² Courtney, E., ‘Ovidian and NonOvidian Heroides’, Bulletin of the Institute of Classical Studies, xii (1965), 63 ff.CrossRefGoogle Scholar

³ Lachmann, Karl, Kleinere Schriften (Berlin, 1876), ii. 56.Google Scholar

⁴ Cf. Dörrie, H., Untersuchungen zur Überlieferungsgeschichte von Ovids Epistulae Heroidum (Nachrichten Akad. Wiss. in Göttingen, Phil.-Hist. Klasse, v, 1960), 171Google Scholar; ‘Die dikterische Absicht Ovids in den Epistulae Heroidum’, A.u.A. xiii (1967), 51 ff.Google Scholar, for which see p. 355 n. 3 below.

⁵ Palmer, A., Ovid: Heroides XIV (London, 1874), xvii f.Google Scholar

⁶ Cf. Shuckburgh, E. S., Ovid: Heroides XIII (London, 1885), 164 ff.Google Scholar; Louis C. Purser in Introduction to Palmer, A., Ovid: Heroides XXI, (Oxford, 1898), xx f.Google Scholar Shuck-burgh appears to have suspected extensive interpolations; Purser concluded that, despite its inferiority, ‘it is difficult to believe that the poem is not by Ovid’. Cf., for example, D'Elia, , Ovidio, pp. 157 ff.Google Scholar; Wilkinson, L. P., Ovid Recalled (Cambridge, 1955), pp. 87, 96.Google Scholar

⁷ The Heroides may rightly be called Ovid's most rhetorical work: cf. Bonner, S. F., Roman Declamation (Liverpool, 1949), 152 ff.Google Scholar; Higham, T. F., ‘Ovid and Rhetoric’, Ovidiana (ed. Herescu, N. L., Paris, 1958),. 32 ff.Google Scholar

¹ This phrase is borrowed from Green, P., ‘Venus Clerke Ovyde’, Essays in Antiquity (London, 1960), 122.Google Scholar

² Fides lies in meaning somewhere between ‘sincerity’ and ‘authenticity’ or ‘credibility’ —cf. Quint. Inst. Or. 5. to. 8 and 10 ff.; cf. Allen, W., ‘“Sincerity” and the Roman Elegists’, CPh. xlv (1950), 146 ff.Google Scholar, reprinted in Sullivan, J. P. (ed.), Critical Essays in Roman Literature: Elegy and Lyric (London, 1962).Google Scholar

³ Purser, loc. cit. xxi.

⁴ Cf., for example, Met. 9. 134 ff., with Her. 9.1 ff.; Met. 9. 182 ff. with Her. 9. 85 ff.; Met. 9. 149 with Her. g. 151. Zingerle, W., Untersuchungen zur Echtheitsfrage der Heroiden Ovid's (Innsbruck, 1878)Google Scholar, cites various parallels and imitations between Heroides 9 and other works of Ovid, believing that this is evidence of the poem's authenticity (pp. 21–3); it is clear, however, that the argument can be used the other way—it is not suggested that the author of the ninth Epistle was unfamiliar with Ovid's technique (Cf., for example, Her. 9. 73 and A.A. 2. 219; Her. 9. 68 and A.A. 2. 400; Her. 9. 136 with Her. 4 and Fast. 3. 20; Her. 9. 102 with Her. 11. 82, Am. I. 7. 20, Rem. 640, etc.).

¹ It can, of course, be said that rhetorical poets often wish to have it both ways—but not usually as blatantly as this.

¹ Courtney, loc. cit.

² Shuckburgh, , op. cit. 164.Google Scholar

³ Occultatio is the introduction of material by stating that there is no need to mention it—a fine example may be found at Statius, , Thebaid i. 3 ff.Google Scholar (but see Rossbach, O., ‘Eine übersehene Erwähnung des Antimachos’, BPhW(1915), 253–6)Google Scholar—the device is common enough.

¹ The reading referetur is extremely dubious; for a discussion cf. Palmer, op. cit., ad loc.

² Line 81 is rejected by Merkel, Palmer, et al. Palmer states (I think correctly) that it is ‘too bombastic for Ovid’ (1898, p. 367). In the event of the poem's not being by Ovid, it could be retained. It is relevant to the situation and was certainly in Planudes' MS. Line 82 ante pedes dominae pertimuisse minas is also rejected in whole or in part because dominae pertimuisse minas is also found at 74. It is supposed that the copyists introduced it in error for a second time at 82. I suspect that it should be excised from 74 instead and was interpolated there by mistake. It is more suitable to the context at 82. Lines 73–4 are in facts taken from A.A. 2. 219 f., where the pentameter reads lanas excoluisse rudes, which could be read at 74 here. Line 83 is justifiably excised as a meaningless and absurd interpolation. This hexameter probably dropped out at an early stage—Planudes had a lacuna in his manuscript; neither eximiis pompis immania semina laudum (marginal reading in P) nor praeconia summa triumphi (G., etc.) makes good sense: the former reading being apparently taken from Statius, , Ach. I. 188:Google Scholar cf. Dilke, O. A. W., Statius: Achilleid (Cambridge, 1954), 98Google Scholar—of course Statius was a frequent imitator of Ovid (cf. Luehr, G., De Statio in Silois priorum poetarum Latinorum imitatore, diss. Konigsberg, 1880Google Scholar; Deipser, B., De Statio Vergilii et Ovidii imitatore, diss. Argentor. 5, 1891)Google Scholar. It is worth noting, in passing, that at Her. 41 (which occurs in a much disturbed passage) the phrase o nimium nimiumque oblite tuorum is repeated at Statius, , Theb. 7. 547Google Scholarnimium mitis nimiumque oblite tuorum.

¹ Cf. the remarks of H. Dörrie in his article in A.u.A. cited p. 349 n. 4: ‘Ein überlang Teil des Briefes, 118 von 168 Versen (70 Prozent), sind eine von indigniertem Hohn gewürtzte Anklagerede gegen Hercules’ (p. 52).

² The refrain can be taken as an indication of non-Ovidian authorship, but both Palmer (1874, p. 88) and Zingerle, op. cit. 24, argue that it favours the traditional ascription, as ‘no imitator would have ventured to introduce anything unusual with Ovid himself’. As a parallel, however, they quote the refrain in the paraklausithuron, Amores 1.6, with its repetitions of excute poste seram—so that an imitator did have some warrant. It is important to notice that the refrain in the Amores is far more apt and more cleverly used; in the Heroides, as I suggest, it is both inappropriate and unimaginatively used—in fact it seems to indicate a poverty of invention rather than anything else.

¹ Purser, , loc. cit. xx.Google Scholar

² Cf. Purser, , loc. cit. xx n. 3.Google Scholar

³ H. Dörrie in his article in A.u.A. has discussed the ninth epistle at some length; he mentions Courtney's article in a footnote (52–3) and, although he recognizes the validity of metrical arguments, he suggests that they may be countered by the possibility of Ovid's experimenting or by interpolation. I hope that my views go some way to show that the objections are too integral to be dismissed as caused by interpolation and that, if Ovid is experimenting, it is a unique and unworthy attempt. However it is interesting to note that Dörrie proposes two postulates: (i) that each heroine is individually characterized by a number of small traits ‘namentlich durch kleine Verstösse gegen Geschmack and Schicklichkeit’; and (ii) that in the characterization of the heroines there is regularly alteration, even downright perversion, within the poem. He admits that these cannot unfortunately be demonstrated from all the epistles; but the main example he discusses is none other than the Ninth, which he rightly characterizes as an ‘ungewöhnlich dramatisch Brief’ (pp. 51–2). His analysis of the poem along these lines is of much interest.

¹ Shuckburgh, , op. cit. 164.Google Scholar

² The use of caerulus Nereus is one of the themes mentioned by Persius as characteristic of the oratio exaggerata of rhetorical poets (Sat. 1. 94).

³ (i) decolor (3): unparalleled in its metaphorical sense in Ovid, who always uses it in a purely physical sense. The passage in Virgil, Aen. 8. 326, cited as a metaphorical use, is in fact a reference to the change from the gold to the bronze age, and is wholly different from the usage here. For the normal Ovidian usage cf. A.A. 3. 130.

(ii) In 45 the phrase arbiter Eurystheus irae Iunonis iniquae, quite apart from the clumsiness of the double genitive (to avoid which Housman proposed substituting astu or furto for irae), is meaningless. Arbiter should mean ‘minister’ or ‘manager’, as Palmer remarks; in the 1874 edn. ad loc. he cites the phrases arbiter bibendi and arbiter Hadriae as parallels, but in 1898 he recognized that they were not really equivalent (p. 363). However, as the verse seems to be corrupt, not much weight can be put on this disparity.

(iii) nympha (50, 103): Palmer (1874) on Ep. 1. 27 suggests that ‘the use of nympha for any young girl... is not classical’. Here the word is used of two females who are not strictly speaking nymphs (Astydameia and Omphale). At 1. 27 we should probably read nuptae with Heinsius as making better sense. The word nympha is, however, used in its non-technical sense at Heroides 16. 128; but of course this may not be Ovid's work, and for this passage we rely on the editio Parmensis of 1477. Ovid does occasionally produce usages more characteristic of the Silver Age than of the Augustan era, and this could be regarded as an example, although two instances may be thought a little suspect.

(iv) In 9–10 the phrase nox... una non tanti ut tantus is manifestly clumsy. This is the reading of P and G and only the codex Moreti reads tanta. Palmer reads tanti, arguing that ‘non tanta can scarcely stand for satis Tonga’ (1898, p. 360). Tanti has been condemned by Madvig, (Opusc. ii. 194)Google Scholar and by Housman, who wished to read luctanti to obviate the difficulty; Slichtenhorst suggested satis.

(v) Line 18 is of dubious meaning: Hercule supposito sidera fulsit Atlans. Showerman in the Loeb edition produces little better than gibberish. The whole point of the story is that when Hercules bore the load Atlas did not need to fulcire the heavens. A verb meaning ‘put down’ is required. An alternative suggestion translates Hercule supposito ‘by means of the substitution of Hercules’, a curious form of expression. One would want a phrase meaning ‘stood easy’, but if sidera is sound it cannot be obtained. Bentley deleted the distich.

(vi) There appears to be a confusion of legal metaphors in lines 107–8. The words rerum mensura tuarum and cede bonis both refer to the process of cessio bonorum, a process akin to bankruptcy, in which a debtor handed over all his property to his creditors: it is defined as ab universitate rerum suarum recedere (v. Palmer, 1878, p. 85). Palmer rightly argues that mensura is equivalent to universitas (cf. Juvenal, Sat. 14. 316). But in heres a slightly different metaphor enters. Palmer tried to explain it by arguing that ‘it was possible to alienate the right of inheritance by the form of cessio bonorum’. However this is not fully satisfactory, since laus is not part of the paterna bona which would be ceded by this process of disinheriting himself; rather it is the principal part of the bona which by bonorum cessio he is said to be ceding to her as creditor. There is a slight change of metaphor here but it cannot be said to be a particularly vicious one.

(vii) Other difficulties include 38 haesuros terna per ora canes (esuros some MSS. with Heinsius, hausuros Postgate); 129 ut (or ab?) Hercule victo; 156 vivus in igne fuit (situs Housman). These may well all be the result of MS. corruption.

(viii) There are a few careless repetitions: e.g. pacatum... pax (13, 15); irae... ira (45, 46—? corrupt); calathum... calathis (73, 76); armenti dives... dives ab ipsa suis (91, 96).

¹ Cf. for example, Her. 4. 99, 11. 13. Platnauer, M., Latin Elegiac Verse (Cambridge, 1951), 58ff.Google Scholar

² Courtney, loc. cit.; cf. Palmer, op. cit., 1898, pp. 509 ff.; Platnauer, op. cit., 58.

³ Platnauer, op. cit. 39; Palmer, op. cit., 1898, p. 356; in favour, Purser, loc. cit. xxx; Courtney, loc. cit.

⁴ e.g. Inachii (Bentley), Aonii (Merkel), Ismeni (Riese), et Echionii (Housman).

⁵ Palmer, op. cit., 1898, p. 371.

¹ Platnauer, op. cit. 60.

² Ibid. 60 n.

³ Courtney, loc. cit. 66.

⁴ The reading tegendo in 526 is plainly corrupt.

⁵ Cf. Platnauer, op. cit. 10.

⁶ In Epistle 7 we quote 9.18 per cent, disregarding one example (17) which is a locus desperatus.

⁷ Platnauer, loc. cit. 10.

⁸ Theodor Birt first used this argument against 15 in Rh.M. xxxii (1877), 390. De Vries cited 9 as a similar case in favour of Ovidian authorship of 15 (Epistula Sapphus, 1885, p. 141), quoted in approval by Purser, op. cit. 421.

⁹ Platnauer, op. cit. 118.

¹ It is worth noting perhaps that the incidence of hexameters beginning with spondees is also relatively high—9.52 per cent. Two other totals are similar to this (3 and 10) whereas the rest are lower (ranging from none to 6.6 per cent). In Epistle 3 (total incidence 9.09 per cent) one line (115) is dubious and condemned by Palmer; and in 10 (with a high incidence of 11.8 per cent), four examples may be deemed dubious—27 was condemned by Palmer, 31 is a locus desperatus, 95 was condemned by Bentley and van Lennep. 83 may be considered suspicious,as 85–6 were condemned by Bentley and 87–8 by Riese as interpolations, in which case 83–4 may also be dubious, being part of the same catalogue, which, it is argued, was interpolated because of 81–2. This would reduce the total percentage to 6.57 per cent. However, even if this is regarded as special pleading, the high total in 9 is still worthy of remark, when related to the high incidence of spondaic pentameters.

² Courtney, loc. cit.

³ Persius, Sat. I. 33 ff.

¹ Lee, A. G., ‘Authorship of the Nux’, Ovidiana (ed. Herescu, N. L., Paris, 1958), 468–9Google Scholar; cf. also E. J. Kenney on Am. 3. 5, in C.Q. N.S. xii (1962), p. 12.