¹ All known MSS. read sunt here. Merrill - the only editor, I believe, who retains this reading – is thereby driven to conjecture insinuanda, positing a most unlikely corruption. I prefer Lambinus' and Marullus' conjecture tune to Lachmann's tumst, which gives identical sense, for the following reasons. Lachmann's objection to tune, which must be inferred from his general remarks ad DRN 1.111, that Lucretius omits the copula only in the presence of the infinitive esse or one of its compounds, is not compelling. Bailey, in his editio maior (Oxford, 1947)Google Scholar states that ‘this arbitrary principle is now rightly rejected by editors’ (p. 103), but he and most others nevertheless follow Lachmann by printing tumst at 5.44. Unfortunately, this conjecture introduces some unusual prosodic elements. While Lucretius does not hesitate to elide most monosyllables, such as the conjunction cum, he generally avoids aphaeresis involving monosyllables: for details, see Soubiran, J., L'Elision dans la poésie latine (Paris, 1966), pp. 393–4Google Scholar. Indeed, I can find no occurrence of tumst in Latin poetry from Plautus to Ovid. Furthermore, Lucretius does not like to use tum before any word that begins with a vowel; the only example I can quote is 5.855, multaque turn interiisse animantum saecla necessest. Conversely, tune, which is much less frequent than turn in Lucretius, does not normally appear except before a vowel. Finally, the sequencetune…turn in consecutive lines (44–5), which some might have thought inelegant, can be paralleled in careful writers (in the same line: Tibullus 1.4.53, Ovid, AA 1.239, Met. 12.526; in consecutive lines: Vergil, G. 1.136–7, Propertius 1.7.21–2, Ovid, AA 1.239–41 quater, 2.321–2, Met. 12.445–6). Note too that in most of these passages tune stands before a vowel.

² Munro, Vol. 3, translation (Cambridge, 1886⁴), p. 117.

³ Bailey, , Lucretius on the Nature of Things (Oxford, 1910), p. 187Google Scholar.

⁴ ‘On Lucretius V. 43 sq.’ CR 19 (1905), 205–6Google Scholar.

⁵ ‘Notes and Emendations in Latin Poets’, CQ 5 (1911), 104–7CrossRefGoogle Scholar. The first edition of Bailey's translation appeared the year before Hardie's paper; Bailey probably first learned of Cole's paper from Merrill's, W. A. commentary (New York, 1907), ad he.Google Scholar Hardie does not mention Cole at all.

⁶ Bailey (1947, p. 1328) begins his discussion by citing Munro's rendering, which Cole (p. 205) also calls ‘the interpretation generally accepted for these lines'; Hardie (p. 105) refers to Munro, to Bailey's first translation, and to J. D. Duff's school edition of Book 5 (Cambridge, 1889).

⁷ Pius: ‘proelia. motus animi. τ πθη: perturbationes scilicet quando in iecore aegro nascuntur domini. sunt insinuandum. indefinitum pro insinuanda: imprimenda et infigenda aio hominibus.’

⁸ Lambinus: ‘id est, quae tune proelia inire, quaeque pericula adire nos oportet, vel invitos? insinuandum: subintellige, est: id est, insinuare, et subire nos oportet.’

⁹ On Pius' ( ≃Munro's) side may be placed Marolles (Paris, 1659), Marchetti (London, 1717), Le Grange (Paris, 1768), Knebel (Leipzig, 1831), Bossart (Berlin, 1865) and Trevelyan (Cambridge, 1937); on that of Lambinus (≃ Cole and Hardie), Paré (Paris, 1631), Le Févre (Saumur, 1662), Du Fay (Paris, 1680), Havercamp (Leiden, 1725), Meinecke (Leipzig, 1795), Wakefield (London, 1796), Lemaire (Paris, 1838), Binder (Stuttgart, 1868–9), Ernout, (Paris, 1920) and Diels (Berlin, 1924). Not all interpreters fall squarely into one or the other camp: e.g. Creech's rather liberal translation (Oxford, 1682) speaks of ‘civil wars’ and of passions that ‘vex the mind', which seems to me closer to the former point of view; but in his notes ad loc. (Oxford, 1695)Google Scholar he closely paraphrases Lambinus.

¹⁰ The renderings of W. H. Brown (New Brunswick, NJ, 1950), R. E. Latham (Harmondsworth, 1951), L. L. Johnson (Fontwell, Sussex, 1963), R. Humphries (Bloomington, IN, 1968) all agree with that of Munro. Bailey is followed, however, by the most recent commentator on DRN 5, C. D. N. Costa (Oxford, 1984). Likewise the new Loeb editor, M. F. Smith, has altered W. H. D. Rouse's rendering, ‘…what battles and perils must we then let in, will we or no!’ (edd. 1–3, Cambridge, MA, 1924, 1928, 1937) to ‘…what battlefields and perils must we then find our way into…’ (ed. 4, 1975). (Smith's own translation [London, 1969] is less committed: ‘…what strifes and dangers we incur against our will'.) Pizzani, U. (in the commentary ad loc. included in Lucreti De Rerum Natura locos praecipue notabiles collegil et illustravit H. Paratore [Roma, 1960])Google Scholar succinctly states the case for both sides and, without mentioning Cole, Hardie, or Bailey, decides in favour of their interpretation.

¹¹ Aalto, P., Untersuchungen über das lateinische Gerundium und Gerundivum (Helsinki, 1949), pp. 92–8Google Scholar.

¹² Risch, E., Gerundium und Gerundivum, Gebrauch im klassischen und älteren Latein, Entstehung und Vorgeschkhte (Berlin and New York, 1984), pp. 110–12 and 186–8CrossRefGoogle Scholar. Cf. Hofmann, J. B. and Szantyr, A., Lateinische Syntax und Stylistik (München, 1963), pp. 371–3Google Scholar (§202 C, a), who elsewhere (p. 369, §201, Zusatz α) admit the difficulty of deciding with certainty whether the form is a gerund or a gerundive.

¹³ For a fairly full list of passages in other authors with forms in -ndum+ accusative + esse see Aalto (above, note 11), loc. cit., with the cautionary remarks of Risch (above, note 12), p. 186.

¹⁴ DRN 1.111, 138, 381, 627, 963; 2.39, 492, 1129; 3.391, 626, 796; 4.778; 5.44, 302; 6.917.

¹⁵ The consistency of these ideas is apparent in Lucretius' diction: cf. 3.492—4 (turbat, 493), 4.337 42 (purgat, 341), 6.374–8 (turbatur, 377), 1119–24 (conturbat, 1121). The same imagery is found at 3.34–40 applied to clear skies, a metaphor for the tranquillity of mind experienced by the Epicurean sage.

¹⁶ He uses it to suggest the way in which lightning (6.89), heat (6.234, 355 bis, 385, 860) and magnetism (6.1032) pass into and through solid objects; to connote the mind's penetration of difficult intellectual matter (1.409); the effect of sensory stimuli upon the body (2.436,684; 4.331, 525; 6.778, 802; morbida vis, 6.955); the occupation of bodies by souls, according to benighted rival theories (1.113, 116, 3.671, 689, 698, 722, 729, 738, 780, 782).

¹⁷ The word sinus in this sense refers primarily to the fold in the toga that draped over the chest, which was used as a pocket; but this meaning easily passes over into ‘breast, chest, or bosom’ (e.g. Ovid, Epist. 1.45; Tacitus, Hist. 3.10.4).

¹⁸ Vergil borrows this idea in Aeneid 2.228–9,

tum vero tremefacta novus per pectora cunctis

insinuat pavor,

and Knox, Bernard M. W. (‘The Serpent and the Flame. The Imagery of the Second Book of the Aeneid’, AJP 71 [1950], 384)Google Scholar finds in this passage a ‘reminiscence’ of DRN 5.73. He also calls insinuare ‘one of Lucretius’ favorite words', observing that ‘Virgil uses it only here’ (cf. the similar remarks of Austin, R. G., P. Vergili Maronis Aeneidos Liber Sextus [Oxford, 1977]Google Scholar, ad loc). It is certainly not common before Lucretius (only three examples in Plautus, Cist. 89, 92 and Mil. 105). Later poetic occurrences (Propertius 3.9.28; Statius, Theb. 5.448, 7.110, Silv. 2.1.234; Rutilius Namatianus 1.590; cf. Manilius 4.604 [conjectured by Scaliger]) show no awareness of this etymology.

¹⁹ Insinuare appears commonly in prose, normally with the meaning ‘t o wind one's way through or into’ by an indirect route (as clearly at DRN 6.1032; cf. OLD s.v. sinuare). Although the verb occurs in several passages that concern battles (e.g. Caesar, Gal. 4.33; Livy 44.41.8), in all it means ‘t o work one's way through (the ranks or the mêlée)'; it is not equivalent to manum conserere, and such passages should not be adduced as parallels to DRN 5.44.

²⁰ See note 16 above.

²¹ Cf. Pius' gloss on proelia (quoted above, note 7); Cicero, Tusc. 4.10: ‘quae Graeci παθη vocant nobis “perturbationes” appellari magis placet quam “morbos”’.

²² This point is forcefully made by West, D., The Imagery and Poetry of Lucretius (Edinburgh, 1969), pp. 1–10Google Scholaret passim; see also Davies, H. Sykes, ‘Notes on Lucretius’, The Criterion 11 (1931), 25–42Google Scholar.

²³ DRN 3.485, 782,4.331.

²⁴ Passive insinuare is found without agent at DRN 3.780, 782, 6.355, 955.

²⁵ See DRN 1.111, 138, 381, 963, 3.391, 626, 796, 5.302, 6.917. An agent appears with this construction only at 1.627 and 4.777–8.

²⁶ There are two examples, both, like 5.44, involving the dative with a compound verb: 2.39, animo prodesse putandum; 2A\28–9, fluere atque recedere rebus / multa manus dandum est. At 2.492 there is perhaps an ellipse of the expected dative in addendum partis alias erit.

²⁷ There has certainly been a pronounced strain of ambiguity, apparently intentional, as well as of obscurity and indecision among commentators and translators from almost the very beginning. Le Grange (Paris, 1768) manages to embrace both interpretations in his version: ‘Mais si nos coeurs ne sont delivres des vices, que de combats interieurs a soutenir! que de perils a vaincre!’ P. H. Wooby (New York, 1973) tends rather less elegantly in the same direction: ‘But if our heart is not purified, what risks and battles, / Then, must infiltrate us dissatisfied, which we must enter.’

This paper was written in order to answer a question put to me by two students, Eric Kyllo and David Rich, with whom I once had the pleasure of reading Lucretius in a graduate seminar. A preliminary version of it was discussed with a group of colleagues and students at the University of Pennsylvania during Spring, 1987, as part of our series of talks on work-in-progress. Several improvements were suggested by the editors of CQ and the anonymous referee. My thanks to all of these benefactors. Any faults that remain are, of course, my own responsibility.