Xantriae F **168.21: if λευκο [ at the start can connote “brilliant, happy”, D.’s ὀ]μ ? μασιν looks probable. Prometheus Lyomenus F 192.4 λίμην (for λίμναν, rightly, cf. F **273a.2 on p.28) μελάνων τρόφον Αἰθιόπων attractively for παντοτρόφον αιθ.. (cf. Eur. Phaethon 4, p.151). (fab. incert.) F 281a.29 καρπουμένη sc. πόλις (28) D., citing KG I.80 (cf. MusCrit), is far better than καρπουμένη[ς ( πόλεως) or – η[ι ( πόλει); other attractive suggestions in 32, 34 and 36.

Sophocles (where D. leaves in his text much more that is corrupt or uncertain than does Lloyd-Jones in the differently intentioned ‘Loeb’): Aiax Locrius F **10c: D.’s methodology perhaps does not allow him to mention that this fragment could be more confidently ascribed to the play if Haslam is right in adding 9-10 to the securely attributed F 15a (not selected); and it is a pity that the seventy-one scraps of this papyrus are so insubstantial that it cannot be known whether they belong to one play, and to this one. In line 2 D. is certainly right to prefer καινὸς to κλεῖνος or κεῖνος as a possible reading of the (corrected) papyrus, given that Athena’s question in 1 starts with ποίου δρύαντος…; in 9 κρηπῖδος (D.) ἐξέστρεψεν (Ll.-J.) makes a convincing combination, printed by both. Achillis Erastae F 149.6: D. obelizes almost all of line 6 including χυμὸς, and he cites only Meineke’s κρυμὸς; both seem better than Meineke’s χῦμα, printed by Radt and Ll.-J. χυμός can perhaps be retained, either literal as ‘melt-water’ (cf. χῦμα νιφάδος Alciphr. 1.23) or metaphorical as ‘taste, flavour’, ambiguous both of snow and love in the sequence of ideas — but the latter seems to require a Passive, possibly ἑθῇ ἀφῇ. Eurypylus F **210: despite the double asterisk, D. could perhaps have ventured more upon the problematic ascription and indeed title, beyond the citation of Plut. Mor.458d against lines 8-9. In 9 D. only hazards, but Ll.-J. has printed, his conjecture ἠραξάτην for Plut.’s ἐρρηξάτην (which, if LSJ are right, is a unique use of the intransitive Active); 47 τοῖν [ν]εκροῖν for τον [ν]εκρον is very good, and certainly a Dative is needed; the two dead are named in D.’s note on 50-3; 66 neither λυγρὸ]ν (Wilamowitz) nor πικρὸ]ν (Hunt) seems quite right here as adjectival with ( ἐρρόθει) στόμα; perhaps ἁδινὸ]ν ? Inachus F **269a: excellent supplements ex. gr. in I.22, 30-1, 32 (a useful note in the apparatus), 32-3 by D., all but the first printed by Ll.-J.; in II.18 D. suggests reading εὖ rather as Epic ἐύ. Ichneutae F 314: D.’s fresh scrutiny of POxy 2369 in ZPE 112 (1996) variously supports or questions earlier readings or conjectures; see particularly 17, 62, 110, 120, 122, 139, 145, 147, 217, 242, 283, 367, 376, 378-95. I note here his attractive supplements in 39, 41, 42, 45, 47 (where τελῶν either as Present or Future properly accommodates τάδε); in 50 he corrects papyrus ἂν to ἢν, cf. 74 and 171; because he prints the whole papyrus, D. includes the very mutilated 64-75; Ll.-J. does not, but might have done, since the excited staging of a search about to begin is readily intelligible (and was vividly evoked in the 1997/8 British performance by Chloe Productions); on 113 D. remarks of the line-end ‘vix [ βοῶν (Hunt), nam ῥοίβδημα bubulcorum est’, so that I am tempted to suggest ῥοίβδημ’ ἐάν τι τῶν [ἐλατῶ]ν (unless this is a letter too few) οὖς [βάλῃ, with Porson’s Law defied in satyric; for the noun ἐλάτης cf. Eur. Phaethon 72, p.153; in 114 ποι[μένο]ς suits both space and sense ( ποιμήν again in 160); in 133 D.’s τ[ίς ἡ τάραξ]ις is printed by Ll.-J., and seems superior to D.’s apparently preferred alternative τ[ί κρᾶτα σείε]ις; for his acceptance of Walker’s ὀνθ[ία in 147 and ἐνήλατ’ ἄξυλ’ ἀρτίγομφα in 316, and Bücherer’s ἀθύρων in 326, one really needs to read ZPE. D. plausibly suggests printing questions in 178 and 224; his excellent suggested supplement in 217 is printed by Ll.-J.; in 226 his correction of φέρων to φορῶν is certain, like his 3. Plur. ἔτεινο ? [ν in 234; in 267 his κρυφ[αῖος may be better than Radt’s κρυφ[αίως and is certainly better than Hunt’s κρυφ[αίαν; in 272 his βρέφος is preferable to Wilamowitz’ τοῦτον; in 333, if papyrus πορεύω is right, the intransitive Active is unique, so that D.’s πολεύω, versor, commends itself, preferably with his ὧπερ (‘causal’, ZPE) rather than Hunt’s ὅπερ. Finally, I have to agree with D. that 369-70 ‘non intelleguntur’, pace Ll.-J. in the ‘Loeb’. Niobe: p.64 hypothesis: good supplements ex. gr. in fr.1.12, 13 and fr.2.23, 26; in F **442 D. makes a tiny improvement upon Snell’s punctuation. Rhizotomi F 534.6: D. daggers the asyndetic participle βοῶσ’, accepted by Ll.-J.; will βοαίς work here? For ἀλαλαζομένη he writes ὀλολυζομένη, citing his important discussion of this verb at Euripidea 479. Tereus F 583.2: ἔβλεψα ( ταύτηι) may perhaps be defended (with Ll.-J.), although LSJ cite only the Accusative of a direct Object with this verb, as ‘behold’ rather than in the sense ‘view with thought’ required here; cf. Eur. Hipp. 379 τῆιδ’ ἀθρητέον τόδε. Tyro F 659: D. obelizes 6-7 and cites a few conjectures, but this is one of the places where one wishes he had tried one of his own. So too in the obelized F 941.7-8 (fab. incert.); Radt’s apparatus shows how scholars have wriggled to save the text, but I too see no rescue; D. transfers 12 to follow 8, for then 13 “gods” follow better upon 9-11 “animals” — but where is “man”? Radt and Ll.-J. mark a lacuna.

Euripides (I must again acknowledge my own and Martin Cropp’s gratitude to James Diggle for allowing us to print many of his readings and conjectures in SFP I in advance of his separate papers and now this volume; I do not discuss them here): Alexander I.2 (F 46): D.’s βρ[έφος διώλεσα is as good as Snell’s βρ[έφος κατέκταμεν, but perhaps neither is as good as Lee’s βρ[έφος κτανεῖν ἔτλην (cited by D.), where ἔτλην will be well picked up by 3 τλήμων γε Πρίαμος; and D. records my own suggestion in 4 ὡς ἴσμεν οἱ παθόντες ο[ἱ τλάντες θ’ ἅμα. At I.12 D.’s στείχουσα]ν better matches the space (and idiom: see his CQ [1997]) than e.g. Wilamowitz’ ἥκουσα]ν. III.1 (F 62a): D.’s apparatus ignores his report in CQ of Lee, who reads possibly ]δ before εκτ in the papyrus; so perhaps τόν]δ’ Ἕκτορα; 4 Δηίφοβον and 8 μεγάλα νο]μίζει D., both very probable; IV.8 (F 62d) ἵκοιτο δ]εῦρό < γ’ > εἰς κτλ., similarly. Antiope : the seventeen appearances of D.’s name in the apparatus do not fully represent his achievement in these fragments: one needs his separate papers to hand. I single out F 187.4, where D.’s apparatus modestly conceals that he had reached, on the basis of analysis of Euripides’ idiom ( CQ [1997]), the unquestionable correction οἴκοι κἀν πόλει before he found he had been anticipated by a “bald” conjecture of R.J. Walker (a private scholar and a most intriguing figure, as the Prefaces to his various books reveal; for D.’s acceptance of his ideas, see also on Soph. Ichneutae 147 and 316 above). Then, F 175 ( POxy 3317: for the attribution, see above), where there are particularly attractive supplements in 2, 6 and 10: see D.’s APf (1996). Last, F 223 (PPetr 1.1-2), where D.’s work is again of exceptional quality (note the ‘Budé’ editor: ‘nous reproduisons pour l’essentiel ses leçons’): for his collation of the papyrus see, together with PCPS (1996), especially 4, 15, 19, 29, 32, 35, 37, 44, 54, 57, 65, 67-8, 71-2, 112 (with revised readings at 31, 43, 58a [discussed at length in PCPS ], 62, 66, 87 and 95) and for his supplements e.g. 10, 37-41 (a very convincing sequence), 46, 48, and especially 54 and 89. I note that for 21 an apparent conjecture τίνες δὲ < καὶ π > ῶς δρῶντες is recorded in Allen-Italie, Concordance, p.170 (unattributed): it fits the papyrus-space, but τίνες δὲ καὶ < τί > δρῶντες would be more idiomatic; D. in fact offers neither. At 22-3 D.’s desideration in PCPS of the sense would be met by something like σημήνατ’, εἴπατ’ ἐκφυγεῖν δ’ αὐτὴν ἐ[ᾶν (or μ[όρον). Archelaus F 228.2: D. is right to cite only West’s ἓν γαίας from the multitude of conjectures for ἐκ γαίας and probably right to follow Tiberius in deleting lines 3-5. Autolycus ‘A’ F 282.12: D. illustrates the “metaphor for simile” τρίβωνες in CQ (1997); he has no new conjecture for the irritating corruption in 23. Bellerophon F 285.12: D. is almost certainly right with ( ὅστις) … σπανίζηι for σπανίζει, but not, I think, with εὐτυχῶν for εὐτυχεῖ; in F 292.6 he obelizes ἀλλὰ τῶι νόμωι : for the reason, see my note on 4ff. in SFP I. Erechtheus F 360.6: for defence and interpretation of the transmitted text see D. in CQ (1997); 38 < δὲ > adds to the monosyllables conjectured, but D. puts none of them in his text. F 362.20 πονηροὺς : D. cites only Matthiae’s λαλοῦντας for a corruption as frustrating as Autolycus F 282.23 (above). In F 370, the Sorbonne papyrus, so productive of suggestions for the whole play’s structure and contemporary import, D. is surely right not to print Χάρων in 2, but keeps it in the apparatus, perhaps inevitably; it is scarcely credible in the context, like the mention of Ares read by some in 14. The end of 14 has so far defeated conjecture (D. cites none); the likely question before the answer in 15, ‘(Eumolpus) has fallen’, is ‘But who (on the other hand: for ἀλλὰ γάρ in questions see Denniston, GP 108) has died?’ or ‘But who still lives?’; yet e.g. ἀλλὰ τίς γὰρ ἔστ’ [ἔτι; means defying the apparently clear letters ιθ before the loss. In 8 a possible revised reading of the papyrus is canvassed. At 21-2 D. is more cautious than Cropp in SFPI in restoring F 361 Nauck to the mutilated papyrus (Kamerbeek). He has new conjectures in 17, 35 and 41, all argued in CQ (1997); here, he has a promising new idea for 37, ἀφῶ δάκρυα‘dummodo προσεῖδον (38) non recte legatur’, suggestions for 38 and 43 and attractive supplements in 62, 81, 82. In 49 he prints his ἔνοσι]ν (PapFlor 7 [1980], 59; not in Euripidea) but in 51 keeps his equally possible σεισμ]ῶν to the apparatus. Perhaps the biggest problem in the whole papyrus is the proposed disentangling from αικορυφηναπαταιθ[ in 42 of some form of ἀπάτη (Austin first, whom some have followed); the whole character of the play must be reinterpreted if either Erechtheus or Praxithea has “deceived” either one or all three of their daughters into sacrificial death; D. wisely keeps the suggestion to the apparatus. Cresphontes: PMich inv. 6973 is still officially without its editio princeps, and D. has been dependent largely upon the text published by Cropp in SFP I. In this text (F 448a), and its overlapping POxy 2458, it is particularly important to note D.’s different line-numbering for this volume. Even such an expert as D. has been unable to reconcile the colometry of POxy with the doubtfully colometric line-divisions of PMich, but he makes small suggestions in 47 and 49. In 16 D. judges Mette’s σμ[ικρ]ὸς too short for the space, unless σμ[εικρ]ὸς was written; but if not σμικρὸς, what? An adjective or noun is required with the participle ὢν; only σκοτιός seems possible, but is also perhaps too “short”; on the other hand, if Cresphontes was presented here by Euripides as Merope’s love-child, it would help to explain how his concealment aided his escape from murder. At the end of this same line, supplementation is made difficult by the elision in ετ [ ]'[ (D. notes similar elisions marked after scriptio plena in 13 and 21: thus ἔτ[ι?] here seems inescapable); so perhaps ἔτ’ [εἴ]χ[ετο (“held” at the breast; for ἐπί cf. Hdt. 3.133 ἐπὶ τοῦ μαστοῦ ἔφυ φῦμα) rather than e.g. ἔτ’ [ἐτ]ρ[ύφα (metrical ‘position’? Cf. Ion 1376 ἐν ἀγκάλαις μητρὸς τρυφῆσαι, where also there is no particular point to ‘luxuriating’); the seeming need for a finite verb rules out an adverb like [A)ς]φ[αλῶς. F 453.9 D.’s [ ἵθ’] ἴθι is supported by reference to Euripidea p.388 n.86, from a typically exhaustive analysis of Euripides’ practice with elision in doubled verbs; it persuades me against Cropp’s defence in SFP I of Bergk’s remedy. Cretes : F 472b.3 τ[αυρωπὸν φορεῖ : the first supplement is very possible, the second certain. In 10 I would now accept Luppe’s supplement as palmary (and D. prints it in preference to his own earlier convincing one in CQ [1997]); certainly a Present verb is needed. F 472e: D. might have recorded that PBerol is now lost, for dependence upon a facsimile both facilitates and endangers conjecture; in 44 his doubt of μ[ after ἔστομωται, identifying rather χ[, is important for future restorers; he does himself not enter the minefield of 45, for example; see also on 51 below. In 22 I have been wondering whether the Future οἴσε[ται may be possible after the Aorist of 21 and before that of 23; if so, then something like οἴσε[ται μέρος πόν]ων or even βλαβὴν δόμ]ων (too long? but cf. 41 τῆς σῆς ἕκατι ζημ[ία]ς)? In 23 D.’s ὅνπερ is a marked improvement. In 27 D. refers to KG I.656-7 for the important emphasis in the express 2. Person pronoun doubted by Wilamowitz. He makes a small orthographic correction in 48, and in 51 generously reports my slightly free conjecture of [Qερ]μὸς. In Melanippe Sophe F 481.12 I am inclined to agree with Cropp in SFP I that ὄνομά τε τοὐμὸν, sc. ἀνοιστέον from 11, is defensible, but D. obelizes and does not mention Wilamowitz’ ἐπ’ ὄνομα τοὐμὸν. Melanippe Desmotis F 494: D. has no conjecture for the baffling ὅ τ’ ἄγαν ἡγούμενοι in 24 but is almost certainly right with 25 ἢν μί’ εὑρεθῆι for εἰ μί’ ηὑρέθη. F 495.17, 19, 23, 24 good supplements ex. gr.; in 30 his ἄλλον for καλὸν is palmary and in 31 his more exact reading of the papyrus leads him to the convincing supplement παῖς αὑ]τὸς. Stheneboea : for the hypothesis see D. in ZPE 77 (1989) = Euripidea 334; in line 13 he prefers Wilamowitz’ παρ’ αὐτῆς to Rabe’s παρὰ του, with a note on the latter ‘obstat hiatus’, but the former does not solve the problem when and how Stheneboea might have warned, or wanted to warn, Bellerophon. F 661.22-5 are deleted, 22-3 being long suspect and 24-5 doubted by Holford-Strevens; certainly the bigger deletion solves many problems. D. leaves 11 λέχος and the whole of 26 obelized, without citing or making conjectures. Telephus F 696.14: possibly Ἕλλην δὲ βαρβάροισιν ἦρχε γειτονῶν for ηρχετεκτονων and then e.g. πολλῶι σὺν ὄλβωι (or ὠλβισμένοισι) πρὶν μολὼν Ἀχαιικὸς in the largely corrupt 15? D. records convincing readings by Maehler at the end of l6. Hypsipyle : most of D.’s new readings in POxy 852 and his conjectures are set out in Eikasmos (1995), but note CQ (1997) on his convincing 14 κ[α]ὶ ο[ὐχὶ] λυ[π]ηροὶ. I draw attention to the inventive δέρκου μὰν for δεῦρο ταν in 46, to the various supplements proposed in 172-5, of which only κακῶν at the end of 172 is convincing, to the revised colometry in 244-5 and to the conjectures by Willink printed in 37 and 40 and canvassed in 284 and 288. Phaethon : D. has issued extremely detailed Epilegomena to his 1970 edition in AC (1996); in this volume the more important new things from there are: 47 a further conjecture, [ ὅσων ἐρᾶις ]; 67 Burges’ δ’ ἐν for δὲ; 79 change of mind in favour of the palimpsest’s ἄφαντος, 82 of Wilamowitz’ ἀχοῦσιν, 170 of Faber’s κάτω διήσει (defended by West at BICS 30 [1983], 77), 240 of Hermann’s βασιλέως and 270-1 of Kannicht’s question mark after καταστάσω (although D. retains the palimpsest’s ἄφαντος against Heiland’s ἄφαντον); at 238 D. prints a good conjecture by Willink and reports a less attractive one in 282 (D. himself suggests various reorderings of the words in 282-3). Phrixus ‘A’ F 822b.1 and 16: attractive supplements suggested. In incert. fab. F 912.5 D.’s προχυταῖον is surely right.

Minores, Adespota: Critias, Pirithous : very suggestive supplements ex. gr. in F 4a.10, 11, 14, 15, 16, 18, 21, 22, 25 and F 5.17, 18, 19, all line-ends; at F 7.11-12 an excellent new idea, ἀλλ’ ὁ (or π[ροστατῶν]) ἐμὴν ἔχεις . In the long Sisyphus book-fragment, F 19, D. has argued in Prometheus 22 (1996) his conjectures 13 δέη for δέος (not, I think, necessary); 25 κύδιστον ironice for ἥδιστον, well possible; and 38-9 τῶν λόγων… | … οὕνεκ’ for τῶι λόγωι… |… οὐκ, very likely right. In the Gyges fragment, F 664, D. makes an attractive suggestion that 7 σι [..] might be read not as σῖ[γα] but as σι[γῆ], presumably either σι[γῆ]ς or σι[γῆ]ι; perhaps then something like καθεῖρξα σι[γῆ]ι πᾶ[σαν] αἰσχύν[ης] βοήν.

The accuracy of the volume is remarkable. I noticed only the (apparent) extrusion into the margin of text at A. F **168.21 (p.23) and Argum (sic) in the apparatus at S. F **269c.25sqq. (p.43).

Notes

1. D. p.172 accepts Wilamowitz’ ascription to Critias rather than to Euripides. While I squeezed the methodological argument for Euripides perhaps too hard in ‘The Pirithous Fragments’ (in J.A. López-Férez, ed. De Homero a Libanio, Madrid 1995, at pp. 183-93), I still have very strong doubts about both Critias and particularly the supposed tetralogy.

2. Antiope is proposed by W. Luppe, most recently at ZPE 102 (1994), 42 n.10, supported implicitly by D. at APf 42 (1996), 164 and O. Taplin, AK 41 (1998), 37.