I

Jupiter allows lovers to break their oaths with impunity (1.4.21–6):

‘Great thanks be to Jupiter: the father himself annuls whatever foolish love should rashly aver. Diana allows you to swear by her arrows with impunity, Minerva by her hair.’ Suspicion is justly cast on crines, for it is not easy to see why a lover, least of all a perjurious lover, should wish to aver his affection by invoking as a witness—of all things—the ‘hair’ of Minerva. That Minerva was fond of her hair is not in doubt: her turning the locks of Medusa, who had preferred her own to the goddess’, into snakes is testament enough to that.Footnote ³ But whether it was the sort of attribute whereby a lover could swear, whether it stood in the same relation to Minerva as the bow and arrow to Diana, is very far from certain. Commentators are at a loss for a parallel.Footnote ⁴ ‘Iuratur per id quod utrique deae carissimum,’ says Dissen, ‘quo contemto maxime irascitur quaeque; sitne ea corporis pars an telum, quid refert? poena ut eo ipso sumatur, per quod peieratum, non opus est.’Footnote ⁵ This is a perfectly reasonable inference, but it does not square neatly with the actual practice of oath-swearing in elegy. Lovers in this genre usually swear by (one of) two things: by something they are willing to lose if they perjure themselves, and by something to ensure that they lose it in such an event.Footnote ⁶ A lover swears by his own hair,Footnote ⁷ and expects to lose it if the oath is broken; he does not swear by the hair of a divine witness. This conception of the lover's oath is most fully worked out by Ovid in Amores 3.3, where it is parodied extensively. There his mistress, having first sworn an oath by all of her most attractive qualities (3–4 long hair, 5–6 rosy cheeks, 7–8 dainty feet, 9–10 sparkling eyes) and then broken it, is permitted—contrary to expectation—not to lose but to keep them, whereas Ovid, because she swore by his eyes also, feels the hurt in his own. When in Am. 3.3.13–14 he writes perque suos illam nuper iurasse recordor | perque meos oculos: en doluere mei!, he is parodying such lines as Prop. 1.15.35–6 hos [sc. ocellos] tu iurabas, si quid mentita fuisses, | ut tibi suppositis exciderent manibus.Footnote ⁸ And when he proceeds to satirize the calling of gods to witness, he concentrates not on their hair but, like Tibullus in 1.4.25, on their weapons (Am. 3.3.27–30):

This passage would seem to controvert Dissen's earlier assertion: in elegy the gods very much do exact punishments using the weapons on which elegiac lovers swear. Such instruments include the bow of Cupid (Ov. Am. 2.7.27–8; Her. 7.157), the torches of Venus (Her. 3.39–40), the arrows of Diana (Fast. 2.157), and, in a different context, the thunderbolt of Jupiter (Verg. Aen. 12.200; Stat. Silu. 3.1.186) and bow of Apollo (Anth. Lat. 199.11 R). Minerva's hair is not only unparalleled as an attribute on which to swear an oath; it is not even of the same class as those attributes which are actually sworn upon.Footnote ⁹ Hence Merli, the only modern scholar besides Luck who seems to acknowledge the oddness of crines, is moved to regard it as intentionally humorous.Footnote ¹⁰ Swearing by the hair of Minerva—upon this theory—is a behaviour illustrative of the ‘foolish love’ of lines 23–4: the lover in his eagerness swears upon whatever first comes to mind (24 iurasset cupide quicquid ineptus amor), and the first thing that here comes to mind, after the arrows of Dictynna, is the hair of Minerva. But with Merli's explanation come two problems. First, if Minerva's hair is acknowledged to be a ridiculous attribute to swear by, then the oath by which it is sworn ought to be accounted as weak or worthless (cf. Ov. Rem. am. 783–4). But if a lover's oath were weak or worthless because he had picked a ridiculous attribute for a witness, Jupiter would have no reason to intervene, and the special exception he makes for lovers (23–4) would not be in request. In truth, the ‘foolishness’ of love consists, as commentators have long remarked, in choosing to swear by two goddesses who by reason of their own castitas uirginalis would be least disposed to forgive the breaking of a lover's oath—not in swearing by an abnormal attribute of one of those two goddesses.Footnote ¹¹ Second, if it were Tibullus’ intention to index some of the strange attributes on which heedless lovers swear, one wonders why he did not assign an equally irreverent item to Diana (her crura, for example).

crines has been explained in other ways. Acknowledging that a parallel to sagittas is required to give force to impune in line 25, Huschke wondered if crines might be a metonym for the plume of Minerva's helmet.Footnote ¹² This is quite impossible. crinis by itself sometimes signifies the material from which a plume is made but never the crest of the helmet itself (cf. E. Lommatzsch, TLL 4.1202.78–1203.2). Moreover, this special use of crinis is usually clarified by context, as in Sil. Pun. 5.133–4 cui uertice surgens | triplex crista iubas effundit crine Sueuo, where the hair clearly belongs to a plumed helmet. Finally, it is not easy to see how a plume might be put to the use of punishing a liar. This last objection holds for Mitscherlich's cristas and for all other conjectures so far made upon this passage.Footnote ¹³

The fact is that, if Tibullus wished to specify a genuinely threatening attribute of Minerva, he would have written hastam … suam.Footnote ¹⁴ That he did is entirely possible. Not only is Minerva's spear a weapon like Diana's arrows; it is also the only attribute of hers which I have found persons in Latin poetry to swear by (cf. Juv. 13.82), and, if read here, would enable line 26 to constitute the line parodied at Ov. Am. 3.3.28.Footnote ¹⁵ The distance travelled by crines from the letters of hastam is contrasted by its closeness to those of the tail-end of affirmes. If a copyist committed an error of dittography, writing, say, affirmes firmes for affirmes hastam, a later scribe could well have changed this into affirmes crines (m = ın), writing suos for suam as a matter of course. Such errors are not uncommon in Tibullus’ manuscripts: cf. 1.1.50 terre ferre (cod. Parisinus lat. 8233) in error for tristes ferre (Z+), 1.4.42 arenti ṭọṛṛẹṇṭị torreat (V) in error for arenti torreat (AGX+), 1.5.11 sulpure puro (Z+) in error for sulpure uiuo (Heinsius, Broukhusius), 1.10.14 telit gerit (H) for tela gerit (Z+).Footnote ¹⁶ The anonymous reader suggests, alternatively, that hastam was first corrupted to castam, owing to the frequent collocation of this adjective with Minerva (cf. Prop. 3.20.7; Ov. Am. 1.7.18; Ilias Latina 78, 333, 532, 894), and then deliberately emended by a scribe to crines.

II

Tibullus explains the origins of a festival (2.1.51–6):

‘The farmer, wearied by constant ploughing, was first to chant rustic words in regular metre, first to play on the oaten pipe a song to sing before decorated gods. And the farmer, drenched with ruddy cinnabar, Bacchus, first led the dance with inexpert skill.’ To what speaker should be attributed the sudden exclamation of Bacche in line 55? To the farmer? But there is no ait or inquit to mark the attribution. Then to Tibullus? Cries of (euhoe) Bacche are the sort of thing one expects to hear from a Dionysian reveller, and a Dionysian revel is indeed what many believe to be here under discussion.Footnote ¹⁷ But Tibullus does not pose as a participant of this festival; he poses as its narrator, and it is not in the style of a narrator to make strange and wild ejaculations of this kind. True, Bacchus is invoked along with Ceres at the beginning of this elegy. But there his presence is welcomed by a second-person pronoun and imperative (3–4):

Here he gets no such fanfare. Bacche is part of no prayer, no earnest exhortation. When Hartman enquires ‘cur vocativus ille Bacche inseritur?’, I do not know how to answer him.Footnote ¹⁸

Whenever Tibullus pronounces a name in the vocative case, he always follows it with a verb or a pronoun of the first or second person.Footnote ¹⁹ The spontaneous apostrophe of Bacchus in line 55 differs from every other address in Tibullus in this particular. No first- or second-person verbs or pronouns are present to bring him into conversation either with Tibullus or with the dancing farmer (for example ‘Bacchus, the farmer leads your dance’), nor is any effect gained by their absence. ‘Intempestiue’, says Statius, ‘Bacchum videtur appellare. Mihi aut aliquid deesse, aut vitio scriptum videbatur.’Footnote ²⁰ If bache (thus MSS AV) is corrupt, the number of possible parts of speech which it could have supplanted is not great. It may hide an adverb qualifying rubentem, such as suaue or, what is much closer to the letters, dulce.Footnote ²¹ But take bache out of the text entirely, and the one thing that appears most to be missing from this sentence is a part of the body in respect of which the farmer has smeared himself with cinnabar. An accusative of respect is so used with suffusus in Ov. Met. 11.368 and [Ov.] Hal. 123; cf. also Ciris 505–6 minioque infecta rubenti | crura. So, where upon their bodies did the holidaying Romans of antiquity apply the tincture of cinnabar? Pliny the Elder (HN 33.111), citing Verrius Flaccus, reports that in old times the Romans used minium to colour the ‘face of a statue of Jupiter’ (Iouis ipsius simulacri faciem diebus festis minio inlini solitam) and the ‘bodies’ of persons going in triumphal procession (triumphantiumque corpora).Footnote ²² Though he claims to be unable to see the origin of this custom (112 cuius rei causam equidem miror), possibly it bears some relation to that practice of painting red the faces of agricultural gods, a custom alluded to not only by Tibullus in line 54 (ornatos … deos) but also by Virgil in Ecl. 6.21–2 Aegle, Naiadum pulcherrima, | iamque uidenti sanguineis frontem moris et tempora pingit and 10.26–7 Pan deus Arcadiae uenit, quem uidimus ipsi | sanguineis ebuli bacis minioque rubentem. Servius and Servius Danielis's comments on these passages are also redolent of Verrius’ dictionary: Verg. Ecl. 6.22 multi ob hoc dictum putant (cf. Plin. HN 33.111 enumerat auctores Verrius), quod robeus color deorum sit: unde et triumphantes facie miniata, et in Capitolio Iuppiter in quadrigis miniatus; 10.27 facie rubra pingitur Pan propter aetheris similitudinem: aether autem est Iuppiter. unde etiam triumphantes … faciem quoque de rubrica inlinunt instar coloris aetherii. So, red faces and red bodies. Since neither ora nor corpora will scan, the best lection according to this line of reasoning would surely be membra. Mention of the farmer's ‘limbs’ (or ‘limbs as comprising the whole body’, OLD s.v. 2) leads neatly into the description of dancing in line 56, membra mouere being a common synonym of this (cf. Tib. 1.7.38; Lucr. 4.980; Hor. Sat. 1.9.24–5; Ov. Fast. 6.677–8). The manuscripts’ membra may have been banalized first to brachia, then corrected to bache for metre's sake;Footnote ²³ else it was abbreviated first to m̃bra, and then expanded by a generous hand into bache, perhaps owing to a scribe's familiarity with Verg. Ecl. 10.27 bacis minioque rubentem, or under the influence of ab arte in the line below (for example suffusus membra rubenti → suffusus ab arte rubenti → suffusus bache rubenti). For membra + perfusus + ablative there is a parallel in Luc. 8.735 perfusus sanguine membra.Footnote ²⁴

III

Tibullus agrees to drink a concoction of magic herbs and drugs, if Nemesis will look kindly on him (2.4.55–60):

When a list of particulars is closed by a noun qualified by mille alii, these words always specify the class to which the items on said list belong, never further items from that class: Sen. Contr. 3.16 laqueus, gladius, praeceps locus, uenenum, naufragium, mille aliae mortes; Celsus, Med. 1.pr.41 metu, dolore, inedia, cruditate, lassitudine, mille aliis mediocribus adfectibus, 3.21.2 famem, sitim, et mille alia taedia; Sen. Dial. 9.5.3 petulantiam, inuidiam, mille alia inertia uitia. We should therefore cast a second glance at herbas in line 60, since it appears to violate this pattern by failing to constitute a class able to comprehend all the items on its list; for uenenum (55) and hippomanes (57–8) are not ‘herbs’ with which a thousand ‘others’ (alias) may be mixed.Footnote ²⁵ A second reason to suspect the text of this passage is its ambiguous syntax. Murgatroyd remarks that quicquid … ueneni (55), quicquid … herbarum (56) and Hippomanes (57–8) can be taken as objects either of misceat (along with mille alias herbas in asyndeton) or of bibam (mille alias herbas misceat illa standing in parenthesis).Footnote ²⁶ The grammar could be disambiguated by making mille alias herbas into an ablative or dative noun phrase, mille aliis herbis, and thus tying the ingredients of lines 55–8 only to misceat: ‘let her mix poison, herbs and hippomanes with a thousand other herbs: provided my Nemesis looks on me kindly, I shall drink.’ But merely changing the case will not solve our first problem: a more generic noun than herbas/-is is required to comprehend all the ingredients of lines 55–8. Sprengel was the first to see that an ablative was needed here; he was also one of the first to notice the need for a noun other than herbis.Footnote ²⁷ His philtris however will not do the job: uenenum, herba and hippomanes are related to philtrum not as species to genus (as in the four examples of mille alii given at the start of this note) but as simples to a compound. A philtre is what Nemesis is making; one does not brew a philtre out of ‘poison, herbs, hippomanes and a thousand other philtres’. Changing alias to a different adjective, such as Valckenaer's malas herbas (or malis herbis), is a solution still less attractive, since it would leave intact the unimaginative repetition of herbas after herbarum in line 56 at the expense of removing an interesting and well-established idiom in mille alias/-is.Footnote ²⁸ Heyne's malos succos (or aliis succis) is better, but strays further from the paradosis than one would like, and it may be doubted whether ‘juice’ is sufficiently generic to comprehend the Thessalian herbs of line 56.Footnote ²⁹ All these emendations are in any case vitiated by the un-Tibullian homoeoteleuton in -īs, -ās or -ōs.Footnote ³⁰ Since the true reading is yet to seek, I conjecture aliis rebus.Footnote ³¹ Ovid employs this very phrase to close the list of ingredients of Medea's magic potion in Met. 7.262–76; cf. lines 273–6:

A variant of this expression, in which alius is used as a neuter substantive (mille alia, ‘a thousand other things’), appears in Livy 29.18.7, Quint. Inst. 1.4.13, 1.6.25, 2.15.23, 9.3.1, and Apul. Apol. 54. This would suggest that mille aliis rebus is a weightier version of a common idiom, and not therefore a weak or colourless phrase on which to end the catalogue.Footnote ³² Further support for rebus is furnished by Manilius 5.468 mille alias rerum species in carmina ducent, where the superfluity of rerum is scarcely less objectionable than the seeming plainness of rebus. How this turned into herbas is not hard to imagine. A scribe with herbarum fluttering before his brain wrote aliis herbis when he meant to say aliis rebus; this was later changed into alias herbas by a copyist who losing track of the syntax sought an object for misceat.