¹ Versions of all or parts of this paper have been read to the Liverpool Latin Seminar (in October 1977) and the Classical Society of the University College of Wales at Aberystwyth (March 1979) and I am grateful to those present on those occasions for helpful discussion. I am also deeply indebted to Ian DuQuesnay, Peter Wiseman and Francis Cairns for detailed comments at various stages and to Professor R. M. Ogilvie for his indulgence in permitting an already by no means short manuscript to grow even longer. All culpa remains mine.

² For the unreliability of the MS headings cf. the table in Mynors, R. A. B., C. Valerii Catulli Carmina (Oxford, 1958), pp. xiv–xvGoogle Scholar. I quote from Mynors' text throughout this paper.

³ cf. Wiseman, T. P., Cinna the Poet (Leicester, 1974), p. 88Google Scholar.

⁴ Macleod, C. W. (CQ n.s. 24 (1974), 82 n. 3)CrossRefGoogle Scholar advances the criticism that Mani would have a solemnity amounting to sarcasm in the context. But that is to assume that sarcasm is not in place. Perhaps that is correct, but to use the point as an argument here is to undermine the objectivity of the names as evidence since some interpretation of 68 a is presupposed. In fact (see text) this problem seems unavoidable.

⁵ Mi + voc. is found in Cat. 10. 25, 13. 1, 28. 3. (The elision can be paralleled in 31.5 (vix mi ipse credens…), 68. 41 (…qua me Allius in re…), 76. 26 (o di reddite mi hoc pro pietate mea), 99. 13 (ut mi ex ambroisa…).

⁶ CQ n.s. 28 (1978), esp. pp. 162 ffGoogle Scholar.

⁷ Wiseman op. cit. (n. 3), p. 89.

⁸ cf. Yardley, J. C., Phoenix 32 (1978), 337–9CrossRefGoogle Scholar.

⁹ cf. Anacreon 96d where ⋯λλ⋯ ὅστις Mουσ⋯ων τε κα⋯ ⋯γλα⋯ δ⋯ρ' 'Aφροδ⋯της / συμμ⋯σγων ⋯ρατ⋯ς μνήσκ∈ται ∈ὐφροσ⋯νης clearly refers to love-poetry.

¹⁰ e.g. Williams, G., Tradition and Originality in Roman Poetry (Oxford, 1968), p. 231Google Scholar.

¹¹ Wiseman, op. cit. (n. 3), p. 95. Plut. Pomp. 2, Domit. Mars. fr. 1 (Morel), Anth. Pal. 12. 104, Plaut. Stich. 727 f. (which anyway does not concern people de meliore nota), Prop. 1. 5 (esp. 29–30) do not, I think, require interpretation in terms of the supposed custom. Only Ascon. ap. Donat. Vit. Verg. 24 explicitly records a proposal by someone (Varius) that his friend (Vergil) should share his amica (Plotia Hieria).

¹² e.g. Cic. Ad Herenn. 4. 57, Caes. BG 4. 29. 3, Hor. AP 20, Ter. Andr. 1. 3. 17.

¹³ vivere = ‘pass one's life, reside, dwell’, Lewis Short s.v. iib; vivere = ‘be alive’, ibid. ia, b, c.

¹⁴ On this passage see Wiseman, op. cit. (n. 3), pp. 96 ff.

¹⁵ e.g. Cic. de amicit. 74; ad fam. 2. 13. 2; 3. 10. 9; 5. 13. 5, 15. 2.

¹⁶ In this context it is a great pity that line 48 has gone missing.

¹⁷ e.g. Cat. 61. 9–10; Ovid, , Met. 10. 1–2, 21. 164, 170Google Scholar.

¹⁸ e.g. Cat. 61. 159 f.; Tib. 1. 3. 19; Ov. Amor. 1. 12. 3; Met. 10. 452; Varro ap. Serv. Eclog. 8. 29; Plut. Quaest. Roman. 29; Lucret. 2. 359.

¹⁹ On this basis lines 41–148 can be divided into a symmetrical structure thus: 41–50 (10 lines); 51–6 (6 lines); 57–72 (16 lines); 73–86 (14 lines); 87–90 (4 lines); 91–100 (10 lines); 101–4 (4 lines); 105–18 (14 lines); 119–34 (16 lines); 135–40 (6 lines); 141–8 (10 lines, assuming lacuna); but this is a purely formal fact (the sections of equivalent length in the two halves of the ring do not have any special substantial reference to one another). One may make of it what one will; all that matters here is that 91–100 comes in the centre of the main body of the poem.

²⁰ cf. Wohlberg, , CP 50 (1955), 44–5Google Scholar; Copley, ibid. 52 (1957), 31–2; Vretska, , WS 79 (1966), 323–5Google Scholar; Wiseman, op. cit. (n. 3), pp. 70 ff.

²¹ I accept argumenti causa that Catullus was responsible for the arrangement of the poems.

²² Further comment on this passage below, p. 136.

²³ cf. Forsyth, P. Y., CJ 66 (1970/1971), 66 fGoogle Scholar.; Sandy, G. N., AJP 92 (1971), 185 fGoogle Scholar.

²⁴ ‘Farfetched’ (Witke, C., Ennarralio Catulliana (Leiden, 1968), 48)Google Scholar; ‘irrelevant in its details’ (Fordyce, C.J., Catullus: a commentary (Oxford, 1961), ad loc)Google Scholar; ‘a style bordering on the grotesque’ (Copley, F., CP 52 (1957), 31Google Scholar); ‘almost impossible to take seriously’ (Quinn, K., Catullus: an interpretation (London, 1972), p. 188Google Scholar); ‘an outrageously long and often unclear comparison’, ‘…the ludicrous tastelessness of comparing Laudamia's love to oozy soil (sic!) or possibly a drainage channel’ (Elder, J. P., HSCP 60 (1951), 103Google Scholar).

²⁵ De Catullo Callimachi Imitatore (Leipzig, 1873), 39–40Google Scholar. Baehrens, A. (N. Jhb. Phil. 115 (1877), 412–13Google Scholar; Catulli Veronensis Liber Vol. ii (Commentarius) (Leipzig, 1885), on LXVIII b 33–4Google Scholar), followed by Lafaye, G. (Catullus et ses modèles (Paris, 1894), 217Google Scholar) speculated about Callimachean sources for Laodamia and Protesilaus, but there is, of course, no reason why the barathrum should come from the same source as that story.

²⁶ op. cit. (n. 3), p. 72 n. 39.

²⁷ On phrases of this sort see e.g. Norden, E., P. Vergilius Maro³ (Leipzig/Berlin 1926), 123ffGoogle Scholar.

²⁸ LSJ⁹ s.v ⋯κο⋯ω iii, 3; cf. Fordyce, op. cit. (n. 24) ad loc., Kroll, , Catullus (Leipzig/Berlin 1923), ad locGoogle Scholar.

²⁹ Hymn 6. 98. (The sense of the words in the two passages is, of course, different.)

³⁰ Hes. Theog. 317; Scut. 165; 416; 433; 465; fr. 25. 23; 26. 33; 33a. 32m/w; Ps.-Theoc. 25. 71; 113; 152. It was later picked up by Vergil (Aen. 8. 103; 214), Propertius (4. 9. 1) and Ovid (Met. 9. 140).

³¹ e.g. Od. 6. 109. 228; Hom. Hym. Aphr. 82, 133; Aesch. Supp. 149; Soph. Elec. 1239; O.C. 1056, 1321; Aj. 450.

³² Apul. Met. 2. 5 (serit blanditias, invadit spiritum, amoris profundi pedicis aeternis alligat) comes as close as anything.

³³ Theoc. 3. 42, where the point is, as in Catullus, the suddenness and totality of the seizure by love.

³⁴ I consign to a note Kroll's belief (op. cit. (n. 28) ad 109) that Catullus was unclear whether the barathrum was a hole or a canal; 110 does not pace Kroll require the latter; cf. also Copley, op. cit. (n. 24) 139–40; Elder, op. cit. (n. 24) 103.

³⁵ Beside Fordyce and Quinn (next note) cf. Copley, F., Catullus – The Complete Poetry (Ann Arbor, 1957)Google Scholar, who translates barathrum as ‘gulf’, suggesting sea-imagery, and Baehrens (op. cit.(n. 25), Commentarius) ad 108 who translates barathrum as periculum, pernicies and regards the water-image as being from the sea.

³⁶ Catullus, , The Poems (London, 1970) ad 108–10Google Scholar.

³⁷ cf. Thes. Ling. Lat. s. vv. absorbeo, aestus, and Lewis and Short s.v. vertex.

³⁸ Aen. 3. 420 f. (cf. Od. 12. 94). Williams, R. D., P. Vergili Maronis Aeneidos Liber Tertius (Oxford, 1962) ad locGoogle Scholar. observes that Vergil's phraseology is markedly similar to Catullus' words in 107 f. There may be some conscious or unconscious connection, but I doubt that this shows that Vergil thought Catullus' barathrum had anything to do with the sea. In Sidon. Carm. 9. 203 we find Syrtium barathrum, but this probably refers not to the sea, but the desert (as well as having overtones of ‘hiding place’ or the like). For that usage elsewhere in Sidonius cf. Carm. 16. 91 ff.; 17. 13 f; Epist. 8. 12. 1 and 4. Carm. 5. 263 f. and 549 are ambiguous, but probably also refer to desert. In Carm. 5. 594 Syrtes might be either land (like the Alps immediately before) or sea (like the mare magnum immediately after). I wonder if in Carm. 9. 203 f. Sidonius might have had in mind the information in Seneca, Epist. 90. 17: non in defosso latent Syrticae gentes quibusque propter nimios solis ardores nullum tegumentum satis repellendis caloribus solidum est nisi ipsa arens humus?

³⁹ Excluding, therefore, references to the underworld (Iliad 8. 14; Ap. Rhod. Arg. 2. 642; 4. 1698; Nonnus, 2. 70, 378; 4. 55; 27. 84; 30. 159; 36. 101, 204; 44. 260; Lucian, , lkaromen. 33Google Scholar) the Athenian punishment-pit (Herod. 7. 133; Aristoph. Eq. 1362, Nub. 1450, Ran. 574, Plut. 431 (cum schol.), 1109; Xen. Hell. 1. 7. 20; Plato, , Gorg. 516dGoogle Scholar; Plut. Arist. 3; Suda s.v. υητραγ⋯ρτης Harp. s.v. β⋯ραβρα Bekker, , Anec. Graec. 1. 219Google Scholar) and other pits of various sorts (Plut. Lyc. 16 (the apothetai, into which deformed children were put)); Od. 12.94, Joseph, , B.J. 1. 405–6Google Scholar, Heliod. Aeth. 5. 2, Galen 17 (1) 10, Antig. Caryst. 135 (various caves, mostly sinister); Nonn.26. 107, 128, 136;.30. 129; 45. 282 (prisons); Phercyd. FGrH 3F51 b (a mantrap); Nonn. 9. 102 (a secure hiding place).

⁴⁰ Poseid. FGrH 87F89 (R. Timavus); Strabo 289 c (Lake Stymphalos/R. Erasinos); Ps. Aristot. Probl. 947a 19 f. (of Arcadia in general); Joseph, . B.J 1. 406Google Scholar; 3. 509 f. Strabo loc. cit. (quoting Eratosthenes, and using the Arcadian dialect form ξ⋯ρ∈θρον, Theophr. Hist. Plant. 3. 1. 2; 5. 4. 6, Plut. Mor. 557c, Paus. 8. 14. 1; 20. 1 (all of Pheneus). The appearance of β⋯ραθρον and its cognates in connection with marshy land belongs in the same context (Polyb. 3. 78. 8; 5. 80. 2; Diod. 1. 30; 16. 46. 5; Steph. Byz. s.v. β⋯ραθρα Non. Theoph. Ep. de cur. morb. 1, 446; Strabo 614c.) βαραθρώδης is used of the sea above the lost continent Atlantis in Philo, de incorr. mund. 26 = de mundo 21Google Scholar.

⁴¹ de rer. nat. 3. 966; 6. 606. For Greek examples, cf. n. 39. Later Latin authors followed Lucretius' lead; Verg. Aen. 8. 245; Val. Flacc. 2. 86, 192; Stat. Theb. 1. 85; 8. 15; Colum. 10. 62; Apul. Met. 2. 6 (cf. 2. 5), etc.

⁴² Whatever the significance of the evil overtones of this, and other, metaphorical usages of barathrum may be for the interpretation of the poem as a whole; see below, p. 132.

⁴³ Lydus is replying to Pistoclerus' iam excessit mi aetas ex magisterio tuo, and the whole context here and throughout the play of Lydus' vain attempts to keep Pistoclerus on the straight and narrow suggests that in despair Lydus feels he should be punished for his failure.

⁴⁴ It is also the only example of metaphorical use of praeligare in Lewis & Short and the Oxf. hat. Dict.

⁴⁵ Esp. 371–4; notice desidiabula (376), gerulifigi (381).

⁴⁶ cf. Platner, S. B./Ashby, T., Topographical Dictionary of Ancient Rome (Oxford, 1919), 126Google Scholar.

⁴⁷ Application of barathrum to the stomach is found also in Hor. Epist. 1. 15. 31; Mart. 1.87. 4; Sidon. carm. 5. 325; Venantius Fortunatus 6. 7. 3.

⁴⁸ Sources relevant to the Pheneate barathra but not mentioning Herakles include Diod. 15.49. 5; Theophr. Hist. Plant. 3. 1. 2; Eratosth. ap. Strab. 389c; Pliny, , NH 31. 54Google Scholar; Ael. Nat. Anim. 3. 58; (Athen. 331 D/E confuses the Pheneate R. Aroanios with the Kleitorian one). For a possible further reference to Herakles building the barathra see next note. The ultimate outlet of the water was the R. Ladon (Eratosth. loc. cit.) See also Baker-Penoyre, J., JHS 22 (1902), 228–40CrossRefGoogle Scholar for recent appearances and disappearances of the lake. At present (Sept. 1978) the plain is dry.

⁴⁹ The association of Herakles and feats of engineering concerning water is not limited to Pheneus; cf. e.g. Paus. 9. 38. 7; Diod. 4. 18. 6 (Lake Copais); Strabo, 7, fr. 44 (Lake Bistonis); Diod. 4. 18. 5 (Pillars of Hercules); ibid. 6; Sen. Here. Fur. 286 (R. Peneius); Diod. 4. 35. 8, Strabo 458c (R. Achelous); ? FGrH 40F 1, line 2 (Orchomenus). Robert, C. (Die griech. Heldensage ii (Berlin, 1921), pp. 528–9)Google Scholar thought that FGrH 40F 1, line 15 referred to Pheneus; Jacoby (ad loc.) and Bölte, , RE xix 1978Google Scholar, take it as referring to Aetolia (R. Achelous). The state of the text (on the Tabula Albana) is far too poor to allow any decisive conclusion.

⁵⁰ In view of these passages Vergil's connection of Evander and Pheneus (Aen. 8. 165) may (pace Servius) be pointful, for Evander also had Heraklean connections; he is first encountered sacrificing to Herakles (102 f.) and his son's belt (Aen. 10,495 f.) has associations with Herakles.(For this latter point I am indebted to an unpublished paper by Margaret Hubbard.) Dion. Perieg. 347 (Geog. Graec. Min. ii. 124) and Eustathius ad loc. (GGM ii. 278. 9 f.), bringing Evander to Rome from ⋯ κατ⋯ 'Aρκαδ⋯αν Kυλλήνη (i.e. the area of Pheneus) may also be registered in this context, as also the claim (Varro and Greek authors ap. Serv. Aen. 3. 167): that the Trojan forebear Dardanus came from Pheneus

⁵¹ Toup, J., Epistola Critica (London, 1767), 162 fGoogle Scholar.; idem., Analecta Alexandrina, 13 f. (non vidi); Thraemer, E., Hermes 25 (1890), 55Google Scholar; Powell, J. U., Collectanea Alexandrina (Oxford, 1925), 40–1Google Scholar; Barigazzi, A., Athen. 26 (1948), 50 fGoogle Scholar.

⁵² Callimachus i (Oxford, 1949), ad Aetia i, fr. 35Google Scholar. He proposed no alternative attribution. H. van Herwerden (quoted in Wilhelm, A., JöAI 14 (1911), 173Google Scholar) thought the lines came from Arctinus' Iliupersis, but what his reasons were is not stated. van Groningen, B. A. (Euphorion (Amsterdam, 1977))Google Scholar puts the lines among Fragments douteux (no. 192) and claims the style is much too simple for Euphorion (or Callimachus). But it is not obvious that the style is damagingly simpler than e.g. frr. 2, 23, 34, 58, 84, 92, 96, 98, etc. (P). Moreover there are features of the lines that would at least make their attribution to archaic epic dangerous, (i) ⋯ναμπ⋯χονοι is unknown to archaic epic, and indeed to most authors except the Hellenistic Pythaenetus (FGrH 299 F 3). (ii) νóσφι κρηδ⋯μνοι? is not an archaic epic way of saying ‘without a headdress’ (Hes. Op. 91, 113; Scut. 15 are hardly comparable, and anyway the presence of ãτερ makes a difference), (iii) The use of ἠοῖαι can be paralleled in [Hes.] Scut. 396 and Hom. Hym. Herm. 17, but is more common in Hellenistic poets (Call. fr. 59. 18; 75. 10; Epigr. 20; Ap. Rhod. 2. 688, 899). In short, the lines do not look un-Hellenistic. de Cuenca, L. Alberto (Euforion de Calcis (Madrid, 1976))Google Scholar prints the lines with the other fragments of Chiliades as no. 79 but marks it Incertum.

⁵³ As is clear from their content and the identity of the parallel sources (Herodotus and Curtius in the first two cases).

⁵⁴ Polyb. 2. 16. 3 f. discusses the Phaethon story as an example of the sort of poetic story that ought to be excluded from historiography.

⁵⁵ Plin. NH 31. 54.

⁵⁶ cf. RE iv A 1005Google Scholar. It is interesting to note that the original R. Crathis, after which the one at Sybaris was named, was near Pheneus (Hdt. 1. 145; Strab. 386c; Paus. 8. 15. 8–9; 18. 4). Can we conceivably detect here the basis on which a poet might effect a transition from one story to the other?

⁵⁷ Call. Aet. i, fr. 35 (Pf); Lycophr. Alex. 1153; Tzetz. ad Lycophr. Alex. 1141; Schol. Lycophr. 1155; Schol. Iliad. 13. 66; Apollod. Epitome 6.22; Iamblichus, , Vita Pythag. 8. 42Google Scholar.

⁵⁸ Diod. 11. 90. 3–4; 12. 9 f. I take the expulsion of the former Sybarites from Sybaris/Thurii (a different site), and their descendants' sufferings at the hands of the Bruttians to be irrelevant, since they do not concern the city on the Crathis.

⁵⁹ Solinus, 2. 10. Intriguingly, the founders are said to have been the Troezenians and Sagaris, son of Locrian Ajax. Could this be another hidden connection between the three Apolline stories? cf. n. 56.

⁶⁰ In these versions her anger was due to (i) the murder of some Crotonian ambassadors (Phylarchus, FGrH 81 f45), or (ii) the murder, at altars, of the supporters of a deposed ruler (Heracleides, On Justice, fr. 49 Wehrli) or (iii) an occasion on which a master stopped beating a slave not when the latter took refuge at an altar but when he fled to the tomb of the master's father. (Steph. Byz. s.v. Σ⋯βαρις; cf. Timaeus, FGrH 566f 50).

⁶¹ e.g. Momigliano, , Secondo Contributo alla Storia degli Studi Classici (Rome, 1960), 448Google Scholar. For the opposite assumption see e.g. Baker–Penoyre, op. cit. (n. 48), 236, 240. Diodorus’ reference (15. 49. 5) to the lake being empty ⋯ν τοῖς προτ⋯ροις χρóοις is also probably (pace Baker–Penoyre) taken over from his source.

⁶² Apollod. Epitome 6. 20 f.; Tzetz. ad Lycophr. Alex. 1141.

⁶³ e.g. Momigliano, op. cit. (n. 61), 446.

⁶⁴ Miscellanea di Studi Alessandrini in memoria di Augusto Rostagni (Turin, 1963), 167–8Google Scholar.

⁶⁵ A. Wilhelm, op. cit. (n. 52), 163; cf. Momigliano, loc. cit., Manni, loc. cit.

⁶⁶ e.g. Timaeus, Lycophron (cf. Momigliano, loc. cit.)

⁶⁷ Momigliano, op. cit. (n. 61), 448 f.

⁶⁸ Most recently, L. Alberto de Cuenca, op. cit. (n. 52), 11 f. and B. A. van Groningen, op. cit. (n. 52), 4 f.

⁶⁹ Heyne, C. G., Excursus III ad Bucolica, p. 258 WagnerGoogle Scholar; Meineke, A., Analecta Alexandrina (Berlin, 1843), 15Google Scholar. The following counter-arguments may be advanced, (i) The marginale in the Suda (π∈ρ⋯ χρησυων ὡς 4ι⋯ χιλωων ⋯τ⋯ν ⋯ποτ∈λο⋯νται) implies that someone thought the reading ⋯τ⋯ν correct, (ii)χιλι⋯ς can mean a period of 1,000 years (cf. Alexan. Aet. fr. 4. 4 (Powell) and the title of a work of Asinius Quadratus variously called ‘PωμαÏκή χιλι⋯ς and ‘PωμαÏκή χιλι⋯ς χιλι∈τηρ⋯ς (testimonia in Peter, H., Historicorum Romanorum Reliquiae² ii (Leipzig, 1906), 142 f.)Google Scholar). (iii) The passage of Aetia i where Callimachus describes the ἄ∈ισμα διην∈κ⋯ς as written ⋯ν πολλα⋯ς χιλι⋯σιν hardly encourages the idea that a poet of Euphorion's type would willingly label a poem Chiliades in this sense, (iv) The text of Suda as it stands is perfectly acceptable linguistically, whereas the proposed emendation (συν⋯γ∈ι δι⋯ χιλ⋯ιν ⋯π⋯ν meaning ‘collected together in groups of 1,000 verses') is less easily paralleled; it would have to be either an instrumental construction analogous to δδ⋯ λóγων συγγ⋯γν∈σθαι (e.g. Plat. Polit. 272 B) or a (metaphorical) spatial one like δι⋯δ⋯κα ⋯παλξ⋯ων (Thuc. 3. 21), neither of which seems very happy, (v) Above all, the point of the poem (ὡς δ⋯κην δοῖ∈ν, κἄν ∈ἰς μακρ⋯ν) seems to require an indication of time in the title.

⁷⁰ Soph. Phil. 758; Xen. Cyr. 1. 4. 28; 5. 5. 41; idem, Mem. 4. 4. 5; idem, Oec. 9. 10; idem, Cyneg. 5. 3; Lys. 1. 12; Plat. Rep. 328B.

⁷¹ Hdt. 3. 27. 3; Aristoph. Plut. 1045.

⁷² Thuc. 6. 15. 4; 6. 91.

⁷³ Plat. Tim. 22D.

⁷⁴ Thuc. 5. 14. 1 and Luc. Nigr. 2 respectively.

⁷⁵ Van Groningen (loc. cit.) further objects that even the riches of Greek myth and history may have been unable to provide Euphorion with a large enough store of stories where the fulfilment of oracles took 1,000 years; but it is better to approach the interpretation of the Suda without any presuppositions on that score. For one thing, we do not know on what scale Euphorion treated his stories or, therefore, how many he needed to fill his poem. But assuming that his interest was in writing poetic narratives (the affectation of teaching his enemies a lesson being merely an excuse) it could be that the scale was fairly large.

⁷⁶ Quaest. Conv. 676f (=fr. 84P = 89 v. Gron. = 20 de Cuenca); ? ibid. 683B (=fr. 175P = 190 v. Gron = Dub. 1 de Cuenca).

⁷⁷ op. cit. (n. 51), 51 n. 3 (cf. frr. 57, 86, 87, 101, 196 (P)). Some but not all of the other arguments advanced here about the attribution of the Apolline stories to Euphorion appear also in Barigazzi's article.

⁷⁸ The connection of Herakles/Pheneus/Elean War was already known to Callimachus (cf. p. 124, Appendix I). It might also be that the association of Apollo and Pheneate flooding in Call. Hymn 4.71 (cf. Appendix I (iii)) assisted the invention.

⁷⁹ One might also note the temple of Apollo mentioned by Pausanias 8. 15. 5. Stories linking Apollo and Herakles in an amicable fashion are rare enough (U. von Wilamowitz-Moellendorf, Pindaros (Berlin, 1922), 80; Parke, H. W. & Boardman, J., JHS 77 (1957), 278)CrossRefGoogle Scholar; another one is the story that Gytheium was a joint foundation of Apollo and Herakles as a gesture of reconciliation after the theft of the tripod (Paus. 3.21. 8). One might hypothesize that knowledge of the Heraklean Apollo temple at Pheneus combined with knowledge of this version of Gytheium's foundation could have helped inspire the idea of the tripod being taken to Pheneus.

⁸⁰ cf. e.g. Fordyce, op. cit. (n. 24) ad. loc, Latte, K., Philol. 90 (1935), 154Google Scholar.

⁸¹ Catullus' phrase refers to victims of the Minotaur; Euphorion's (fr. 24c61 Van Groningen = 38 c, 59 de Cuenca) to the death of Comaetho's father, eaten by wild beasts. For the comparison, cf. Latte, loc. cit. (n. 80).

⁸² Maas, E., Hermes 24 (1889), 528 fGoogle Scholar.; Castiglioni, L., Studi Alexandrini, i: Arianna e Theseo (Pisa, 1907)Google Scholar; Barigazzi, A., Miscellanea…Rostagni (cf. n. 64 above), 422 fGoogle Scholar. idem, Maia 17 (1963), 163; Fordyce, op. cit. (n. 24), 272.

⁸³ The large number of fragments is one indication of this. The catchphrase Cantores Euphorionis is another (cf. my ‘Cantores Euphorionis’, Papers of the Liverpool Latin Seminar i (= Area (Classical and Medieval Papers and Monographs) ii) (Liverpool, 1977), I f. and ‘Cantores Euphorionis Again’, CQ N.S. 29 (1979), 358 fCrossRefGoogle Scholar.).

⁸⁴ cf. Wiseman, op. cit. (n. 3), 47–59 (and further literature quoted there).

⁸⁵ He knew Euphorion's poetry (cf. Erot. Path. 13, 26, 28) and, if the reported tastes of the emperor Tiberius (Suet. Tib. 70) are a consistent indicator, his own poetry may have owed something to Euphorion.

⁸⁶ cf. Wiseman, op. cit. (n. 3), 72 n. 39.

⁸⁷ RE Suppibd. iii. 1021–2Google Scholar. The geographical proximity of Pheneus and Stymphalos (and the fact that L. Stymphalos is drained by barathra) would help.

⁸⁸ FHG iii. 151, fr. 8.

⁸⁹ Diod. 4. 33; Paus. 8. 14. 1 f.

⁹⁰ cf. above p. 124.

⁹¹ cf. Friedrich, , Catulli Veronensis Liber (Leipzig/Berlin, 1908), ad 109Google Scholar; de Gubernatis, M. Lenchantin, Catullo (Turin, n.d.) ad 112Google Scholar; Granarolo, J., L'Oeuvre de Catulle (Paris, 1967), 123 n. 1 (all without arguments)Google Scholar.

⁹² Wiseman, T. P. (letter, 22 April 1978). The point will be discussed in his forthcoming book Clio's Cosmetics: Three Studies in Greco-Roman Literature (Leicester University Press)Google Scholar; cf. also Cairns, F., ‘Catullus I’, Mnemos. 22 (1969), 153 ffCrossRefGoogle Scholar.

⁹³ There is at least as much (or as little) evidence that this synchronism had been thought out before Catullus' time as there is for the other one hypothesized above (p. 130); more if Catullus' use of it and claim that it derived from Grai count as evidence.

⁹⁴ Friedrich, op. cit. (n. 91) ad 107 (grief at loss of Protesilaus); Riese, A., Gedichte des Catulls (Leipzig, 1884), ad 117 (misfortune of losing her husband)Google Scholar; Baehrens, op. cit. (n. 25, Commentar.), ad loc. (fate of unrequited love at Protesilaus' death); Wohlberg, CP 50 (1955), 44 (‘love in grief’). All these views have difficulty' with 117–18 and require detulerat (108) to signify the speed of Laodamia's fall rather than to be taken at its face value (viz. that Laodamia's falling in love preceded her loss of Protesilaus). cf. also n. 118.

⁹⁵ cf. Lieberg, G., Puella Divina (Amsterdam, 1962), 209 ffGoogle Scholar.

⁹⁶ cf. above p. 120.

⁹⁷ Artemis: Soph. Elec. 1239; Athena: idem.Aj. 450. In Od. 6. 109 it is applied to Nausicaa, who is also compared to Artemis.

⁹⁸ Atalanta: Soph. O.C. 1321; Danaids: Aesch. Suppl. 149. Antigone and Ismene, who spend their lives with their blind father/brother Oedipus instead of marrying are τ⋯ς διστóλους ⋯δμῇτας ⋯δ∈λφ⋯ς (Soph. O.C. 1056).

⁹⁹ Ars Am. 2. 185.

¹⁰⁰ p. 120 and n. 33.

¹⁰¹ cf. e.g. Prop. 2. 3. 47 f.; Hor. Od. 1. 33. 10 f.; 3. 9. 18 f.

¹⁰² flagrans…amore (73); avidum…amorem (83); sed tu horum magnos vicislifurores… (129).

¹⁰³ cf. Lieberg, op. cit. (n. 95), 241 f.; Pepe, L., GIF 6 (1953), 107 fGoogle Scholar.; Witke, op. cit. (n. 24), 34 f.; Quinn, op. cit. (n. 24), 181 f., 266. Others have, of course, questioned Laodamia's guilt (e.g. Pennisi, G., Emerita 27 (1959), 89 f., 213 f.Google Scholar; Bardon, H., Propositions sur Catulle (Brussels, 1970), 98Google Scholar; Hering, W., Acta Classica Univ. Scient. Debrecen. 8 (1972), 36 fGoogle Scholar. (I am indebted to Dr Hering for sending me a copy of this article).

¹⁰⁴ LSJ⁹ s.v. π⋯πτ∈ιν BIII2; ⋯μπ⋯πτ∈ιν 4b.

¹⁰⁵ cf. nn. 33, 100.

¹⁰⁶ n. 40 above and Appendix II. (It would be unduly pedantic to object that with sink-holes it was sometimes the case that the water that disappeared into them reappeared elsewhere.)

¹⁰⁷ cf. a similar point in Witke, op. cit. (n. 24), 36.

¹⁰⁸ Witke, op. cit. (n. 24), 36; Offerman, , Philol. 119 (1975), 65CrossRefGoogle Scholar.

¹⁰⁹ Notwithstanding the various ‘Chinese-box’ analyses of poem 68 (Skutsch, F., RhM 47 (1892), 138Google Scholar f. = Kl. Schr. (Leipzig/Berlin, 1914), 49Google Scholar; J. Wohlberg, op. cit. (n. 94), 42; Vretska, K., WS 79 (1966), 313 ffGoogle Scholar.; Hering, W., Wiss. Zeitschr. der Univ. Rostok, Gesellsch. und sprachwiss. Reihe 19 (1970), 599 fGoogle Scholar; idem, op. cit. (n. 103), 40 f., 51 f.) which would bring 105/7–18 into at least formal relation with passages as diverse as 51–6 (plus eight conjecturally lost lines) (Hering), 73–86 (Skutsch) and 85–104 (less 91–100) (Vretska, Wohlberg), one is surely entitled to find parallels and contrasts for the barathrum passage anywhere in the poem. The diversity of the schemata hitherto proposed suggests that no single one can assimilate all the cross-references that the poem contains.

¹¹⁰ The centre of a ‘proper’ home, cf. Oxf. Lat. Diet. s.v. focus 1 a, 2ab.

¹¹¹ cf. above p. 117. Notice also: (i) Laodamia’ s passion exceeds the attachment of a grandfather to his new grandson and a dove to her mate (119 ff.). Doves were an exemplum of conjugal propriety and fidelity (e.g. Prop. 2. 15. 27); the grandfather and grandson have a legitimate blood-relation. Laodamia's exceeding of what was proper in such cases is another sign of the impropriety of her love. (ii) conciliata (130), a neutral word instead of ‘married', and the overtones could be sinister (cf. Lucr. 5. 963 f., Suet. DJ 50).

¹¹² Lieberg, op. cit. (n. 95), 253 f.; Reynen, H., Mus. Helv. 31 (1974), 150 fGoogle Scholar.

¹¹³ But cf. Witke, op. cit. (n. 24), 32 ff.

¹¹⁴ It is the only example of eri used of the gods collectively in Thes. Ling. Latin. Heusch, H., Das Archaische in der Sprache Catulls (Bonn, 1954), 42 fGoogle Scholar. regards the use of erus in Catullus in general as a deliberate archaism, but has no specific comment about 68. 76, 78.

¹¹⁵ cf. above n. 18. A comparison of 68. 71–3 with Theocr. 2.103–4 (drawn by K. Gantar, Grazer Beiträge 5 (1976), 117 f.) helps to highlight the fact that Catullus has made Lesbia step on the threshold, for in Theocritus we read ⋯γὼ δ⋯ νιν ὡς ⋯νóησα/ἄρτι θ⋯ρας Ὁπ⋯ρ οὉδóν ⋯μ∈ιβóμ∈νον ποδ⋯ κο⋯ρῳ.

¹¹⁶ A referee suggests that the prominence of the ianua-motif might, if anything, draw attention to the propriety of Hercules' marriage.

¹¹⁷ cf. Part I, above.

¹¹⁸ The eventual acceptance into Olympus of falsiparens Amphitryoniades (the product of one of Jupiter's furta) might also be seen as an example of that capacity to control anger which Catullus implicitly praises in Juno (138–40).

¹¹⁹ Unrequited in the case of Catullus; and the same may be true of Laodamia in 107–18. Contrast qui tamen indomitam ferre iugum docuit (118) with ut semel es flavo conciliata viro (130). The second two similes deal with Laodamia's passion after the conciliatio, the first with the love that compelled her to undergo the conciliatio (cf. Hering, op. cit. (n. 103), 51).

¹²⁰ Thermopylae: Suda s.v. Θ∈ρμοπ⋯λαι Et. Mag. 447.19; Schol. Ar. Nub. 1047; Strabo, 428; Hdt. 7. 176. The association of Oeta and Hercules' immortality is very strong: cf. Accius (ap. Cic. de nat. deor. 3. 41); Cic. Tusc. Disp. 2. 19; Prop. 1. 13. 24; 3. 1. 31 (Oetaeus deus); Ovid Ibis 345 (Oetaeus gener); idem.Met. 9. 135 ff; idem.Heroid. 9. 147; Lucan. 8. 800; 3. 177; Stat Theb. 4. 158, etc. Oeta was of course associated with marriage in its own right (cf. Cat. 62. 7; Verg. Ecl. 8. 30; Culex 203). It has recently been independently argued (Robson, A. G., TAPA 103 (1972), 433 f.)Google Scholar that 53 should read cum tantum arderem quantum Trachinia rupes. If that were correct then Catullus' love is described in purely Herculean terms.

¹²¹ cf. Witke, op. cit. (n. 24), p. 42.

¹²² cf. above, p. 132.

¹²³ e.g. Hor. Od. 1. 12. 25; 3. 3. 9; 4. 5. 35–6; Epp. 2. 1. 5 ff.; Cic. de leg. 2. 19; TD 1. 28; de nat. deor. 2. 62; Curt. 8. 5. 8; Xen. Symp. 8. 29; Aristot. (Poet. Mel. Graec. 842, 9 f.); Arr. Anab. 4. 8. 3; Dio Chrys. 69. 1, etc.

¹²⁴ Soph. Trach. 811; Eur. HF 183; Aristoph. Nub. 1049.

¹²⁵ Esp. Pindar; see Galinsky, G. K., The Hercules Theme (Oxford, 1972), pp. 23 ffGoogle Scholar.

¹²⁶ Most famously, Prodicus, Diels-Kranz 84b 1, 2. cf. also Cic. de off. 3. 25; Sen. Const. 2. 1; Epict. 1. 6. 32 f. Also, the allegorization of the Labours as conquests over one's own improper desires and pleasures in e.g. Herodorus, FGrH 31f14; Dio Chrys. 5. 22 f. (and cf. Epict. 2. 16. 44f.;Apul. Florid. 22).

¹²⁷ cf. Epict. 3. 24. 13 f. Note also the idea in e.g. Cic. de fin. 3. 65 (= Chrysipp. SVF iii. 84. 6); 2. 118–19; de off. 3.25, that one should play one's role as part of the community of mankind and not seek to live in self-centred separation from it, an idea exemplified by Hercules.

¹²⁸ See now C. W. Macleod, op. cit. (n. 4), 82 f.; and above, p. 121.

¹²⁹ cf. Part I above.

¹³⁰ cf. Clausen, W. V., HSCP 74 (1970), 90 fGoogle Scholar. for comments to this effect.